Dynomotion

Group: DynoMotion Message: 4063 From: himykabibble Date: 2/23/2012
Subject: v429z Problem
Tom,

At this point, there is not a doubt in my mind that v429z is broken. I have run dozens of tests, and v428 always works, v429z always fails. With v429z, the lower I set acceleration, or feedrate, in KMotionCNC, the worse it performs. With accel set to 25 in/sec^2, I get small, but absolutely consistent, position errors, and NO servo faults. With accel set to 0.1 in/sec^2, I get nearly constant servo faults, with attendant massive position loss. The exact same g-code will work perfectly time after time with v428, with acceleration set as low or as high as I choose, and with feed rates cranked up to near rapid speeds. Even worse, it will sometimes move the Z axis, and there is not a single Z-axis command anywhere in the G-code! On the crooked, non-straight segments, I can put my fingers on the ballscrews, and I can feel clear, smooth motion on the axis that is supposed to be stationary. No chance this is noise - these are deliberate, constant speed moves, always in exactly the same place, and they ARE reflected in the DROs.

I've posted the G-code I'm using in RayLs Files.

Regards,
Ray L.
Group: DynoMotion Message: 4064 From: himykabibble Date: 2/23/2012
Subject: Re: v429z Problem
Tom,

I'm posting some captures of ch[0-2].Dest for versions v428 and v429z. My machine was not even turned on for this. The problem is quite evident.

Regards,
Ray L.

--- In DynoMotion@yahoogroups.com, "himykabibble" <jagboy@...> wrote:
>
> Tom,
>
> At this point, there is not a doubt in my mind that v429z is broken. I have run dozens of tests, and v428 always works, v429z always fails. With v429z, the lower I set acceleration, or feedrate, in KMotionCNC, the worse it performs. With accel set to 25 in/sec^2, I get small, but absolutely consistent, position errors, and NO servo faults. With accel set to 0.1 in/sec^2, I get nearly constant servo faults, with attendant massive position loss. The exact same g-code will work perfectly time after time with v428, with acceleration set as low or as high as I choose, and with feed rates cranked up to near rapid speeds. Even worse, it will sometimes move the Z axis, and there is not a single Z-axis command anywhere in the G-code! On the crooked, non-straight segments, I can put my fingers on the ballscrews, and I can feel clear, smooth motion on the axis that is supposed to be stationary. No chance this is noise - these are deliberate, constant speed moves, always in exactly the same place, and they ARE reflected in the DROs.
>
> I've posted the G-code I'm using in RayLs Files.
>
> Regards,
> Ray L.
>
Group: DynoMotion Message: 4065 From: himykabibble Date: 2/23/2012
Subject: Re: v429z Problem
OK, three spreadsheets uploaded to RayLs Files. The v428 trace looks the same, regardless of parameter setting. The v429 trace gets goofier the lower acceleration is set. Two captures, for accel = 30.0, and accel = 0.10 are given. All captures were recording Dest on every 50th pass of my Init program. First column is time, second thru fourth are X, Y, Z position.

Regards,
Ray L.

--- In DynoMotion@yahoogroups.com, "himykabibble" <jagboy@...> wrote:
>
> Tom,
>
> I'm posting some captures of ch[0-2].Dest for versions v428 and v429z. My machine was not even turned on for this. The problem is quite evident.
>
> Regards,
> Ray L.
>
> --- In DynoMotion@yahoogroups.com, "himykabibble" <jagboy@> wrote:
> >
> > Tom,
> >
> > At this point, there is not a doubt in my mind that v429z is broken. I have run dozens of tests, and v428 always works, v429z always fails. With v429z, the lower I set acceleration, or feedrate, in KMotionCNC, the worse it performs. With accel set to 25 in/sec^2, I get small, but absolutely consistent, position errors, and NO servo faults. With accel set to 0.1 in/sec^2, I get nearly constant servo faults, with attendant massive position loss. The exact same g-code will work perfectly time after time with v428, with acceleration set as low or as high as I choose, and with feed rates cranked up to near rapid speeds. Even worse, it will sometimes move the Z axis, and there is not a single Z-axis command anywhere in the G-code! On the crooked, non-straight segments, I can put my fingers on the ballscrews, and I can feel clear, smooth motion on the axis that is supposed to be stationary. No chance this is noise - these are deliberate, constant speed moves, always in exactly the same place, and they ARE reflected in the DROs.
> >
> > I've posted the G-code I'm using in RayLs Files.
> >
> > Regards,
> > Ray L.
> >
>
Group: DynoMotion Message: 4066 From: himykabibble Date: 2/23/2012
Subject: Re: v429z Problem
BTW - I can duplicate this with two different KFlops, and two different PCs....

--- In DynoMotion@yahoogroups.com, "himykabibble" <jagboy@...> wrote:
>
> OK, three spreadsheets uploaded to RayLs Files. The v428 trace looks the same, regardless of parameter setting. The v429 trace gets goofier the lower acceleration is set. Two captures, for accel = 30.0, and accel = 0.10 are given. All captures were recording Dest on every 50th pass of my Init program. First column is time, second thru fourth are X, Y, Z position.
>
> Regards,
> Ray L.
>
> --- In DynoMotion@yahoogroups.com, "himykabibble" <jagboy@> wrote:
> >
> > Tom,
> >
> > I'm posting some captures of ch[0-2].Dest for versions v428 and v429z. My machine was not even turned on for this. The problem is quite evident.
> >
> > Regards,
> > Ray L.
> >
> > --- In DynoMotion@yahoogroups.com, "himykabibble" <jagboy@> wrote:
> > >
> > > Tom,
> > >
> > > At this point, there is not a doubt in my mind that v429z is broken. I have run dozens of tests, and v428 always works, v429z always fails. With v429z, the lower I set acceleration, or feedrate, in KMotionCNC, the worse it performs. With accel set to 25 in/sec^2, I get small, but absolutely consistent, position errors, and NO servo faults. With accel set to 0.1 in/sec^2, I get nearly constant servo faults, with attendant massive position loss. The exact same g-code will work perfectly time after time with v428, with acceleration set as low or as high as I choose, and with feed rates cranked up to near rapid speeds. Even worse, it will sometimes move the Z axis, and there is not a single Z-axis command anywhere in the G-code! On the crooked, non-straight segments, I can put my fingers on the ballscrews, and I can feel clear, smooth motion on the axis that is supposed to be stationary. No chance this is noise - these are deliberate, constant speed moves, always in exactly the same place, and they ARE reflected in the DROs.
> > >
> > > I've posted the G-code I'm using in RayLs Files.
> > >
> > > Regards,
> > > Ray L.
> > >
> >
>
Group: DynoMotion Message: 4067 From: Tom Kerekes Date: 2/23/2012
Subject: Re: v429z Problem
Hi Ray,
 
I think I see what is going on.  I re-plotted your data and uploaded the Excel Sheet.   Looks like there is a problem with Arcs getting out of order with Line segments due to the Grouping into blocks of segments that we did in V4.29 to improve USB traffic.  Everything works fine for Albert (and me) because we weren't doing any Arcs.  Give me some time to come up with a fix.
 
Looks like it wasn't noise :}
 
Regards
TK

Group: DynoMotion Message: 4068 From: himykabibble Date: 2/23/2012
Subject: Re: v429z Problem
Tom,

Well, how many people really NEED to do arcs anyway.... :-)

Regards,
Ray L.

--- In DynoMotion@yahoogroups.com, Tom Kerekes <tk@...> wrote:
>
> Hi Ray,
>  
> I think I see what is going on.  I re-plotted your data and uploaded the Excel Sheet.   Looks like there is a problem with Arcs getting out of order with Line segments due to the Grouping into blocks of segments that we did in V4.29 to improve USB traffic.  Everything works fine for Albert (and me) because we weren't doing any Arcs.  Give me some time to come up with a fix.
>  
> Looks like it wasn't noise :}
>  
> Regards
> TK
>
> From: himykabibble <jagboy@...>
> To: DynoMotion@yahoogroups.com
> Sent: Thursday, February 23, 2012 12:05 PM
> Subject: [DynoMotion] Re: v429z Problem
>
>
>  
> BTW - I can duplicate this with two different KFlops, and two different PCs....
>
> --- In DynoMotion@yahoogroups.com, "himykabibble" <jagboy@> wrote:
> >
> > OK, three spreadsheets uploaded to RayLs Files. The v428 trace looks the same, regardless of parameter setting. The v429 trace gets goofier the lower acceleration is set. Two captures, for accel = 30.0, and accel = 0.10 are given. All captures were recording Dest on every 50th pass of my Init program. First column is time, second thru fourth are X, Y, Z position.
> >
> > Regards,
> > Ray L.
> >
> > --- In DynoMotion@yahoogroups.com, "himykabibble" <jagboy@> wrote:
> > >
> > > Tom,
> > >
> > > I'm posting some captures of ch[0-2].Dest for versions v428 and v429z. My machine was not even turned on for this. The problem is quite evident.
> > >
> > > Regards,
> > > Ray L.
> > >
> > > --- In DynoMotion@yahoogroups.com, "himykabibble" <jagboy@> wrote:
> > > >
> > > > Tom,
> > > >
> > > > At this point, there is not a doubt in my mind that v429z is broken. I have run dozens of tests, and v428 always works, v429z always fails. With v429z, the lower I set acceleration, or feedrate, in KMotionCNC, the worse it performs. With accel set to 25 in/sec^2, I get small, but absolutely consistent, position errors, and NO servo faults. With accel set to 0.1 in/sec^2, I get nearly constant servo faults, with attendant massive position loss. The exact same g-code will work perfectly time after time with v428, with acceleration set as low or as high as I choose, and with feed rates cranked up to near rapid speeds. Even worse, it will sometimes move the Z axis, and there is not a single Z-axis command anywhere in the G-code! On the crooked, non-straight segments, I can put my fingers on the ballscrews, and I can feel clear, smooth motion on the axis that is supposed to be stationary. No chance this is noise - these are deliberate, constant speed
> moves, always in exactly the same place, and they ARE reflected in the DROs.
> > > >
> > > > I've posted the G-code I'm using in RayLs Files.
> > > >
> > > > Regards,
> > > > Ray L.
> > > >
> > >
> >
>
Group: DynoMotion Message: 4070 From: himykabibble Date: 2/23/2012
Subject: Re: v429z Problem
Tom,

Does this also explain the uncommanded moves I've seen? That one cost me a good test indicator this AM, when I did a "G0X0Y0" and the KFlop decided to also jam the quill (Z) into the table at full rapid speed....

Regards,
Ray L.

--- In DynoMotion@yahoogroups.com, "himykabibble" <jagboy@...> wrote:
>
> Tom,
>
> Well, how many people really NEED to do arcs anyway.... :-)
>
> Regards,
> Ray L.
>
> --- In DynoMotion@yahoogroups.com, Tom Kerekes <tk@> wrote:
> >
> > Hi Ray,
> >  
> > I think I see what is going on.  I re-plotted your data and uploaded the Excel Sheet.   Looks like there is a problem with Arcs getting out of order with Line segments due to the Grouping into blocks of segments that we did in V4.29 to improve USB traffic.  Everything works fine for Albert (and me) because we weren't doing any Arcs.  Give me some time to come up with a fix.
> >  
> > Looks like it wasn't noise :}
> >  
> > Regards
> > TK
> >
> > From: himykabibble <jagboy@>
> > To: DynoMotion@yahoogroups.com
> > Sent: Thursday, February 23, 2012 12:05 PM
> > Subject: [DynoMotion] Re: v429z Problem
> >
> >
> >  
> > BTW - I can duplicate this with two different KFlops, and two different PCs....
> >
> > --- In DynoMotion@yahoogroups.com, "himykabibble" <jagboy@> wrote:
> > >
> > > OK, three spreadsheets uploaded to RayLs Files. The v428 trace looks the same, regardless of parameter setting. The v429 trace gets goofier the lower acceleration is set. Two captures, for accel = 30.0, and accel = 0.10 are given. All captures were recording Dest on every 50th pass of my Init program. First column is time, second thru fourth are X, Y, Z position.
> > >
> > > Regards,
> > > Ray L.
> > >
> > > --- In DynoMotion@yahoogroups.com, "himykabibble" <jagboy@> wrote:
> > > >
> > > > Tom,
> > > >
> > > > I'm posting some captures of ch[0-2].Dest for versions v428 and v429z. My machine was not even turned on for this. The problem is quite evident.
> > > >
> > > > Regards,
> > > > Ray L.
> > > >
> > > > --- In DynoMotion@yahoogroups.com, "himykabibble" <jagboy@> wrote:
> > > > >
> > > > > Tom,
> > > > >
> > > > > At this point, there is not a doubt in my mind that v429z is broken. I have run dozens of tests, and v428 always works, v429z always fails. With v429z, the lower I set acceleration, or feedrate, in KMotionCNC, the worse it performs. With accel set to 25 in/sec^2, I get small, but absolutely consistent, position errors, and NO servo faults. With accel set to 0.1 in/sec^2, I get nearly constant servo faults, with attendant massive position loss. The exact same g-code will work perfectly time after time with v428, with acceleration set as low or as high as I choose, and with feed rates cranked up to near rapid speeds. Even worse, it will sometimes move the Z axis, and there is not a single Z-axis command anywhere in the G-code! On the crooked, non-straight segments, I can put my fingers on the ballscrews, and I can feel clear, smooth motion on the axis that is supposed to be stationary. No chance this is noise - these are deliberate, constant speed
> > moves, always in exactly the same place, and they ARE reflected in the DROs.
> > > > >
> > > > > I've posted the G-code I'm using in RayLs Files.
> > > > >
> > > > > Regards,
> > > > > Ray L.
> > > > >
> > > >
> > >
> >
>
Group: DynoMotion Message: 4072 From: Tom Kerekes Date: 2/23/2012
Subject: Re: v429z Problem
I don't think so.
 
TK

Group: DynoMotion Message: 4084 From: Tom Kerekes Date: 2/25/2012
Subject: Re: v429z Problem
Ray,
 
Please try this patch to fix GCode Arcs for V4.29z.  Copy to the C:\Kmotion429z\Kmotion\Release directory
 
 
Regards
TK
 

Group: DynoMotion Message: 4091 From: himykabibble Date: 2/25/2012
Subject: Re: v429z Problem
Tom,

I was only able to run a quick test, but it appears to be working correctly now. I'll try to run some real parts tomorrow, once I get my machine back together (finally installing limit switches on X, Y and A axes).

Can I get the modified source, as I need to make my tool length comp mods before I can run any of my existing G-code.

Regards,
Ray L.

--- In DynoMotion@yahoogroups.com, Tom Kerekes <tk@...> wrote:
>
> Ray,
>  
> Please try this patch to fix GCode Arcs for V4.29z.  Copy to the C:\Kmotion429z\Kmotion\Release directory
>  
> http://dynomotion.com/Software/Patch/FixTP429Arcs/GCodeInterpreter.dll
>  
> Regards
> TK
>  
>
> From: himykabibble <jagboy@...>
> To: DynoMotion@yahoogroups.com
> Sent: Thursday, February 23, 2012 2:25 PM
> Subject: [DynoMotion] Re: v429z Problem
>
>
>  
> Tom,
>
> Well, how many people really NEED to do arcs anyway.... :-)
>
> Regards,
> Ray L.
>
> --- In DynoMotion@yahoogroups.com, Tom Kerekes <tk@> wrote:
> >
> > Hi Ray,
> >  
> > I think I see what is going on.  I re-plotted your data and uploaded the Excel Sheet.   Looks like there is a problem with Arcs getting out of order with Line segments due to the Grouping into blocks of segments that we did in V4.29 to improve USB traffic.  Everything works fine for Albert (and me) because we weren't doing any Arcs.  Give me some time to come up with a fix.
> >  
> > Looks like it wasn't noise :}
> >  
> > Regards
> > TK
> >
> > From: himykabibble <jagboy@>
> > To: DynoMotion@yahoogroups.com
> > Sent: Thursday, February 23, 2012 12:05 PM
> > Subject: [DynoMotion] Re: v429z Problem
> >
> >
> >  
> > BTW - I can duplicate this with two different KFlops, and two different PCs....
> >
> > --- In DynoMotion@yahoogroups.com, "himykabibble" <jagboy@> wrote:
> > >
> > > OK, three spreadsheets uploaded to RayLs Files. The v428 trace looks the same, regardless of parameter setting. The v429 trace gets goofier the lower acceleration is set. Two captures, for accel = 30.0, and accel = 0.10 are given. All captures were recording Dest on every 50th pass of my Init program. First column is time, second thru fourth are X, Y, Z position.
> > >
> > > Regards,
> > > Ray L.
> > >
> > > --- In DynoMotion@yahoogroups.com, "himykabibble" <jagboy@> wrote:
> > > >
> > > > Tom,
> > > >
> > > > I'm posting some captures of ch[0-2].Dest for versions v428 and v429z. My machine was not even turned on for this. The problem is quite evident.
> > > >
> > > > Regards,
> > > > Ray L.
> > > >
> > > > --- In DynoMotion@yahoogroups.com, "himykabibble" <jagboy@> wrote:
> > > > >
> > > > > Tom,
> > > > >
> > > > > At this point, there is not a doubt in my mind that v429z is broken. I have run dozens of tests, and v428 always works, v429z always fails. With v429z, the lower I set acceleration, or feedrate, in KMotionCNC, the worse it performs. With accel set to 25 in/sec^2, I get small, but absolutely consistent, position errors, and NO servo faults. With accel set to 0.1 in/sec^2, I get nearly constant servo faults, with attendant massive position loss. The exact same g-code will work perfectly time after time with v428, with acceleration set as low or as high as I choose, and with feed rates cranked up to near rapid speeds. Even worse, it will sometimes move the Z axis, and there is not a single Z-axis command anywhere in the G-code! On the crooked, non-straight segments, I can put my fingers on the ballscrews, and I can feel clear, smooth motion on the axis that is supposed to be stationary. No chance this is noise - these are deliberate, constant speed
> > moves, always in exactly the same place, and they ARE reflected in the DROs.
> > > > >
> > > > > I've posted the G-code I'm using in RayLs Files.
> > > > >
> > > > > Regards,
> > > > > Ray L.
> > > > >
> > > >
> > >
> >
>
Group: DynoMotion Message: 4092 From: Tom Kerekes Date: 2/25/2012
Subject: Re: v429z Problem
Hi Ray
 
Here are the sources changed from V4.29z:
 
 
 
Regards
TK

Group: DynoMotion Message: 4102 From: himykabibble Date: 2/26/2012
Subject: Re: v429z Problem
Tom,

Today I ran a batch of the same parts that failed a few days ago, and that problem seems to be fixed. I was able to run one batch of parts that came out perfect.

However, while setting up, I did have a different problem that seems might be related - When I first started up my app, and started doing the setup, I found the interpreter was hanging if I did any MDI command that involved axis motion. I have on-screen indicators that show whether the interpreter is running or not, and it would stay on after any MDI that required axis motion, and the motion would not occur. Any non-move commands, like M3, M7, G54, etc. worked fine. I tried KMotionCNC and got the same behavior. I went back to my app, loaded my program, and started it, and it executed to the first move, then hung. I stopped it, did a rewind, and this time single-stepped through the first few moves, and all seemed fine. So I started it running from the start, and the problem did not re-occur. I'll see if it repeats tomorrow.

Regards,
Ray L.

--- In DynoMotion@yahoogroups.com, Tom Kerekes <tk@...> wrote:
>
> Hi Ray
>  
> Here are the sources changed from V4.29z:
>  
> http://dynomotion.com/Software/Patch/FixTP429Arcs/CoordMotion.cpp
>  
> http://dynomotion.com/Software/Patch/FixTP429Arcs/TrajectoryPlanner.cpp
>  
> Regards
> TK
>
> From: himykabibble <jagboy@...>
> To: DynoMotion@yahoogroups.com
> Sent: Saturday, February 25, 2012 2:35 PM
> Subject: [DynoMotion] Re: v429z Problem
>
>
>  
> Tom,
>
> I was only able to run a quick test, but it appears to be working correctly now. I'll try to run some real parts tomorrow, once I get my machine back together (finally installing limit switches on X, Y and A axes).
>
> Can I get the modified source, as I need to make my tool length comp mods before I can run any of my existing G-code.
>
> Regards,
> Ray L.
>
> --- In DynoMotion@yahoogroups.com, Tom Kerekes <tk@> wrote:
> >
> > Ray,
> >  
> > Please try this patch to fix GCode Arcs for V4.29z.  Copy to the C:\Kmotion429z\Kmotion\Release directory
> >  
> > http://dynomotion.com/Software/Patch/FixTP429Arcs/GCodeInterpreter.dll
> >  
> > Regards
> > TK
> >  
> >
> > From: himykabibble <jagboy@>
> > To: DynoMotion@yahoogroups.com
> > Sent: Thursday, February 23, 2012 2:25 PM
> > Subject: [DynoMotion] Re: v429z Problem
> >
> >
> >  
> > Tom,
> >
> > Well, how many people really NEED to do arcs anyway.... :-)
> >
> > Regards,
> > Ray L.
> >
> > --- In DynoMotion@yahoogroups.com, Tom Kerekes <tk@> wrote:
> > >
> > > Hi Ray,
> > >  
> > > I think I see what is going on.  I re-plotted your data and uploaded the Excel Sheet.   Looks like there is a problem with Arcs getting out of order with Line segments due to the Grouping into blocks of segments that we did in V4.29 to improve USB traffic.  Everything works fine for Albert (and me) because we weren't doing any Arcs.  Give me some time to come up with a fix.
> > >  
> > > Looks like it wasn't noise :}
> > >  
> > > Regards
> > > TK
> > >
> > > From: himykabibble <jagboy@>
> > > To: DynoMotion@yahoogroups.com
> > > Sent: Thursday, February 23, 2012 12:05 PM
> > > Subject: [DynoMotion] Re: v429z Problem
> > >
> > >
> > >  
> > > BTW - I can duplicate this with two different KFlops, and two different PCs....
> > >
> > > --- In DynoMotion@yahoogroups.com, "himykabibble" <jagboy@> wrote:
> > > >
> > > > OK, three spreadsheets uploaded to RayLs Files. The v428 trace looks the same, regardless of parameter setting. The v429 trace gets goofier the lower acceleration is set. Two captures, for accel = 30.0, and accel = 0.10 are given. All captures were recording Dest on every 50th pass of my Init program. First column is time, second thru fourth are X, Y, Z position.
> > > >
> > > > Regards,
> > > > Ray L.
> > > >
> > > > --- In DynoMotion@yahoogroups.com, "himykabibble" <jagboy@> wrote:
> > > > >
> > > > > Tom,
> > > > >
> > > > > I'm posting some captures of ch[0-2].Dest for versions v428 and v429z. My machine was not even turned on for this. The problem is quite evident.
> > > > >
> > > > > Regards,
> > > > > Ray L.
> > > > >
> > > > > --- In DynoMotion@yahoogroups.com, "himykabibble" <jagboy@> wrote:
> > > > > >
> > > > > > Tom,
> > > > > >
> > > > > > At this point, there is not a doubt in my mind that v429z is broken. I have run dozens of tests, and v428 always works, v429z always fails. With v429z, the lower I set acceleration, or feedrate, in KMotionCNC, the worse it performs. With accel set to 25 in/sec^2, I get small, but absolutely consistent, position errors, and NO servo faults. With accel set to 0.1 in/sec^2, I get nearly constant servo faults, with attendant massive position loss. The exact same g-code will work perfectly time after time with v428, with acceleration set as low or as high as I choose, and with feed rates cranked up to near rapid speeds. Even worse, it will sometimes move the Z axis, and there is not a single Z-axis command anywhere in the G-code! On the crooked, non-straight segments, I can put my fingers on the ballscrews, and I can feel clear, smooth motion on the axis that is supposed to be stationary. No chance this is noise - these are deliberate, constant speed
> > > moves, always in exactly the same place, and they ARE reflected in the DROs.
> > > > > >
> > > > > > I've posted the G-code I'm using in RayLs Files.
> > > > > >
> > > > > > Regards,
> > > > > > Ray L.
> > > > > >
> > > > >
> > > >
> > >
> >
>
Group: DynoMotion Message: 4109 From: himykabibble Date: 2/27/2012
Subject: Re: v429z Problem
Tom,

Whatever that weirdness was yesterday, it did not repeat today, so I think we're OK.

Any word on when we'll have a dotNet release where Connected works properly?

Also, is MainStatus now working correctly in dotNet?

Regards,
Ray L.

--- In DynoMotion@yahoogroups.com, "himykabibble" <jagboy@...> wrote:
>
> Tom,
>
> Today I ran a batch of the same parts that failed a few days ago, and that problem seems to be fixed. I was able to run one batch of parts that came out perfect.
>
> However, while setting up, I did have a different problem that seems might be related - When I first started up my app, and started doing the setup, I found the interpreter was hanging if I did any MDI command that involved axis motion. I have on-screen indicators that show whether the interpreter is running or not, and it would stay on after any MDI that required axis motion, and the motion would not occur. Any non-move commands, like M3, M7, G54, etc. worked fine. I tried KMotionCNC and got the same behavior. I went back to my app, loaded my program, and started it, and it executed to the first move, then hung. I stopped it, did a rewind, and this time single-stepped through the first few moves, and all seemed fine. So I started it running from the start, and the problem did not re-occur. I'll see if it repeats tomorrow.
>
> Regards,
> Ray L.
>
> --- In DynoMotion@yahoogroups.com, Tom Kerekes <tk@> wrote:
> >
> > Hi Ray
> >  
> > Here are the sources changed from V4.29z:
> >  
> > http://dynomotion.com/Software/Patch/FixTP429Arcs/CoordMotion.cpp
> >  
> > http://dynomotion.com/Software/Patch/FixTP429Arcs/TrajectoryPlanner.cpp
> >  
> > Regards
> > TK
> >
> > From: himykabibble <jagboy@>
> > To: DynoMotion@yahoogroups.com
> > Sent: Saturday, February 25, 2012 2:35 PM
> > Subject: [DynoMotion] Re: v429z Problem
> >
> >
> >  
> > Tom,
> >
> > I was only able to run a quick test, but it appears to be working correctly now. I'll try to run some real parts tomorrow, once I get my machine back together (finally installing limit switches on X, Y and A axes).
> >
> > Can I get the modified source, as I need to make my tool length comp mods before I can run any of my existing G-code.
> >
> > Regards,
> > Ray L.
> >
> > --- In DynoMotion@yahoogroups.com, Tom Kerekes <tk@> wrote:
> > >
> > > Ray,
> > >  
> > > Please try this patch to fix GCode Arcs for V4.29z.  Copy to the C:\Kmotion429z\Kmotion\Release directory
> > >  
> > > http://dynomotion.com/Software/Patch/FixTP429Arcs/GCodeInterpreter.dll
> > >  
> > > Regards
> > > TK
> > >  
> > >
> > > From: himykabibble <jagboy@>
> > > To: DynoMotion@yahoogroups.com
> > > Sent: Thursday, February 23, 2012 2:25 PM
> > > Subject: [DynoMotion] Re: v429z Problem
> > >
> > >
> > >  
> > > Tom,
> > >
> > > Well, how many people really NEED to do arcs anyway.... :-)
> > >
> > > Regards,
> > > Ray L.
> > >
> > > --- In DynoMotion@yahoogroups.com, Tom Kerekes <tk@> wrote:
> > > >
> > > > Hi Ray,
> > > >  
> > > > I think I see what is going on.  I re-plotted your data and uploaded the Excel Sheet.   Looks like there is a problem with Arcs getting out of order with Line segments due to the Grouping into blocks of segments that we did in V4.29 to improve USB traffic.  Everything works fine for Albert (and me) because we weren't doing any Arcs.  Give me some time to come up with a fix.
> > > >  
> > > > Looks like it wasn't noise :}
> > > >  
> > > > Regards
> > > > TK
> > > >
> > > > From: himykabibble <jagboy@>
> > > > To: DynoMotion@yahoogroups.com
> > > > Sent: Thursday, February 23, 2012 12:05 PM
> > > > Subject: [DynoMotion] Re: v429z Problem
> > > >
> > > >
> > > >  
> > > > BTW - I can duplicate this with two different KFlops, and two different PCs....
> > > >
> > > > --- In DynoMotion@yahoogroups.com, "himykabibble" <jagboy@> wrote:
> > > > >
> > > > > OK, three spreadsheets uploaded to RayLs Files. The v428 trace looks the same, regardless of parameter setting. The v429 trace gets goofier the lower acceleration is set. Two captures, for accel = 30.0, and accel = 0.10 are given. All captures were recording Dest on every 50th pass of my Init program. First column is time, second thru fourth are X, Y, Z position.
> > > > >
> > > > > Regards,
> > > > > Ray L.
> > > > >
> > > > > --- In DynoMotion@yahoogroups.com, "himykabibble" <jagboy@> wrote:
> > > > > >
> > > > > > Tom,
> > > > > >
> > > > > > I'm posting some captures of ch[0-2].Dest for versions v428 and v429z. My machine was not even turned on for this. The problem is quite evident.
> > > > > >
> > > > > > Regards,
> > > > > > Ray L.
> > > > > >
> > > > > > --- In DynoMotion@yahoogroups.com, "himykabibble" <jagboy@> wrote:
> > > > > > >
> > > > > > > Tom,
> > > > > > >
> > > > > > > At this point, there is not a doubt in my mind that v429z is broken. I have run dozens of tests, and v428 always works, v429z always fails. With v429z, the lower I set acceleration, or feedrate, in KMotionCNC, the worse it performs. With accel set to 25 in/sec^2, I get small, but absolutely consistent, position errors, and NO servo faults. With accel set to 0.1 in/sec^2, I get nearly constant servo faults, with attendant massive position loss. The exact same g-code will work perfectly time after time with v428, with acceleration set as low or as high as I choose, and with feed rates cranked up to near rapid speeds. Even worse, it will sometimes move the Z axis, and there is not a single Z-axis command anywhere in the G-code! On the crooked, non-straight segments, I can put my fingers on the ballscrews, and I can feel clear, smooth motion on the axis that is supposed to be stationary. No chance this is noise - these are deliberate, constant speed
> > > > moves, always in exactly the same place, and they ARE reflected in the DROs.
> > > > > > >
> > > > > > > I've posted the G-code I'm using in RayLs Files.
> > > > > > >
> > > > > > > Regards,
> > > > > > > Ray L.
> > > > > > >
> > > > > >
> > > > >
> > > >
> > >
> >
>
Group: DynoMotion Message: 4112 From: Brad Murry Date: 2/27/2012
Subject: Re: v429z Problem

MainStatus works well.

 

I have not tried the Connected() method with the new patches.

 

-Brad

 

From: DynoMotion@yahoogroups.com [mailto:DynoMotion@yahoogroups.com] On Behalf Of himykabibble
Sent: Monday, February 27, 2012 11:59 AM
To: DynoMotion@yahoogroups.com
Subject: [DynoMotion] Re: v429z Problem

 

 

Tom,

Whatever that weirdness was yesterday, it did not repeat today, so I think we're OK.

Any word on when we'll have a dotNet release where Connected works properly?

Also, is MainStatus now working correctly in dotNet?

Regards,
Ray L.

--- In DynoMotion@yahoogroups.com, "himykabibble" <jagboy@...> wrote:
>
> Tom,
>
> Today I ran a batch of the same parts that failed a few days ago, and that problem seems to be fixed. I was able to run one batch of parts that came out perfect.
>
> However, while setting up, I did have a different problem that seems might be related - When I first started up my app, and started doing the setup, I found the interpreter was hanging if I did any MDI command that involved axis motion. I have on-screen indicators that show whether the interpreter is running or not, and it would stay on after any MDI that required axis motion, and the motion would not occur. Any non-move commands, like M3, M7, G54, etc. worked fine. I tried KMotionCNC and got the same behavior. I went back to my app, loaded my program, and started it, and it executed to the first move, then hung. I stopped it, did a rewind, and this time single-stepped through the first few moves, and all seemed fine. So I started it running from the start, and the problem did not re-occur. I'll see if it repeats tomorrow.
>
> Regards,
> Ray L.
>
> --- In DynoMotion@yahoogroups.com, Tom Kerekes <tk@> wrote:
> >
> > Hi Ray
> >  
> > Here are the sources changed from V4.29z:
> >  
> > http://dynomotion.com/Software/Patch/FixTP429Arcs/CoordMotion.cpp
> >  
> > http://dynomotion.com/Software/Patch/FixTP429Arcs/TrajectoryPlanner.cpp
> >  
> > Regards
> > TK
> >
> > From: himykabibble <jagboy@>
> > To: DynoMotion@yahoogroups.com
> > Sent: Saturday, February 25, 2012 2:35 PM
> > Subject: [DynoMotion] Re: v429z Problem
> >
> >
> >  
> > Tom,
> >
> > I was only able to run a quick test, but it appears to be working correctly now. I'll try to run some real parts tomorrow, once I get my machine back together (finally installing limit switches on X, Y and A axes).
> >
> > Can I get the modified source, as I need to make my tool length comp mods before I can run any of my existing G-code.
> >
> > Regards,
> > Ray L.
> >
> > --- In DynoMotion@yahoogroups.com, Tom Kerekes <tk@> wrote:
> > >
> > > Ray,
> > >  
> > > Please try this patch to fix GCode Arcs for V4.29z.  Copy to the C:\Kmotion429z\Kmotion\Release directory
> > >  
> > > http://dynomotion.com/Software/Patch/FixTP429Arcs/GCodeInterpreter.dll
> > >  
> > > Regards
> > > TK
> > >  
> > >
> > > From: himykabibble <jagboy@>
> > > To: DynoMotion@yahoogroups.com
> > > Sent: Thursday, February 23, 2012 2:25 PM
> > > Subject: [DynoMotion] Re: v429z Problem
> > >
> > >
> > >  
> > > Tom,
> > >
> > > Well, how many people really NEED to do arcs anyway.... :-)
> > >
> > > Regards,
> > > Ray L.
> > >
> > > --- In DynoMotion@yahoogroups.com, Tom Kerekes <tk@> wrote:
> > > >
> > > > Hi Ray,
> > > >  
> > > > I think I see what is going on.  I re-plotted your data and uploaded the Excel Sheet.   Looks like there is a problem with Arcs getting out of order with Line segments due to the Grouping into blocks of segments that we did in V4.29 to improve USB traffic.  Everything works fine for Albert (and me) because we weren't doing any Arcs.  Give me some time to come up with a fix.
> > > >  
> > > > Looks like it wasn't noise :}
> > > >  
> > > > Regards
> > > > TK
> > > >
> > > > From: himykabibble <jagboy@>
> > > > To: DynoMotion@yahoogroups.com
> > > > Sent: Thursday, February 23, 2012 12:05 PM
> > > > Subject: [DynoMotion] Re: v429z Problem
> > > >
> > > >
> > > >  
> > > > BTW - I can duplicate this with two different KFlops, and two different PCs....
> > > >
> > > > --- In DynoMotion@yahoogroups.com, "himykabibble" <jagboy@> wrote:
> > > > >
> > > > > OK, three spreadsheets uploaded to RayLs Files. The v428 trace looks the same, regardless of parameter setting. The v429 trace gets goofier the lower acceleration is set. Two captures, for accel = 30.0, and accel = 0.10 are given. All captures were recording Dest on every 50th pass of my Init program. First column is time, second thru fourth are X, Y, Z position.
> > > > >
> > > > > Regards,
> > > > > Ray L.
> > > > >
> > > > > --- In DynoMotion@yahoogroups.com, "himykabibble" <jagboy@> wrote:
> > > > > >
> > > > > > Tom,
> > > > > >
> > > > > > I'm posting some captures of ch[0-2].Dest for versions v428 and v429z. My machine was not even turned on for this. The problem is quite evident.
> > > > > >
> > > > > > Regards,
> > > > > > Ray L.
> > > > > >
> > > > > > --- In DynoMotion@yahoogroups.com, "himykabibble" <jagboy@> wrote:
> > > > > > >
> > > > > > > Tom,
> > > > > > >
> > > > > > > At this point, there is not a doubt in my mind that v429z is broken. I have run dozens of tests, and v428 always works, v429z always fails. With v429z, the lower I set acceleration, or feedrate, in KMotionCNC, the worse it performs. With accel set to 25 in/sec^2, I get small, but absolutely consistent, position errors, and NO servo faults. With accel set to 0.1 in/sec^2, I get nearly constant servo faults, with attendant massive position loss. The exact same g-code will work perfectly time after time with v428, with acceleration set as low or as high as I choose, and with feed rates cranked up to near rapid speeds. Even worse, it will sometimes move the Z axis, and there is not a single Z-axis command anywhere in the G-code! On the crooked, non-straight segments, I can put my fingers on the ballscrews, and I can feel clear, smooth motion on the axis that is supposed to be stationary. No chance this is noise - these are deliberate, constant speed
> > > > moves, always in exactly the same place, and they ARE reflected in the DROs.
> > > > > > >
> > > > > > > I've posted the G-code I'm using in RayLs Files.
> > > > > > >
> > > > > > > Regards,
> > > > > > > Ray L.
> > > > > > >
> > > > > >
> > > > >
> > > >
> > >
> >
>
Group: DynoMotion Message: 4113 From: himykabibble Date: 2/27/2012
Subject: Re: v429z Problem
Brad,

"new patches"?? When/Where? All I've seen is the 429z release about 10 days ago.

Regards,
Ray L.

--- In DynoMotion@yahoogroups.com, Brad Murry <bradodarb@...> wrote:
>
> MainStatus works well.
>
>
>
> I have not tried the Connected() method with the new patches.
>
>
>
> -Brad
>
>
>
> From: DynoMotion@yahoogroups.com [mailto:DynoMotion@yahoogroups.com] On Behalf Of himykabibble
> Sent: Monday, February 27, 2012 11:59 AM
> To: DynoMotion@yahoogroups.com
> Subject: [DynoMotion] Re: v429z Problem
>
>
>
>
>
> Tom,
>
> Whatever that weirdness was yesterday, it did not repeat today, so I think we're OK.
>
> Any word on when we'll have a dotNet release where Connected works properly?
>
> Also, is MainStatus now working correctly in dotNet?
>
> Regards,
> Ray L.
>
> --- In DynoMotion@yahoogroups.com <mailto:DynoMotion%40yahoogroups.com> , "himykabibble" <jagboy@> wrote:
> >
> > Tom,
> >
> > Today I ran a batch of the same parts that failed a few days ago, and that problem seems to be fixed. I was able to run one batch of parts that came out perfect.
> >
> > However, while setting up, I did have a different problem that seems might be related - When I first started up my app, and started doing the setup, I found the interpreter was hanging if I did any MDI command that involved axis motion. I have on-screen indicators that show whether the interpreter is running or not, and it would stay on after any MDI that required axis motion, and the motion would not occur. Any non-move commands, like M3, M7, G54, etc. worked fine. I tried KMotionCNC and got the same behavior. I went back to my app, loaded my program, and started it, and it executed to the first move, then hung. I stopped it, did a rewind, and this time single-stepped through the first few moves, and all seemed fine. So I started it running from the start, and the problem did not re-occur. I'll see if it repeats tomorrow.
> >
> > Regards,
> > Ray L.
> >
> > --- In DynoMotion@yahoogroups.com <mailto:DynoMotion%40yahoogroups.com> , Tom Kerekes <tk@> wrote:
> > >
> > > Hi Ray
> > > Â
> > > Here are the sources changed from V4.29z:
> > > Â
> > > http://dynomotion.com/Software/Patch/FixTP429Arcs/CoordMotion.cpp
> > > Â
> > > http://dynomotion.com/Software/Patch/FixTP429Arcs/TrajectoryPlanner.cpp
> > > Â
> > > Regards
> > > TK
> > >
> > > From: himykabibble <jagboy@>
> > > To: DynoMotion@yahoogroups.com <mailto:DynoMotion%40yahoogroups.com>
> > > Sent: Saturday, February 25, 2012 2:35 PM
> > > Subject: [DynoMotion] Re: v429z Problem
> > >
> > >
> > > Â
> > > Tom,
> > >
> > > I was only able to run a quick test, but it appears to be working correctly now. I'll try to run some real parts tomorrow, once I get my machine back together (finally installing limit switches on X, Y and A axes).
> > >
> > > Can I get the modified source, as I need to make my tool length comp mods before I can run any of my existing G-code.
> > >
> > > Regards,
> > > Ray L.
> > >
> > > --- In DynoMotion@yahoogroups.com <mailto:DynoMotion%40yahoogroups.com> , Tom Kerekes <tk@> wrote:
> > > >
> > > > Ray,
> > > > ÂÂ
> > > > Please try this patch to fix GCode Arcs for V4.29z. Copy to the C:\Kmotion429z\Kmotion\Release directory
> > > > ÂÂ
> > > > http://dynomotion.com/Software/Patch/FixTP429Arcs/GCodeInterpreter.dll
> > > > ÂÂ
> > > > Regards
> > > > TK
> > > > ÂÂ
> > > >
> > > > From: himykabibble <jagboy@>
> > > > To: DynoMotion@yahoogroups.com <mailto:DynoMotion%40yahoogroups.com>
> > > > Sent: Thursday, February 23, 2012 2:25 PM
> > > > Subject: [DynoMotion] Re: v429z Problem
> > > >
> > > >
> > > > ÂÂ
> > > > Tom,
> > > >
> > > > Well, how many people really NEED to do arcs anyway.... :-)
> > > >
> > > > Regards,
> > > > Ray L.
> > > >
> > > > --- In DynoMotion@yahoogroups.com <mailto:DynoMotion%40yahoogroups.com> , Tom Kerekes <tk@> wrote:
> > > > >
> > > > > Hi Ray,
> > > > > ÃÆ'‚ÂÂ
> > > > > I think I see what is going on.ÃÆ'‚ÂÂ I re-plotted your data and uploaded the Excel Sheet.ÃÆ'‚ÂÂ ÃÆ'‚ÂÂ Looks like there is a problem with Arcs getting out of order with Line segments due to the Grouping into blocks of segments that we did in V4.29 to improve USB traffic.ÃÆ'‚ÂÂ Everything works fine for Albert (and me) because we weren't doing any Arcs.ÃÆ'‚ÂÂ Give me some time to come up with a fix.
> > > > > ÃÆ'‚ÂÂ
> > > > > Looks like it wasn't noise :}
> > > > > ÃÆ'‚ÂÂ
> > > > > Regards
> > > > > TK
> > > > >
> > > > > From: himykabibble <jagboy@>
> > > > > To: DynoMotion@yahoogroups.com <mailto:DynoMotion%40yahoogroups.com>
> > > > > Sent: Thursday, February 23, 2012 12:05 PM
> > > > > Subject: [DynoMotion] Re: v429z Problem
> > > > >
> > > > >
> > > > > ÃÆ'‚ÂÂ
> > > > > BTW - I can duplicate this with two different KFlops, and two different PCs....
> > > > >
> > > > > --- In DynoMotion@yahoogroups.com <mailto:DynoMotion%40yahoogroups.com> , "himykabibble" <jagboy@> wrote:
> > > > > >
> > > > > > OK, three spreadsheets uploaded to RayLs Files. The v428 trace looks the same, regardless of parameter setting. The v429 trace gets goofier the lower acceleration is set. Two captures, for accel = 30.0, and accel = 0.10 are given. All captures were recording Dest on every 50th pass of my Init program. First column is time, second thru fourth are X, Y, Z position.
> > > > > >
> > > > > > Regards,
> > > > > > Ray L.
> > > > > >
> > > > > > --- In DynoMotion@yahoogroups.com <mailto:DynoMotion%40yahoogroups.com> , "himykabibble" <jagboy@> wrote:
> > > > > > >
> > > > > > > Tom,
> > > > > > >
> > > > > > > I'm posting some captures of ch[0-2].Dest for versions v428 and v429z. My machine was not even turned on for this. The problem is quite evident.
> > > > > > >
> > > > > > > Regards,
> > > > > > > Ray L.
> > > > > > >
> > > > > > > --- In DynoMotion@yahoogroups.com <mailto:DynoMotion%40yahoogroups.com> , "himykabibble" <jagboy@> wrote:
> > > > > > > >
> > > > > > > > Tom,
> > > > > > > >
> > > > > > > > At this point, there is not a doubt in my mind that v429z is broken. I have run dozens of tests, and v428 always works, v429z always fails. With v429z, the lower I set acceleration, or feedrate, in KMotionCNC, the worse it performs. With accel set to 25 in/sec^2, I get small, but absolutely consistent, position errors, and NO servo faults. With accel set to 0.1 in/sec^2, I get nearly constant servo faults, with attendant massive position loss. The exact same g-code will work perfectly time after time with v428, with acceleration set as low or as high as I choose, and with feed rates cranked up to near rapid speeds. Even worse, it will sometimes move the Z axis, and there is not a single Z-axis command anywhere in the G-code! On the crooked, non-straight segments, I can put my fingers on the ballscrews, and I can feel clear, smooth motion on the axis that is supposed to be stationary. No chance this is noise - these are deliberate, constant speed
> > > > > moves, always in exactly the same place, and they ARE reflected in the DROs.
> > > > > > > >
> > > > > > > > I've posted the G-code I'm using in RayLs Files.
> > > > > > > >
> > > > > > > > Regards,
> > > > > > > > Ray L.
> > > > > > > >
> > > > > > >
> > > > > >
> > > > >
> > > >
> > >
> >
>
Group: DynoMotion Message: 4115 From: fireup_kev Date: 2/27/2012
Subject: Re: v429z Problem
For me, MainStatus still as a bug in the upper bits.

GetKFlopBitDiection does not return the correct values for bit 32 to 64

Kevin

--- In DynoMotion@yahoogroups.com, Brad Murry <bradodarb@...> wrote:
>
> MainStatus works well.
>
>
>
> I have not tried the Connected() method with the new patches.
>
>
>
> -Brad
>
>
>
> From: DynoMotion@yahoogroups.com [mailto:DynoMotion@yahoogroups.com] On Behalf Of himykabibble
> Sent: Monday, February 27, 2012 11:59 AM
> To: DynoMotion@yahoogroups.com
> Subject: [DynoMotion] Re: v429z Problem
>
>
>
>
>
> Tom,
>
> Whatever that weirdness was yesterday, it did not repeat today, so I think we're OK.
>
> Any word on when we'll have a dotNet release where Connected works properly?
>
> Also, is MainStatus now working correctly in dotNet?
>
> Regards,
> Ray L.
>
> --- In DynoMotion@yahoogroups.com <mailto:DynoMotion%40yahoogroups.com> , "himykabibble" <jagboy@> wrote:
> >
> > Tom,
> >
> > Today I ran a batch of the same parts that failed a few days ago, and that problem seems to be fixed. I was able to run one batch of parts that came out perfect.
> >
> > However, while setting up, I did have a different problem that seems might be related - When I first started up my app, and started doing the setup, I found the interpreter was hanging if I did any MDI command that involved axis motion. I have on-screen indicators that show whether the interpreter is running or not, and it would stay on after any MDI that required axis motion, and the motion would not occur. Any non-move commands, like M3, M7, G54, etc. worked fine. I tried KMotionCNC and got the same behavior. I went back to my app, loaded my program, and started it, and it executed to the first move, then hung. I stopped it, did a rewind, and this time single-stepped through the first few moves, and all seemed fine. So I started it running from the start, and the problem did not re-occur. I'll see if it repeats tomorrow.
> >
> > Regards,
> > Ray L.
> >
> > --- In DynoMotion@yahoogroups.com <mailto:DynoMotion%40yahoogroups.com> , Tom Kerekes <tk@> wrote:
> > >
> > > Hi Ray
> > > Â
> > > Here are the sources changed from V4.29z:
> > > Â
> > > http://dynomotion.com/Software/Patch/FixTP429Arcs/CoordMotion.cpp
> > > Â
> > > http://dynomotion.com/Software/Patch/FixTP429Arcs/TrajectoryPlanner.cpp
> > > Â
> > > Regards
> > > TK
> > >
> > > From: himykabibble <jagboy@>
> > > To: DynoMotion@yahoogroups.com <mailto:DynoMotion%40yahoogroups.com>
> > > Sent: Saturday, February 25, 2012 2:35 PM
> > > Subject: [DynoMotion] Re: v429z Problem
> > >
> > >
> > > Â
> > > Tom,
> > >
> > > I was only able to run a quick test, but it appears to be working correctly now. I'll try to run some real parts tomorrow, once I get my machine back together (finally installing limit switches on X, Y and A axes).
> > >
> > > Can I get the modified source, as I need to make my tool length comp mods before I can run any of my existing G-code.
> > >
> > > Regards,
> > > Ray L.
> > >
> > > --- In DynoMotion@yahoogroups.com <mailto:DynoMotion%40yahoogroups.com> , Tom Kerekes <tk@> wrote:
> > > >
> > > > Ray,
> > > > ÂÂ
> > > > Please try this patch to fix GCode Arcs for V4.29z. Copy to the C:\Kmotion429z\Kmotion\Release directory
> > > > ÂÂ
> > > > http://dynomotion.com/Software/Patch/FixTP429Arcs/GCodeInterpreter.dll
> > > > ÂÂ
> > > > Regards
> > > > TK
> > > > ÂÂ
> > > >
> > > > From: himykabibble <jagboy@>
> > > > To: DynoMotion@yahoogroups.com <mailto:DynoMotion%40yahoogroups.com>
> > > > Sent: Thursday, February 23, 2012 2:25 PM
> > > > Subject: [DynoMotion] Re: v429z Problem
> > > >
> > > >
> > > > ÂÂ
> > > > Tom,
> > > >
> > > > Well, how many people really NEED to do arcs anyway.... :-)
> > > >
> > > > Regards,
> > > > Ray L.
> > > >
> > > > --- In DynoMotion@yahoogroups.com <mailto:DynoMotion%40yahoogroups.com> , Tom Kerekes <tk@> wrote:
> > > > >
> > > > > Hi Ray,
> > > > > ÃÆ'‚ÂÂ
> > > > > I think I see what is going on.ÃÆ'‚ÂÂ I re-plotted your data and uploaded the Excel Sheet.ÃÆ'‚ÂÂ ÃÆ'‚ÂÂ Looks like there is a problem with Arcs getting out of order with Line segments due to the Grouping into blocks of segments that we did in V4.29 to improve USB traffic.ÃÆ'‚ÂÂ Everything works fine for Albert (and me) because we weren't doing any Arcs.ÃÆ'‚ÂÂ Give me some time to come up with a fix.
> > > > > ÃÆ'‚ÂÂ
> > > > > Looks like it wasn't noise :}
> > > > > ÃÆ'‚ÂÂ
> > > > > Regards
> > > > > TK
> > > > >
> > > > > From: himykabibble <jagboy@>
> > > > > To: DynoMotion@yahoogroups.com <mailto:DynoMotion%40yahoogroups.com>
> > > > > Sent: Thursday, February 23, 2012 12:05 PM
> > > > > Subject: [DynoMotion] Re: v429z Problem
> > > > >
> > > > >
> > > > > ÃÆ'‚ÂÂ
> > > > > BTW - I can duplicate this with two different KFlops, and two different PCs....
> > > > >
> > > > > --- In DynoMotion@yahoogroups.com <mailto:DynoMotion%40yahoogroups.com> , "himykabibble" <jagboy@> wrote:
> > > > > >
> > > > > > OK, three spreadsheets uploaded to RayLs Files. The v428 trace looks the same, regardless of parameter setting. The v429 trace gets goofier the lower acceleration is set. Two captures, for accel = 30.0, and accel = 0.10 are given. All captures were recording Dest on every 50th pass of my Init program. First column is time, second thru fourth are X, Y, Z position.
> > > > > >
> > > > > > Regards,
> > > > > > Ray L.
> > > > > >
> > > > > > --- In DynoMotion@yahoogroups.com <mailto:DynoMotion%40yahoogroups.com> , "himykabibble" <jagboy@> wrote:
> > > > > > >
> > > > > > > Tom,
> > > > > > >
> > > > > > > I'm posting some captures of ch[0-2].Dest for versions v428 and v429z. My machine was not even turned on for this. The problem is quite evident.
> > > > > > >
> > > > > > > Regards,
> > > > > > > Ray L.
> > > > > > >
> > > > > > > --- In DynoMotion@yahoogroups.com <mailto:DynoMotion%40yahoogroups.com> , "himykabibble" <jagboy@> wrote:
> > > > > > > >
> > > > > > > > Tom,
> > > > > > > >
> > > > > > > > At this point, there is not a doubt in my mind that v429z is broken. I have run dozens of tests, and v428 always works, v429z always fails. With v429z, the lower I set acceleration, or feedrate, in KMotionCNC, the worse it performs. With accel set to 25 in/sec^2, I get small, but absolutely consistent, position errors, and NO servo faults. With accel set to 0.1 in/sec^2, I get nearly constant servo faults, with attendant massive position loss. The exact same g-code will work perfectly time after time with v428, with acceleration set as low or as high as I choose, and with feed rates cranked up to near rapid speeds. Even worse, it will sometimes move the Z axis, and there is not a single Z-axis command anywhere in the G-code! On the crooked, non-straight segments, I can put my fingers on the ballscrews, and I can feel clear, smooth motion on the axis that is supposed to be stationary. No chance this is noise - these are deliberate, constant speed
> > > > > moves, always in exactly the same place, and they ARE reflected in the DROs.
> > > > > > > >
> > > > > > > > I've posted the G-code I'm using in RayLs Files.
> > > > > > > >
> > > > > > > > Regards,
> > > > > > > > Ray L.
> > > > > > > >
> > > > > > >
> > > > > >
> > > > >
> > > >
> > >
> >
>
Group: DynoMotion Message: 4127 From: Tom Kerekes Date: 2/27/2012
Subject: Re: v429z Problem
Hi Kevin,
 
Yes it looks like (KMotion_dotNet\KM_MainStatus.cs)
 
public int GetKFlopBitDirection(int index)
{
    if (index > 31)
    {
        return GetValue(_BitsDirection[1], index - 31, 1);
    }
    else
    {
        return GetValue(_BitsDirection[0], index, 1);
    }
}
 
should be:
 
public int GetKFlopBitDirection(int index)
{
if (index > 31)
{
return GetValue(_BitsDirection[1], index - 32, 1);
}
else
{
return GetValue(_BitsDirection[0], index, 1);
}
}
 
Thanks
TK

Group: DynoMotion Message: 4132 From: fireup_kev Date: 2/28/2012
Subject: Re: v429z Problem
Thanks Tom

So this, and

http://dynomotion.com/Software/Patch/FixTP429Arcs/CoordMotion.cpp

http://dynomotion.com/Software/Patch/FixTP429Arcs/TrajectoryPlanner.cpp

are the changes I need to make to the source to take care of the current bugs reported so far?

Kevin

--- In DynoMotion@yahoogroups.com, Tom Kerekes <tk@...> wrote:
>
> Hi Kevin,
>  
> Yes it looks like (KMotion_dotNet\KM_MainStatus.cs)
>  {
>     {
>     }publicintGetKFlopBitDirection(intindex)    if(index > 31)        returnGetValue(_BitsDirection[1], index - 31, 1);    else    {
>     }
> }        returnGetValue(_BitsDirection[0], index, 1);
>  
> should be:
>  publicintGetKFlopBitDirection(intindex)if(index > 31)returnGetValue(_BitsDirection[1], index - 32, 1);else{
> }
> }returnGetValue(_BitsDirection[0], index, 1);
>  
> Thanks
> TK
>
> From: fireup_kev <kliboon@...>
> To: DynoMotion@yahoogroups.com
> Sent: Monday, February 27, 2012 2:18 PM
> Subject: [DynoMotion] Re: v429z Problem
>
>
>  
> For me, MainStatus still as a bug in the upper bits.
>
> GetKFlopBitDiection does not return the correct values for bit 32 to 64
>
> Kevin
>
> --- In DynoMotion@yahoogroups.com, Brad Murry <bradodarb@> wrote:
> >
> > MainStatus works well.
> >
> >
> >
> > I have not tried the Connected() method with the new patches.
> >
> >
> >
> > -Brad
> >
> >
> >
> > From: DynoMotion@yahoogroups.com [mailto:DynoMotion@yahoogroups.com] On Behalf Of himykabibble
> > Sent: Monday, February 27, 2012 11:59 AM
> > To: DynoMotion@yahoogroups.com
> > Subject: [DynoMotion] Re: v429z Problem
> >
> >
> >
> >
> >
> > Tom,
> >
> > Whatever that weirdness was yesterday, it did not repeat today, so I think we're OK.
> >
> > Any word on when we'll have a dotNet release where Connected works properly?
> >
> > Also, is MainStatus now working correctly in dotNet?
> >
> > Regards,
> > Ray L.
> >
> > --- In DynoMotion@yahoogroups.com <mailto:DynoMotion%40yahoogroups.com> , "himykabibble" <jagboy@> wrote:
> > >
> > > Tom,
> > >
> > > Today I ran a batch of the same parts that failed a few days ago, and that problem seems to be fixed. I was able to run one batch of parts that came out perfect.
> > >
> > > However, while setting up, I did have a different problem that seems might be related - When I first started up my app, and started doing the setup, I found the interpreter was hanging if I did any MDI command that involved axis motion. I have on-screen indicators that show whether the interpreter is running or not, and it would stay on after any MDI that required axis motion, and the motion would not occur. Any non-move commands, like M3, M7, G54, etc. worked fine. I tried KMotionCNC and got the same behavior. I went back to my app, loaded my program, and started it, and it executed to the first move, then hung. I stopped it, did a rewind, and this time single-stepped through the first few moves, and all seemed fine. So I started it running from the start, and the problem did not re-occur. I'll see if it repeats tomorrow.
> > >
> > > Regards,
> > > Ray L.
> > >
> > > --- In DynoMotion@yahoogroups.com <mailto:DynoMotion%40yahoogroups.com> , Tom Kerekes <tk@> wrote:
> > > >
> > > > Hi Ray
> > > > Â
> > > > Here are the sources changed from V4.29z:
> > > > Â
> > > > http://dynomotion.com/Software/Patch/FixTP429Arcs/CoordMotion.cpp
> > > > Â
> > > > http://dynomotion.com/Software/Patch/FixTP429Arcs/TrajectoryPlanner.cpp
> > > > Â
> > > > Regards
> > > > TK
> > > >
> > > > From: himykabibble <jagboy@>
> > > > To: DynoMotion@yahoogroups.com <mailto:DynoMotion%40yahoogroups.com>
> > > > Sent: Saturday, February 25, 2012 2:35 PM
> > > > Subject: [DynoMotion] Re: v429z Problem
> > > >
> > > >
> > > > Â
> > > > Tom,
> > > >
> > > > I was only able to run a quick test, but it appears to be working correctly now. I'll try to run some real parts tomorrow, once I get my machine back together (finally installing limit switches on X, Y and A axes).
> > > >
> > > > Can I get the modified source, as I need to make my tool length comp mods before I can run any of my existing G-code.
> > > >
> > > > Regards,
> > > > Ray L.
> > > >
> > > > --- In DynoMotion@yahoogroups.com <mailto:DynoMotion%40yahoogroups.com> , Tom Kerekes <tk@> wrote:
> > > > >
> > > > > Ray,
> > > > > ÃÆ'‚Â
> > > > > Please try this patch to fix GCode Arcs for V4.29z.ÃÆ'‚Â Copy to the C:\Kmotion429z\Kmotion\Release directory
> > > > > ÃÆ'‚Â
> > > > > http://dynomotion.com/Software/Patch/FixTP429Arcs/GCodeInterpreter.dll
> > > > > ÃÆ'‚Â
> > > > > Regards
> > > > > TK
> > > > > ÃÆ'‚Â
> > > > >
> > > > > From: himykabibble <jagboy@>
> > > > > To: DynoMotion@yahoogroups.com <mailto:DynoMotion%40yahoogroups.com>
> > > > > Sent: Thursday, February 23, 2012 2:25 PM
> > > > > Subject: [DynoMotion] Re: v429z Problem
> > > > >
> > > > >
> > > > > ÃÆ'‚Â
> > > > > Tom,
> > > > >
> > > > > Well, how many people really NEED to do arcs anyway.... :-)
> > > > >
> > > > > Regards,
> > > > > Ray L.
> > > > >
> > > > > --- In DynoMotion@yahoogroups.com <mailto:DynoMotion%40yahoogroups.com> , Tom Kerekes <tk@> wrote:
> > > > > >
> > > > > > Hi Ray,
> > > > > > ÃÆ'Æ'‚ÃÆ'‚Â
> > > > > > I think I see what is going on.ÃÆ'Æ'‚ÃÆ'‚Â I re-plotted your data and uploaded the Excel Sheet.ÃÆ'Æ'‚ÃÆ'‚Â ÃÆ'Æ'‚ÃÆ'‚Â Looks like there is a problem with Arcs getting out of order with Line segments due to the Grouping into blocks of segments that we did in V4.29 to improve USB traffic.ÃÆ'Æ'‚ÃÆ'‚Â Everything works fine for Albert (and me) because we weren't doing any Arcs.ÃÆ'Æ'‚ÃÆ'‚Â Give me some time to come up with a fix.
> > > > > > ÃÆ'Æ'‚ÃÆ'‚Â
> > > > > > Looks like it wasn't noise :}
> > > > > > ÃÆ'Æ'‚ÃÆ'‚Â
> > > > > > Regards
> > > > > > TK
> > > > > >
> > > > > > From: himykabibble <jagboy@>
> > > > > > To: DynoMotion@yahoogroups.com <mailto:DynoMotion%40yahoogroups.com>
> > > > > > Sent: Thursday, February 23, 2012 12:05 PM
> > > > > > Subject: [DynoMotion] Re: v429z Problem
> > > > > >
> > > > > >
> > > > > > ÃÆ'Æ'‚ÃÆ'‚Â
> > > > > > BTW - I can duplicate this with two different KFlops, and two different PCs....
> > > > > >
> > > > > > --- In DynoMotion@yahoogroups.com <mailto:DynoMotion%40yahoogroups.com> , "himykabibble" <jagboy@> wrote:
> > > > > > >
> > > > > > > OK, three spreadsheets uploaded to RayLs Files. The v428 trace looks the same, regardless of parameter setting. The v429 trace gets goofier the lower acceleration is set. Two captures, for accel = 30.0, and accel = 0.10 are given. All captures were recording Dest on every 50th pass of my Init program. First column is time, second thru fourth are X, Y, Z position.
> > > > > > >
> > > > > > > Regards,
> > > > > > > Ray L.
> > > > > > >
> > > > > > > --- In DynoMotion@yahoogroups.com <mailto:DynoMotion%40yahoogroups.com> , "himykabibble" <jagboy@> wrote:
> > > > > > > >
> > > > > > > > Tom,
> > > > > > > >
> > > > > > > > I'm posting some captures of ch[0-2].Dest for versions v428 and v429z. My machine was not even turned on for this. The problem is quite evident.
> > > > > > > >
> > > > > > > > Regards,
> > > > > > > > Ray L.
> > > > > > > >
> > > > > > > > --- In DynoMotion@yahoogroups.com <mailto:DynoMotion%40yahoogroups.com> , "himykabibble" <jagboy@> wrote:
> > > > > > > > >
> > > > > > > > > Tom,
> > > > > > > > >
> > > > > > > > > At this point, there is not a doubt in my mind that v429z is broken. I have run dozens of tests, and v428 always works, v429z always fails. With v429z, the lower I set acceleration, or feedrate, in KMotionCNC, the worse it performs. With accel set to 25 in/sec^2, I get small, but absolutely consistent, position errors, and NO servo faults. With accel set to 0.1 in/sec^2, I get nearly constant servo faults, with attendant massive position loss. The exact same g-code will work perfectly time after time with v428, with acceleration set as low or as high as I choose, and with feed rates cranked up to near rapid speeds. Even worse, it will sometimes move the Z axis, and there is not a single Z-axis command anywhere in the G-code! On the crooked, non-straight segments, I can put my fingers on the ballscrews, and I can feel clear, smooth motion on the axis that is supposed to be stationary. No chance this is noise - these are deliberate,
> constant speed
> > > > > > moves, always in exactly the same place, and they ARE reflected in the DROs.
> > > > > > > > >
> > > > > > > > > I've posted the G-code I'm using in RayLs Files.
> > > > > > > > >
> > > > > > > > > Regards,
> > > > > > > > > Ray L.
> > > > > > > > >
> > > > > > > >
> > > > > > >
> > > > > >
> > > > >
> > > >
> > >
> >
>
>
>
>
>
> {
> {
> }
>
Group: DynoMotion Message: 4133 From: Tom Kerekes Date: 2/28/2012
Subject: Re: v429z Problem
Hi Kevin,
 
Yes those should be the only changes.
 
TK

Group: DynoMotion Message: 4144 From: himykabibble Date: 2/29/2012
Subject: Re: v429z Problem
Tom,

Thought you'd like to know - I've been using my machine quite a bit the last several days with 429z, and not a single problem since putting in the planner patches. Everything has worked absolutely perfectly.

Kinda funny - I've gotten so used to Mach3 doing unpredictable things over the last few years, today I thought something weird was happening, and hit the stop button, only to find it was doing the right thing, and I was confused about where it was in the program. I actually ended up trusting it enough to leave the machine running while I went into the house for lunch. I could NEVER do that with Mach3.

Regards,
Ray L.

--- In DynoMotion@yahoogroups.com, Tom Kerekes <tk@...> wrote:
>
> Hi Kevin,
>  
> Yes those should be the only changes.
>  
> TK
>
>
> ________________________________
> From: fireup_kev <kliboon@...>
> To: DynoMotion@yahoogroups.com
> Sent: Tuesday, February 28, 2012 7:33 AM
> Subject: [DynoMotion] Re: v429z Problem
>
>
>  
> Thanks Tom
>
> So this, and
>
> http://dynomotion.com/Software/Patch/FixTP429Arcs/CoordMotion.cpp
>
> http://dynomotion.com/Software/Patch/FixTP429Arcs/TrajectoryPlanner.cpp
>
> are the changes I need to make to the source to take care of the current bugs reported so far?
>
> Kevin
>
> --- In DynoMotion@yahoogroups.com, Tom Kerekes <tk@> wrote:
> >
> > Hi Kevin,
> >  
> > Yes it looks like (KMotion_dotNet\KM_MainStatus.cs)
> >  {
> >     {
> >     }publicintGetKFlopBitDirection(intindex)    if(index > 31)        returnGetValue(_BitsDirection[1], index - 31, 1);    else    {
> >     }
> > }        returnGetValue(_BitsDirection[0], index, 1);
> >  
> > should be:
> >  publicintGetKFlopBitDirection(intindex)if(index > 31)returnGetValue(_BitsDirection[1], index - 32, 1);else{
> > }
> > }returnGetValue(_BitsDirection[0], index, 1);
> >  
> > Thanks
> > TK
> >
> > From: fireup_kev <kliboon@>
> > To: DynoMotion@yahoogroups.com
> > Sent: Monday, February 27, 2012 2:18 PM
> > Subject: [DynoMotion] Re: v429z Problem
> >
> >
> >  
> > For me, MainStatus still as a bug in the upper bits.
> >
> > GetKFlopBitDiection does not return the correct values for bit 32 to 64
> >
> > Kevin
> >
> > --- In DynoMotion@yahoogroups.com, Brad Murry <bradodarb@> wrote:
> > >
> > > MainStatus works well.
> > >
> > >
> > >
> > > I have not tried the Connected() method with the new patches.
> > >
> > >
> > >
> > > -Brad
> > >
> > >
> > >
> > > From: DynoMotion@yahoogroups.com [mailto:DynoMotion@yahoogroups.com] On Behalf Of himykabibble
> > > Sent: Monday, February 27, 2012 11:59 AM
> > > To: DynoMotion@yahoogroups.com
> > > Subject: [DynoMotion] Re: v429z Problem
> > >
> > >
> > >
> > >
> > >
> > > Tom,
> > >
> > > Whatever that weirdness was yesterday, it did not repeat today, so I think we're OK.
> > >
> > > Any word on when we'll have a dotNet release where Connected works properly?
> > >
> > > Also, is MainStatus now working correctly in dotNet?
> > >
> > > Regards,
> > > Ray L.
> > >
> > > --- In DynoMotion@yahoogroups.com <mailto:DynoMotion%40yahoogroups.com> , "himykabibble" <jagboy@> wrote:
> > > >
> > > > Tom,
> > > >
> > > > Today I ran a batch of the same parts that failed a few days ago, and that problem seems to be fixed. I was able to run one batch of parts that came out perfect.
> > > >
> > > > However, while setting up, I did have a different problem that seems might be related - When I first started up my app, and started doing the setup, I found the interpreter was hanging if I did any MDI command that involved axis motion. I have on-screen indicators that show whether the interpreter is running or not, and it would stay on after any MDI that required axis motion, and the motion would not occur. Any non-move commands, like M3, M7, G54, etc. worked fine. I tried KMotionCNC and got the same behavior. I went back to my app, loaded my program, and started it, and it executed to the first move, then hung. I stopped it, did a rewind, and this time single-stepped through the first few moves, and all seemed fine. So I started it running from the start, and the problem did not re-occur. I'll see if it repeats tomorrow.
> > > >
> > > > Regards,
> > > > Ray L.
> > > >
> > > > --- In DynoMotion@yahoogroups.com <mailto:DynoMotion%40yahoogroups.com> , Tom Kerekes <tk@> wrote:
> > > > >
> > > > > Hi Ray
> > > > > ÃÆ'‚
> > > > > Here areÃÆ'‚ the sources changed from V4.29z:
> > > > > ÃÆ'‚
> > > > > http://dynomotion.com/Software/Patch/FixTP429Arcs/CoordMotion.cpp
> > > > > ÃÆ'‚
> > > > > http://dynomotion.com/Software/Patch/FixTP429Arcs/TrajectoryPlanner.cpp
> > > > > ÃÆ'‚
> > > > > Regards
> > > > > TK
> > > > >
> > > > > From: himykabibble <jagboy@>
> > > > > To: DynoMotion@yahoogroups.com <mailto:DynoMotion%40yahoogroups.com>
> > > > > Sent: Saturday, February 25, 2012 2:35 PM
> > > > > Subject: [DynoMotion] Re: v429z Problem
> > > > >
> > > > >
> > > > > ÃÆ'‚
> > > > > Tom,
> > > > >
> > > > > I was only able to run a quick test, but it appears to be working correctly now. I'll try to run some real parts tomorrow, once I get my machine back together (finally installing limit switches on X, Y and A axes).
> > > > >
> > > > > Can I get the modified source, as I need to make my tool length comp mods before I can run any of my existing G-code.
> > > > >
> > > > > Regards,
> > > > > Ray L.
> > > > >
> > > > > --- In DynoMotion@yahoogroups.com <mailto:DynoMotion%40yahoogroups.com> , Tom Kerekes <tk@> wrote:
> > > > > >
> > > > > > Ray,
> > > > > > ÃÆ'Æ'‚ÃÆ'‚
> > > > > > Please try this patch to fix GCode Arcs for V4.29z.ÃÆ'Æ'‚ÃÆ'‚ Copy to the C:\Kmotion429z\Kmotion\Release directory
> > > > > > ÃÆ'Æ'‚ÃÆ'‚
> > > > > > http://dynomotion.com/Software/Patch/FixTP429Arcs/GCodeInterpreter.dll
> > > > > > ÃÆ'Æ'‚ÃÆ'‚
> > > > > > Regards
> > > > > > TK
> > > > > > ÃÆ'Æ'‚ÃÆ'‚
> > > > > >
> > > > > > From: himykabibble <jagboy@>
> > > > > > To: DynoMotion@yahoogroups.com <mailto:DynoMotion%40yahoogroups.com>
> > > > > > Sent: Thursday, February 23, 2012 2:25 PM
> > > > > > Subject: [DynoMotion] Re: v429z Problem
> > > > > >
> > > > > >
> > > > > > ÃÆ'Æ'‚ÃÆ'‚
> > > > > > Tom,
> > > > > >
> > > > > > Well, how many people really NEED to do arcs anyway.... :-)
> > > > > >
> > > > > > Regards,
> > > > > > Ray L.
> > > > > >
> > > > > > --- In DynoMotion@yahoogroups.com <mailto:DynoMotion%40yahoogroups.com> , Tom Kerekes <tk@> wrote:
> > > > > > >
> > > > > > > Hi Ray,
> > > > > > > ÃÆ'Æ'Æ'ÃÆ'¢â‚¬Å¡ÃÆ'Æ'‚ÃÆ'‚
> > > > > > > I think I see what is going on.ÃÆ'Æ'Æ'ÃÆ'¢â‚¬Å¡ÃÆ'Æ'‚ÃÆ'‚ I re-plotted your data and uploaded the Excel Sheet.ÃÆ'Æ'Æ'ÃÆ'¢â‚¬Å¡ÃÆ'Æ'‚ÃÆ'‚ ÃÆ'Æ'Æ'ÃÆ'¢â‚¬Å¡ÃÆ'Æ'‚ÃÆ'‚ Looks like there is a problem with Arcs getting out of order with Line segments due to the Grouping into blocks of segments that we did in V4.29 to improve USB traffic.ÃÆ'Æ'Æ'ÃÆ'¢â‚¬Å¡ÃÆ'Æ'‚ÃÆ'‚ Everything works fine for Albert (and me) because we weren't doing any Arcs.ÃÆ'Æ'Æ'ÃÆ'¢â‚¬Å¡ÃÆ'Æ'‚ÃÆ'‚ Give me some time to come up with a fix.
> > > > > > > ÃÆ'Æ'Æ'ÃÆ'¢â‚¬Å¡ÃÆ'Æ'‚ÃÆ'‚
> > > > > > > Looks like it wasn't noise :}
> > > > > > > ÃÆ'Æ'Æ'ÃÆ'¢â‚¬Å¡ÃÆ'Æ'‚ÃÆ'‚
> > > > > > > Regards
> > > > > > > TK
> > > > > > >
> > > > > > > From: himykabibble <jagboy@>
> > > > > > > To: DynoMotion@yahoogroups.com <mailto:DynoMotion%40yahoogroups.com>
> > > > > > > Sent: Thursday, February 23, 2012 12:05 PM
> > > > > > > Subject: [DynoMotion] Re: v429z Problem
> > > > > > >
> > > > > > >
> > > > > > > ÃÆ'Æ'Æ'ÃÆ'¢â‚¬Å¡ÃÆ'Æ'‚ÃÆ'‚
> > > > > > > BTW - I can duplicate this with two different KFlops, and two different PCs....
> > > > > > >
> > > > > > > --- In DynoMotion@yahoogroups.com <mailto:DynoMotion%40yahoogroups.com> , "himykabibble" <jagboy@> wrote:
> > > > > > > >
> > > > > > > > OK, three spreadsheets uploaded to RayLs Files. The v428 trace looks the same, regardless of parameter setting. The v429 trace gets goofier the lower acceleration is set. Two captures, for accel = 30.0, and accel = 0.10 are given. All captures were recording Dest on every 50th pass of my Init program. First column is time, second thru fourth are X, Y, Z position.
> > > > > > > >
> > > > > > > > Regards,
> > > > > > > > Ray L.
> > > > > > > >
> > > > > > > > --- In DynoMotion@yahoogroups.com <mailto:DynoMotion%40yahoogroups.com> , "himykabibble" <jagboy@> wrote:
> > > > > > > > >
> > > > > > > > > Tom,
> > > > > > > > >
> > > > > > > > > I'm posting some captures of ch[0-2].Dest for versions v428 and v429z. My machine was not even turned on for this. The problem is quite evident.
> > > > > > > > >
> > > > > > > > > Regards,
> > > > > > > > > Ray L.
> > > > > > > > >
> > > > > > > > > --- In DynoMotion@yahoogroups.com <mailto:DynoMotion%40yahoogroups.com> , "himykabibble" <jagboy@> wrote:
> > > > > > > > > >
> > > > > > > > > > Tom,
> > > > > > > > > >
> > > > > > > > > > At this point, there is not a doubt in my mind that v429z is broken. I have run dozens of tests, and v428 always works, v429z always fails. With v429z, the lower I set acceleration, or feedrate, in KMotionCNC, the worse it performs. With accel set to 25 in/sec^2, I get small, but absolutely consistent, position errors, and NO servo faults. With accel set to 0.1 in/sec^2, I get nearly constant servo faults, with attendant massive position loss. The exact same g-code will work perfectly time after time with v428, with acceleration set as low or as high as I choose, and with feed rates cranked up to near rapid speeds. Even worse, it will sometimes move the Z axis, and there is not a single Z-axis command anywhere in the G-code! On the crooked, non-straight segments, I can put my fingers on the ballscrews, and I can feel clear, smooth motion on the axis that is supposed to be stationary. No chance this is noise - these are deliberate,
> > constant speed
> > > > > > > moves, always in exactly the same place, and they ARE reflected in the DROs.
> > > > > > > > > >
> > > > > > > > > > I've posted the G-code I'm using in RayLs Files.
> > > > > > > > > >
> > > > > > > > > > Regards,
> > > > > > > > > > Ray L.
> > > > > > > > > >
> > > > > > > > >
> > > > > > > >
> > > > > > >
> > > > > >
> > > > >
> > > >
> > >
> >
> >
> >
> >
> >
> > {
> > {
> > }
> >
>
Group: DynoMotion Message: 4145 From: fireup_kev Date: 2/29/2012
Subject: Re: v429z Problem
I found another problem....

When I move my application out to it's own directory, it gives me an unable to open file "\GCode Programs\emc.var" error. Last version I specified CoordMotion.Interpreter.MainPathRoot but 4.29z no longer has this value and CoordMotion.Interpreter.SetupFile does not seem to be the same thing.

Thanks,
Kevin

--- In DynoMotion@yahoogroups.com, Tom Kerekes <tk@...> wrote:
>
> Hi Kevin,
>  
> Yes those should be the only changes.
>  
> TK
>
>
> ________________________________
> From: fireup_kev <kliboon@...>
> To: DynoMotion@yahoogroups.com
> Sent: Tuesday, February 28, 2012 7:33 AM
> Subject: [DynoMotion] Re: v429z Problem
>
>
>  
> Thanks Tom
>
> So this, and
>
> http://dynomotion.com/Software/Patch/FixTP429Arcs/CoordMotion.cpp
>
> http://dynomotion.com/Software/Patch/FixTP429Arcs/TrajectoryPlanner.cpp
>
> are the changes I need to make to the source to take care of the current bugs reported so far?
>
> Kevin
>
> --- In DynoMotion@yahoogroups.com, Tom Kerekes <tk@> wrote:
> >
> > Hi Kevin,
> >  
> > Yes it looks like (KMotion_dotNet\KM_MainStatus.cs)
> >  {
> >     {
> >     }publicintGetKFlopBitDirection(intindex)    if(index > 31)        returnGetValue(_BitsDirection[1], index - 31, 1);    else    {
> >     }
> > }        returnGetValue(_BitsDirection[0], index, 1);
> >  
> > should be:
> >  publicintGetKFlopBitDirection(intindex)if(index > 31)returnGetValue(_BitsDirection[1], index - 32, 1);else{
> > }
> > }returnGetValue(_BitsDirection[0], index, 1);
> >  
> > Thanks
> > TK
> >
> > From: fireup_kev <kliboon@>
> > To: DynoMotion@yahoogroups.com
> > Sent: Monday, February 27, 2012 2:18 PM
> > Subject: [DynoMotion] Re: v429z Problem
> >
> >
> >  
> > For me, MainStatus still as a bug in the upper bits.
> >
> > GetKFlopBitDiection does not return the correct values for bit 32 to 64
> >
> > Kevin
> >
> > --- In DynoMotion@yahoogroups.com, Brad Murry <bradodarb@> wrote:
> > >
> > > MainStatus works well.
> > >
> > >
> > >
> > > I have not tried the Connected() method with the new patches.
> > >
> > >
> > >
> > > -Brad
> > >
> > >
> > >
> > > From: DynoMotion@yahoogroups.com [mailto:DynoMotion@yahoogroups.com] On Behalf Of himykabibble
> > > Sent: Monday, February 27, 2012 11:59 AM
> > > To: DynoMotion@yahoogroups.com
> > > Subject: [DynoMotion] Re: v429z Problem
> > >
> > >
> > >
> > >
> > >
> > > Tom,
> > >
> > > Whatever that weirdness was yesterday, it did not repeat today, so I think we're OK.
> > >
> > > Any word on when we'll have a dotNet release where Connected works properly?
> > >
> > > Also, is MainStatus now working correctly in dotNet?
> > >
> > > Regards,
> > > Ray L.
> > >
> > > --- In DynoMotion@yahoogroups.com <mailto:DynoMotion%40yahoogroups.com> , "himykabibble" <jagboy@> wrote:
> > > >
> > > > Tom,
> > > >
> > > > Today I ran a batch of the same parts that failed a few days ago, and that problem seems to be fixed. I was able to run one batch of parts that came out perfect.
> > > >
> > > > However, while setting up, I did have a different problem that seems might be related - When I first started up my app, and started doing the setup, I found the interpreter was hanging if I did any MDI command that involved axis motion. I have on-screen indicators that show whether the interpreter is running or not, and it would stay on after any MDI that required axis motion, and the motion would not occur. Any non-move commands, like M3, M7, G54, etc. worked fine. I tried KMotionCNC and got the same behavior. I went back to my app, loaded my program, and started it, and it executed to the first move, then hung. I stopped it, did a rewind, and this time single-stepped through the first few moves, and all seemed fine. So I started it running from the start, and the problem did not re-occur. I'll see if it repeats tomorrow.
> > > >
> > > > Regards,
> > > > Ray L.
> > > >
> > > > --- In DynoMotion@yahoogroups.com <mailto:DynoMotion%40yahoogroups.com> , Tom Kerekes <tk@> wrote:
> > > > >
> > > > > Hi Ray
> > > > > ÃÆ'‚
> > > > > Here areÃÆ'‚ the sources changed from V4.29z:
> > > > > ÃÆ'‚
> > > > > http://dynomotion.com/Software/Patch/FixTP429Arcs/CoordMotion.cpp
> > > > > ÃÆ'‚
> > > > > http://dynomotion.com/Software/Patch/FixTP429Arcs/TrajectoryPlanner.cpp
> > > > > ÃÆ'‚
> > > > > Regards
> > > > > TK
> > > > >
> > > > > From: himykabibble <jagboy@>
> > > > > To: DynoMotion@yahoogroups.com <mailto:DynoMotion%40yahoogroups.com>
> > > > > Sent: Saturday, February 25, 2012 2:35 PM
> > > > > Subject: [DynoMotion] Re: v429z Problem
> > > > >
> > > > >
> > > > > ÃÆ'‚
> > > > > Tom,
> > > > >
> > > > > I was only able to run a quick test, but it appears to be working correctly now. I'll try to run some real parts tomorrow, once I get my machine back together (finally installing limit switches on X, Y and A axes).
> > > > >
> > > > > Can I get the modified source, as I need to make my tool length comp mods before I can run any of my existing G-code.
> > > > >
> > > > > Regards,
> > > > > Ray L.
> > > > >
> > > > > --- In DynoMotion@yahoogroups.com <mailto:DynoMotion%40yahoogroups.com> , Tom Kerekes <tk@> wrote:
> > > > > >
> > > > > > Ray,
> > > > > > ÃÆ'Æ'‚ÃÆ'‚
> > > > > > Please try this patch to fix GCode Arcs for V4.29z.ÃÆ'Æ'‚ÃÆ'‚ Copy to the C:\Kmotion429z\Kmotion\Release directory
> > > > > > ÃÆ'Æ'‚ÃÆ'‚
> > > > > > http://dynomotion.com/Software/Patch/FixTP429Arcs/GCodeInterpreter.dll
> > > > > > ÃÆ'Æ'‚ÃÆ'‚
> > > > > > Regards
> > > > > > TK
> > > > > > ÃÆ'Æ'‚ÃÆ'‚
> > > > > >
> > > > > > From: himykabibble <jagboy@>
> > > > > > To: DynoMotion@yahoogroups.com <mailto:DynoMotion%40yahoogroups.com>
> > > > > > Sent: Thursday, February 23, 2012 2:25 PM
> > > > > > Subject: [DynoMotion] Re: v429z Problem
> > > > > >
> > > > > >
> > > > > > ÃÆ'Æ'‚ÃÆ'‚
> > > > > > Tom,
> > > > > >
> > > > > > Well, how many people really NEED to do arcs anyway.... :-)
> > > > > >
> > > > > > Regards,
> > > > > > Ray L.
> > > > > >
> > > > > > --- In DynoMotion@yahoogroups.com <mailto:DynoMotion%40yahoogroups.com> , Tom Kerekes <tk@> wrote:
> > > > > > >
> > > > > > > Hi Ray,
> > > > > > > ÃÆ'Æ'Æ'ÃÆ'¢â‚¬Å¡ÃÆ'Æ'‚ÃÆ'‚
> > > > > > > I think I see what is going on.ÃÆ'Æ'Æ'ÃÆ'¢â‚¬Å¡ÃÆ'Æ'‚ÃÆ'‚ I re-plotted your data and uploaded the Excel Sheet.ÃÆ'Æ'Æ'ÃÆ'¢â‚¬Å¡ÃÆ'Æ'‚ÃÆ'‚ ÃÆ'Æ'Æ'ÃÆ'¢â‚¬Å¡ÃÆ'Æ'‚ÃÆ'‚ Looks like there is a problem with Arcs getting out of order with Line segments due to the Grouping into blocks of segments that we did in V4.29 to improve USB traffic.ÃÆ'Æ'Æ'ÃÆ'¢â‚¬Å¡ÃÆ'Æ'‚ÃÆ'‚ Everything works fine for Albert (and me) because we weren't doing any Arcs.ÃÆ'Æ'Æ'ÃÆ'¢â‚¬Å¡ÃÆ'Æ'‚ÃÆ'‚ Give me some time to come up with a fix.
> > > > > > > ÃÆ'Æ'Æ'ÃÆ'¢â‚¬Å¡ÃÆ'Æ'‚ÃÆ'‚
> > > > > > > Looks like it wasn't noise :}
> > > > > > > ÃÆ'Æ'Æ'ÃÆ'¢â‚¬Å¡ÃÆ'Æ'‚ÃÆ'‚
> > > > > > > Regards
> > > > > > > TK
> > > > > > >
> > > > > > > From: himykabibble <jagboy@>
> > > > > > > To: DynoMotion@yahoogroups.com <mailto:DynoMotion%40yahoogroups.com>
> > > > > > > Sent: Thursday, February 23, 2012 12:05 PM
> > > > > > > Subject: [DynoMotion] Re: v429z Problem
> > > > > > >
> > > > > > >
> > > > > > > ÃÆ'Æ'Æ'ÃÆ'¢â‚¬Å¡ÃÆ'Æ'‚ÃÆ'‚
> > > > > > > BTW - I can duplicate this with two different KFlops, and two different PCs....
> > > > > > >
> > > > > > > --- In DynoMotion@yahoogroups.com <mailto:DynoMotion%40yahoogroups.com> , "himykabibble" <jagboy@> wrote:
> > > > > > > >
> > > > > > > > OK, three spreadsheets uploaded to RayLs Files. The v428 trace looks the same, regardless of parameter setting. The v429 trace gets goofier the lower acceleration is set. Two captures, for accel = 30.0, and accel = 0.10 are given. All captures were recording Dest on every 50th pass of my Init program. First column is time, second thru fourth are X, Y, Z position.
> > > > > > > >
> > > > > > > > Regards,
> > > > > > > > Ray L.
> > > > > > > >
> > > > > > > > --- In DynoMotion@yahoogroups.com <mailto:DynoMotion%40yahoogroups.com> , "himykabibble" <jagboy@> wrote:
> > > > > > > > >
> > > > > > > > > Tom,
> > > > > > > > >
> > > > > > > > > I'm posting some captures of ch[0-2].Dest for versions v428 and v429z. My machine was not even turned on for this. The problem is quite evident.
> > > > > > > > >
> > > > > > > > > Regards,
> > > > > > > > > Ray L.
> > > > > > > > >
> > > > > > > > > --- In DynoMotion@yahoogroups.com <mailto:DynoMotion%40yahoogroups.com> , "himykabibble" <jagboy@> wrote:
> > > > > > > > > >
> > > > > > > > > > Tom,
> > > > > > > > > >
> > > > > > > > > > At this point, there is not a doubt in my mind that v429z is broken. I have run dozens of tests, and v428 always works, v429z always fails. With v429z, the lower I set acceleration, or feedrate, in KMotionCNC, the worse it performs. With accel set to 25 in/sec^2, I get small, but absolutely consistent, position errors, and NO servo faults. With accel set to 0.1 in/sec^2, I get nearly constant servo faults, with attendant massive position loss. The exact same g-code will work perfectly time after time with v428, with acceleration set as low or as high as I choose, and with feed rates cranked up to near rapid speeds. Even worse, it will sometimes move the Z axis, and there is not a single Z-axis command anywhere in the G-code! On the crooked, non-straight segments, I can put my fingers on the ballscrews, and I can feel clear, smooth motion on the axis that is supposed to be stationary. No chance this is noise - these are deliberate,
> > constant speed
> > > > > > > moves, always in exactly the same place, and they ARE reflected in the DROs.
> > > > > > > > > >
> > > > > > > > > > I've posted the G-code I'm using in RayLs Files.
> > > > > > > > > >
> > > > > > > > > > Regards,
> > > > > > > > > > Ray L.
> > > > > > > > > >
> > > > > > > > >
> > > > > > > >
> > > > > > >
> > > > > >
> > > > >
> > > >
> > >
> >
> >
> >
> >
> >
> > {
> > {
> > }
> >
>
Group: DynoMotion Message: 4146 From: Tom Kerekes Date: 2/29/2012
Subject: Re: v429z Problem
Hi Kevin,
 
I can't duplicate the poblem.  CoordinateMotion should now automatically set the MainPathRoot to the root of where the GCodeInterpreter.dll is located.  It will then use that to find \GCode Programs\emc.var.  It shouldn't matter where your app is located, but the DLL has to be placed relative to that file correctly.
 
Otherwise could you set a breakpoint on the line:
 
CString Name=(CString)(GC->CoordMotion->MainPathRoot) +
\\GCode Programs\\emc.var;
 
and see what value it has?
 
Thanks
TK

Group: DynoMotion Message: 4147 From: fireup_kev Date: 2/29/2012
Subject: Re: v429z Problem
I also moved GCodeIinterperter.dll and all dlls needed for my application to a new directory. It gives me garbage for (GC->CoordMotion->MainPathRoot).

I changed:
CString Name=(CString)(GC->CoordMotion->MainPathRoot) + ""\\GCode Programs\\emc.var";

to CString Name = "emc.var"; and just copy that emc.var file to my exe directory and it works for now...

--- In DynoMotion@yahoogroups.com, Tom Kerekes <tk@...> wrote:
>
> Hi Kevin,
>  
> I can't duplicate the poblem.  CoordinateMotion should now automatically set the MainPathRoot to the root of where the GCodeInterpreter.dll is located.  It will then use that to find \GCode Programs\emc.var.  It shouldn't matter where your app is located, but the DLL has to be placed relative to that file correctly.
>  
> Otherwise could you set a breakpoint on the line:
>  CString Name=(CString)(GC->CoordMotion->MainPathRoot) +
>  
> and see what value it has?
>  
> Thanks
> TK\\GCode Programs\\emc.var;
>
> From: fireup_kev <kliboon@...>
> To: DynoMotion@yahoogroups.com
> Sent: Wednesday, February 29, 2012 8:52 PM
> Subject: [DynoMotion] Re: v429z Problem
>
>
>  
> I found another problem....
>
> When I move my application out to it's own directory, it gives me an unable to open file "\GCode Programs\emc.var" error. Last version I specified CoordMotion.Interpreter.MainPathRoot but 4.29z no longer has this value and CoordMotion.Interpreter.SetupFile does not seem to be the same thing.
>
> Thanks,
> Kevin
>
> --- In DynoMotion@yahoogroups.com, Tom Kerekes <tk@> wrote:
> >
> > Hi Kevin,
> >  
> > Yes those should be the only changes.
> >  
> > TK
> >
> >
> > ________________________________
> > From: fireup_kev <kliboon@>
> > To: DynoMotion@yahoogroups.com
> > Sent: Tuesday, February 28, 2012 7:33 AM
> > Subject: [DynoMotion] Re: v429z Problem
> >
> >
> >  
> > Thanks Tom
> >
> > So this, and
> >
> > http://dynomotion.com/Software/Patch/FixTP429Arcs/CoordMotion.cpp
> >
> > http://dynomotion.com/Software/Patch/FixTP429Arcs/TrajectoryPlanner.cpp
> >
> > are the changes I need to make to the source to take care of the current bugs reported so far?
> >
> > Kevin
> >
> > --- In DynoMotion@yahoogroups.com, Tom Kerekes <tk@> wrote:
> > >
> > > Hi Kevin,
> > >  
> > > Yes it looks like (KMotion_dotNet\KM_MainStatus.cs)
> > >  {
> > >     {
> > >     }publicintGetKFlopBitDirection(intindex)    if(index > 31)        returnGetValue(_BitsDirection[1], index - 31, 1);    else    {
> > >     }
> > > }        returnGetValue(_BitsDirection[0], index, 1);
> > >  
> > > should be:
> > >  publicintGetKFlopBitDirection(intindex)if(index > 31)returnGetValue(_BitsDirection[1], index - 32, 1);else{
> > > }
> > > }returnGetValue(_BitsDirection[0], index, 1);
> > >  
> > > Thanks
> > > TK
> > >
> > > From: fireup_kev <kliboon@>
> > > To: DynoMotion@yahoogroups.com
> > > Sent: Monday, February 27, 2012 2:18 PM
> > > Subject: [DynoMotion] Re: v429z Problem
> > >
> > >
> > >  
> > > For me, MainStatus still as a bug in the upper bits.
> > >
> > > GetKFlopBitDiection does not return the correct values for bit 32 to 64
> > >
> > > Kevin
> > >
> > > --- In DynoMotion@yahoogroups.com, Brad Murry <bradodarb@> wrote:
> > > >
> > > > MainStatus works well.
> > > >
> > > >
> > > >
> > > > I have not tried the Connected() method with the new patches.
> > > >
> > > >
> > > >
> > > > -Brad
> > > >
> > > >
> > > >
> > > > From: DynoMotion@yahoogroups.com [mailto:DynoMotion@yahoogroups.com] On Behalf Of himykabibble
> > > > Sent: Monday, February 27, 2012 11:59 AM
> > > > To: DynoMotion@yahoogroups.com
> > > > Subject: [DynoMotion] Re: v429z Problem
> > > >
> > > >
> > > >
> > > >
> > > >
> > > > Tom,
> > > >
> > > > Whatever that weirdness was yesterday, it did not repeat today, so I think we're OK.
> > > >
> > > > Any word on when we'll have a dotNet release where Connected works properly?
> > > >
> > > > Also, is MainStatus now working correctly in dotNet?
> > > >
> > > > Regards,
> > > > Ray L.
> > > >
> > > > --- In DynoMotion@yahoogroups.com <mailto:DynoMotion%40yahoogroups.com> , "himykabibble" <jagboy@> wrote:
> > > > >
> > > > > Tom,
> > > > >
> > > > > Today I ran a batch of the same parts that failed a few days ago, and that problem seems to be fixed. I was able to run one batch of parts that came out perfect.
> > > > >
> > > > > However, while setting up, I did have a different problem that seems might be related - When I first started up my app, and started doing the setup, I found the interpreter was hanging if I did any MDI command that involved axis motion. I have on-screen indicators that show whether the interpreter is running or not, and it would stay on after any MDI that required axis motion, and the motion would not occur. Any non-move commands, like M3, M7, G54, etc. worked fine. I tried KMotionCNC and got the same behavior. I went back to my app, loaded my program, and started it, and it executed to the first move, then hung. I stopped it, did a rewind, and this time single-stepped through the first few moves, and all seemed fine. So I started it running from the start, and the problem did not re-occur. I'll see if it repeats tomorrow.
> > > > >
> > > > > Regards,
> > > > > Ray L.
> > > > >
> > > > > --- In DynoMotion@yahoogroups.com <mailto:DynoMotion%40yahoogroups.com> , Tom Kerekes <tk@> wrote:
> > > > > >
> > > > > > Hi Ray
> > > > > > ÃÆ'Æ'‚
> > > > > > Here areÃÆ'Æ'‚ the sources changed from V4.29z:
> > > > > > ÃÆ'Æ'‚
> > > > > > http://dynomotion.com/Software/Patch/FixTP429Arcs/CoordMotion.cpp
> > > > > > ÃÆ'Æ'‚
> > > > > > http://dynomotion.com/Software/Patch/FixTP429Arcs/TrajectoryPlanner.cpp
> > > > > > ÃÆ'Æ'‚
> > > > > > Regards
> > > > > > TK
> > > > > >
> > > > > > From: himykabibble <jagboy@>
> > > > > > To: DynoMotion@yahoogroups.com <mailto:DynoMotion%40yahoogroups.com>
> > > > > > Sent: Saturday, February 25, 2012 2:35 PM
> > > > > > Subject: [DynoMotion] Re: v429z Problem
> > > > > >
> > > > > >
> > > > > > ÃÆ'Æ'‚
> > > > > > Tom,
> > > > > >
> > > > > > I was only able to run a quick test, but it appears to be working correctly now. I'll try to run some real parts tomorrow, once I get my machine back together (finally installing limit switches on X, Y and A axes).
> > > > > >
> > > > > > Can I get the modified source, as I need to make my tool length comp mods before I can run any of my existing G-code.
> > > > > >
> > > > > > Regards,
> > > > > > Ray L.
> > > > > >
> > > > > > --- In DynoMotion@yahoogroups.com <mailto:DynoMotion%40yahoogroups.com> , Tom Kerekes <tk@> wrote:
> > > > > > >
> > > > > > > Ray,
> > > > > > > ÃÆ'Æ'Æ'ÃÆ'¢â‚¬Å¡ÃÆ'Æ'‚
> > > > > > > Please try this patch to fix GCode Arcs for V4.29z.ÃÆ'Æ'Æ'ÃÆ'¢â‚¬Å¡ÃÆ'Æ'‚ Copy to the C:\Kmotion429z\Kmotion\Release directory
> > > > > > > ÃÆ'Æ'Æ'ÃÆ'¢â‚¬Å¡ÃÆ'Æ'‚
> > > > > > > http://dynomotion.com/Software/Patch/FixTP429Arcs/GCodeInterpreter.dll
> > > > > > > ÃÆ'Æ'Æ'ÃÆ'¢â‚¬Å¡ÃÆ'Æ'‚
> > > > > > > Regards
> > > > > > > TK
> > > > > > > ÃÆ'Æ'Æ'ÃÆ'¢â‚¬Å¡ÃÆ'Æ'‚
> > > > > > >
> > > > > > > From: himykabibble <jagboy@>
> > > > > > > To: DynoMotion@yahoogroups.com <mailto:DynoMotion%40yahoogroups.com>
> > > > > > > Sent: Thursday, February 23, 2012 2:25 PM
> > > > > > > Subject: [DynoMotion] Re: v429z Problem
> > > > > > >
> > > > > > >
> > > > > > > ÃÆ'Æ'Æ'ÃÆ'¢â‚¬Å¡ÃÆ'Æ'‚
> > > > > > > Tom,
> > > > > > >
> > > > > > > Well, how many people really NEED to do arcs anyway.... :-)
> > > > > > >
> > > > > > > Regards,
> > > > > > > Ray L.
> > > > > > >
> > > > > > > --- In DynoMotion@yahoogroups.com <mailto:DynoMotion%40yahoogroups.com> , Tom Kerekes <tk@> wrote:
> > > > > > > >
> > > > > > > > Hi Ray,
> > > > > > > > ÃÆ'Æ'Æ'ÃÆ'†'ÃÆ'Æ'¢ÃÆ'¢â€šÂ¬ÃÆ'…¡ÃÆ'Æ'Æ'ÃÆ'¢â‚¬Å¡ÃÆ'Æ'‚
> > > > > > > > I think I see what is going on.ÃÆ'Æ'Æ'ÃÆ'†'ÃÆ'Æ'¢ÃÆ'¢â€šÂ¬ÃÆ'…¡ÃÆ'Æ'Æ'ÃÆ'¢â‚¬Å¡ÃÆ'Æ'‚ I re-plotted your data and uploaded the Excel Sheet.ÃÆ'Æ'Æ'ÃÆ'†'ÃÆ'Æ'¢ÃÆ'¢â€šÂ¬ÃÆ'…¡ÃÆ'Æ'Æ'ÃÆ'¢â‚¬Å¡ÃÆ'Æ'‚ ÃÆ'Æ'Æ'ÃÆ'†'ÃÆ'Æ'¢ÃÆ'¢â€šÂ¬ÃÆ'…¡ÃÆ'Æ'Æ'ÃÆ'¢â‚¬Å¡ÃÆ'Æ'‚ Looks like there is a problem with Arcs getting out of order with Line segments due to the Grouping into blocks of segments that we did in V4.29 to improve USB traffic.ÃÆ'Æ'Æ'ÃÆ'†'ÃÆ'Æ'¢ÃÆ'¢â€šÂ¬ÃÆ'…¡ÃÆ'Æ'Æ'ÃÆ'¢â‚¬Å¡ÃÆ'Æ'‚ Everything works fine for Albert (and me) because we
> weren't doing any Arcs.ÃÆ'Æ'Æ'ÃÆ'†'ÃÆ'Æ'¢ÃÆ'¢â€šÂ¬ÃÆ'…¡ÃÆ'Æ'Æ'ÃÆ'¢â‚¬Å¡ÃÆ'Æ'‚ Give me some time to come up with a fix.
> > > > > > > > ÃÆ'Æ'Æ'ÃÆ'†'ÃÆ'Æ'¢ÃÆ'¢â€šÂ¬ÃÆ'…¡ÃÆ'Æ'Æ'ÃÆ'¢â‚¬Å¡ÃÆ'Æ'‚
> > > > > > > > Looks like it wasn't noise :}
> > > > > > > > ÃÆ'Æ'Æ'ÃÆ'†'ÃÆ'Æ'¢ÃÆ'¢â€šÂ¬ÃÆ'…¡ÃÆ'Æ'Æ'ÃÆ'¢â‚¬Å¡ÃÆ'Æ'‚
> > > > > > > > Regards
> > > > > > > > TK
> > > > > > > >
> > > > > > > > From: himykabibble <jagboy@>
> > > > > > > > To: DynoMotion@yahoogroups.com <mailto:DynoMotion%40yahoogroups.com>
> > > > > > > > Sent: Thursday, February 23, 2012 12:05 PM
> > > > > > > > Subject: [DynoMotion] Re: v429z Problem
> > > > > > > >
> > > > > > > >
> > > > > > > > ÃÆ'Æ'Æ'ÃÆ'†'ÃÆ'Æ'¢ÃÆ'¢â€šÂ¬ÃÆ'…¡ÃÆ'Æ'Æ'ÃÆ'¢â‚¬Å¡ÃÆ'Æ'‚
> > > > > > > > BTW - I can duplicate this with two different KFlops, and two different PCs....
> > > > > > > >
> > > > > > > > --- In DynoMotion@yahoogroups.com <mailto:DynoMotion%40yahoogroups.com> , "himykabibble" <jagboy@> wrote:
> > > > > > > > >
> > > > > > > > > OK, three spreadsheets uploaded to RayLs Files. The v428 trace looks the same, regardless of parameter setting. The v429 trace gets goofier the lower acceleration is set. Two captures, for accel = 30.0, and accel = 0.10 are given. All captures were recording Dest on every 50th pass of my Init program. First column is time, second thru fourth are X, Y, Z position.
> > > > > > > > >
> > > > > > > > > Regards,
> > > > > > > > > Ray L.
> > > > > > > > >
> > > > > > > > > --- In DynoMotion@yahoogroups.com <mailto:DynoMotion%40yahoogroups.com> , "himykabibble" <jagboy@> wrote:
> > > > > > > > > >
> > > > > > > > > > Tom,
> > > > > > > > > >
> > > > > > > > > > I'm posting some captures of ch[0-2].Dest for versions v428 and v429z. My machine was not even turned on for this. The problem is quite evident.
> > > > > > > > > >
> > > > > > > > > > Regards,
> > > > > > > > > > Ray L.
> > > > > > > > > >
> > > > > > > > > > --- In DynoMotion@yahoogroups.com <mailto:DynoMotion%40yahoogroups.com> , "himykabibble" <jagboy@> wrote:
> > > > > > > > > > >
> > > > > > > > > > > Tom,
> > > > > > > > > > >
> > > > > > > > > > > At this point, there is not a doubt in my mind that v429z is broken. I have run dozens of tests, and v428 always works, v429z always fails. With v429z, the lower I set acceleration, or feedrate, in KMotionCNC, the worse it performs. With accel set to 25 in/sec^2, I get small, but absolutely consistent, position errors, and NO servo faults. With accel set to 0.1 in/sec^2, I get nearly constant servo faults, with attendant massive position loss. The exact same g-code will work perfectly time after time with v428, with acceleration set as low or as high as I choose, and with feed rates cranked up to near rapid speeds. Even worse, it will sometimes move the Z axis, and there is not a single Z-axis command anywhere in the G-code! On the crooked, non-straight segments, I can put my fingers on the ballscrews, and I can feel clear, smooth motion on the axis that is supposed to be stationary. No chance this is noise - these are deliberate,
> > > constant speed
> > > > > > > > moves, always in exactly the same place, and they ARE reflected in the DROs.
> > > > > > > > > > >
> > > > > > > > > > > I've posted the G-code I'm using in RayLs Files.
> > > > > > > > > > >
> > > > > > > > > > > Regards,
> > > > > > > > > > > Ray L.
> > > > > > > > > > >
> > > > > > > > > >
> > > > > > > > >
> > > > > > > >
> > > > > > >
> > > > > >
> > > > >
> > > >
> > >
> > >
> > >
> > >
> > >
> > > {
> > > {
> > > }
> > >
> >
>
Group: DynoMotion Message: 4148 From: Brad Murry Date: 2/29/2012
Subject: Re: v429z Problem

Kevin,

 

 

     It was recommended to leave the MainPathRoot out of the API and let cKMotionDLL.cpp define that.

 

The underlying libraries expect a very specific folder structure.  Even if you copied the files in the KMotion\Release things will likely become unstable.

 

I have found it is  reliable to leave the KMotion4.xx folder alone and just load the KMotion_dotNet.dll at runtime.

 

I use the following code in MM.  It loads a dll that provides interfaces of some basic machine objects.  You can simplify it to only load the KMotion_dotNet.dll

 

        private void LoadMachineAssembly()

        {

            var machinedll = _AssemblyPath;

            if (File.Exists(machinedll))

            {

                var type = typeof(IMachineProvider);

                Assembly a = Assembly.LoadFrom(machinedll);

                var machineprovider = a.GetTypes().Where(p => type.IsAssignableFrom(p)).FirstOrDefault();

                if (machineprovider != null)

                {

                    var instance = Activator.CreateInstance(machineprovider);

                    _Provider = instance as IMachineProvider;

                }

                else

                {

                    throw new Exception("No valid Machine Providers found; Library must contain at least one class that implements IMachineProvider.");

                }

            }

            else

            {

                throw new FileNotFoundException("Invalid Machine Provider File");

            }

        }

 

The IMachineProvider interface can be as simple as you like and call it anything you want.  Implement this interface  with a class that houses an instance of KM_Controller.

 

As I said, you could also just load KMotion_dotNet.dll with the above code.  Your app already has a reference to the dll, just set it’s Copy Local to false and load the DLL like above.

 

-Brad Murry

 

 

From: DynoMotion@yahoogroups.com [mailto:DynoMotion@yahoogroups.com] On Behalf Of fireup_kev
Sent: Wednesday, February 29, 2012 9:52 PM
To: DynoMotion@yahoogroups.com
Subject: [DynoMotion] Re: v429z Problem

 

 

I found another problem....

When I move my application out to it's own directory, it gives me an unable to open file "\GCode Programs\emc.var" error. Last version I specified CoordMotion.Interpreter.MainPathRoot but 4.29z no longer has this value and CoordMotion.Interpreter.SetupFile does not seem to be the same thing.

Thanks,
Kevin

--- In DynoMotion@yahoogroups.com, Tom Kerekes <tk@...> wrote:
>
> Hi Kevin,
>  
> Yes those should be the only changes.
>  
> TK
>
>
> ________________________________
> From: fireup_kev <kliboon@...>
> To: DynoMotion@yahoogroups.com
> Sent: Tuesday, February 28, 2012 7:33 AM
> Subject: [DynoMotion] Re: v429z Problem
>
>
>  
> Thanks Tom
>
> So this, and
>
> http://dynomotion.com/Software/Patch/FixTP429Arcs/CoordMotion.cpp
>
> http://dynomotion.com/Software/Patch/FixTP429Arcs/TrajectoryPlanner.cpp
>
> are the changes I need to make to the source to take care of the current bugs reported so far?
>
> Kevin
>
> --- In DynoMotion@yahoogroups.com, Tom Kerekes <tk@> wrote:
> >
> > Hi Kevin,
> >  
> > Yes it looks like (KMotion_dotNet\KM_MainStatus.cs)
> >  {
> >     {
> >     }publicintGetKFlopBitDirection(intindex)    if(index > 31)        returnGetValue(_BitsDirection[1], index - 31, 1);    else    {
> >     }
> > }        returnGetValue(_BitsDirection[0], index, 1);
> >  
> > should be:
> >  publicintGetKFlopBitDirection(intindex)if(index > 31)returnGetValue(_BitsDirection[1], index - 32, 1);else{
> > }
> > }returnGetValue(_BitsDirection[0], index, 1);
> >  
> > Thanks
> > TK
> >
> > From: fireup_kev <kliboon@>
> > To: DynoMotion@yahoogroups.com
> > Sent: Monday, February 27, 2012 2:18 PM
> > Subject: [DynoMotion] Re: v429z Problem
> >
> >
> >  
> > For me, MainStatus still as a bug in the upper bits.
> >
> > GetKFlopBitDiection does not return the correct values for bit 32 to 64
> >
> > Kevin
> >
> > --- In DynoMotion@yahoogroups.com, Brad Murry <bradodarb@> wrote:
> > >
> > > MainStatus works well.
> > >
> > >
> > >
> > > I have not tried the Connected() method with the new patches.
> > >
> > >
> > >
> > > -Brad
> > >
> > >
> > >
> > > From: DynoMotion@yahoogroups.com [mailto:DynoMotion@yahoogroups.com] On Behalf Of himykabibble
> > > Sent: Monday, February 27, 2012 11:59 AM
> > > To: DynoMotion@yahoogroups.com
> > > Subject: [DynoMotion] Re: v429z Problem
> > >
> > >
> > >
> > >
> > >
> > > Tom,
> > >
> > > Whatever that weirdness was yesterday, it did not repeat today, so I think we're OK.
> > >
> > > Any word on when we'll have a dotNet release where Connected works properly?
> > >
> > > Also, is MainStatus now working correctly in dotNet?
> > >
> > > Regards,
> > > Ray L.
> > >
> > > --- In DynoMotion@yahoogroups.com <mailto:DynoMotion%40yahoogroups.com> , "himykabibble" <jagboy@> wrote:
> > > >
> > > > Tom,
> > > >
> > > > Today I ran a batch of the same parts that failed a few days ago, and that problem seems to be fixed. I was able to run one batch of parts that came out perfect.
> > > >
> > > > However, while setting up, I did have a different problem that seems might be related - When I first started up my app, and started doing the setup, I found the interpreter was hanging if I did any MDI command that involved axis motion. I have on-screen indicators that show whether the interpreter is running or not, and it would stay on after any MDI that required axis motion, and the motion would not occur. Any non-move commands, like M3, M7, G54, etc. worked fine. I tried KMotionCNC and got the same behavior. I went back to my app, loaded my program, and started it, and it executed to the first move, then hung. I stopped it, did a rewind, and this time single-stepped through the first few moves, and all seemed fine. So I started it running from the start, and the problem did not re-occur. I'll see if it repeats tomorrow.
> > > >
> > > > Regards,
> > > > Ray L.
> > > >
> > > > --- In DynoMotion@yahoogroups.com <mailto:DynoMotion%40yahoogroups.com> , Tom Kerekes <tk@> wrote:
> > > > >
> > > > > Hi Ray
> > > > > ÃÆ'‚
> > > > > Here areÃÆ'‚ the sources changed from V4.29z:
> > > > > ÃÆ'‚
> > > > > http://dynomotion.com/Software/Patch/FixTP429Arcs/CoordMotion.cpp
> > > > > ÃÆ'‚
> > > > > http://dynomotion.com/Software/Patch/FixTP429Arcs/TrajectoryPlanner.cpp
> > > > > ÃÆ'‚
> > > > > Regards
> > > > > TK
> > > > >
> > > > > From: himykabibble <jagboy@>
> > > > > To: DynoMotion@yahoogroups.com <mailto:DynoMotion%40yahoogroups.com>
> > > > > Sent: Saturday, February 25, 2012 2:35 PM
> > > > > Subject: [DynoMotion] Re: v429z Problem
> > > > >
> > > > >
> > > > > ÃÆ'‚
> > > > > Tom,
> > > > >
> > > > > I was only able to run a quick test, but it appears to be working correctly now. I'll try to run some real parts tomorrow, once I get my machine back together (finally installing limit switches on X, Y and A axes).
> > > > >
> > > > > Can I get the modified source, as I need to make my tool length comp mods before I can run any of my existing G-code.
> > > > >
> > > > > Regards,
> > > > > Ray L.
> > > > >
> > > > > --- In DynoMotion@yahoogroups.com <mailto:DynoMotion%40yahoogroups.com> , Tom Kerekes <tk@> wrote:
> > > > > >
> > > > > > Ray,
> > > > > > ÃÆ'Æ'‚ÃÆ'‚
> > > > > > Please try this patch to fix GCode Arcs for V4.29z.ÃÆ'Æ'‚ÃÆ'‚ Copy to the C:\Kmotion429z\Kmotion\Release directory
> > > > > > ÃÆ'Æ'‚ÃÆ'‚
> > > > > > http://dynomotion.com/Software/Patch/FixTP429Arcs/GCodeInterpreter.dll
> > > > > > ÃÆ'Æ'‚ÃÆ'‚
> > > > > > Regards
> > > > > > TK
> > > > > > ÃÆ'Æ'‚ÃÆ'‚
> > > > > >
> > > > > > From: himykabibble <jagboy@>
> > > > > > To: DynoMotion@yahoogroups.com <mailto:DynoMotion%40yahoogroups.com>
> > > > > > Sent: Thursday, February 23, 2012 2:25 PM
> > > > > > Subject: [DynoMotion] Re: v429z Problem
> > > > > >
> > > > > >
> > > > > > ÃÆ'Æ'‚ÃÆ'‚
> > > > > > Tom,
> > > > > >
> > > > > > Well, how many people really NEED to do arcs anyway.... :-)
> > > > > >
> > > > > > Regards,
> > > > > > Ray L.
> > > > > >
> > > > > > --- In DynoMotion@yahoogroups.com <mailto:DynoMotion%40yahoogroups.com> , Tom Kerekes <tk@> wrote:
> > > > > > >
> > > > > > > Hi Ray,
> > > > > > > ÃÆ'Æ'Æ'ÃÆ'¢â‚¬Å¡ÃÆ'Æ'‚ÃÆ'‚
> > > > > > > I think I see what is going on.ÃÆ'Æ'Æ'ÃÆ'¢â‚¬Å¡ÃÆ'Æ'‚ÃÆ'‚ I re-plotted your data and uploaded the Excel Sheet.ÃÆ'Æ'Æ'ÃÆ'¢â‚¬Å¡ÃÆ'Æ'‚ÃÆ'‚ ÃÆ'Æ'Æ'ÃÆ'¢â‚¬Å¡ÃÆ'Æ'‚ÃÆ'‚ Looks like there is a problem with Arcs getting out of order with Line segments due to the Grouping into blocks of segments that we did in V4.29 to improve USB traffic.ÃÆ'Æ'Æ'ÃÆ'¢â‚¬Å¡ÃÆ'Æ'‚ÃÆ'‚ Everything works fine for Albert (and me) because we weren't doing any Arcs.ÃÆ'Æ'Æ'ÃÆ'¢â‚¬Å¡ÃÆ'Æ'‚ÃÆ'‚ Give me some time to come up with a fix.
> > > > > > > ÃÆ'Æ'Æ'ÃÆ'¢â‚¬Å¡ÃÆ'Æ'‚ÃÆ'‚
> > > > > > > Looks like it wasn't noise :}
> > > > > > > ÃÆ'Æ'Æ'ÃÆ'¢â‚¬Å¡ÃÆ'Æ'‚ÃÆ'‚
> > > > > > > Regards
> > > > > > > TK
> > > > > > >
> > > > > > > From: himykabibble <jagboy@>
> > > > > > > To: DynoMotion@yahoogroups.com <mailto:DynoMotion%40yahoogroups.com>
> > > > > > > Sent: Thursday, February 23, 2012 12:05 PM
> > > > > > > Subject: [DynoMotion] Re: v429z Problem
> > > > > > >
> > > > > > >
> > > > > > > ÃÆ'Æ'Æ'ÃÆ'¢â‚¬Å¡ÃÆ'Æ'‚ÃÆ'‚
> > > > > > > BTW - I can duplicate this with two different KFlops, and two different PCs....
> > > > > > >
> > > > > > > --- In DynoMotion@yahoogroups.com <mailto:DynoMotion%40yahoogroups.com> , "himykabibble" <jagboy@> wrote:
> > > > > > > >
> > > > > > > > OK, three spreadsheets uploaded to RayLs Files. The v428 trace looks the same, regardless of parameter setting. The v429 trace gets goofier the lower acceleration is set. Two captures, for accel = 30.0, and accel = 0.10 are given. All captures were recording Dest on every 50th pass of my Init program. First column is time, second thru fourth are X, Y, Z position.
> > > > > > > >
> > > > > > > > Regards,
> > > > > > > > Ray L.
> > > > > > > >
> > > > > > > > --- In DynoMotion@yahoogroups.com <mailto:DynoMotion%40yahoogroups.com> , "himykabibble" <jagboy@> wrote:
> > > > > > > > >
> > > > > > > > > Tom,
> > > > > > > > >
> > > > > > > > > I'm posting some captures of ch[0-2].Dest for versions v428 and v429z. My machine was not even turned on for this. The problem is quite evident.
> > > > > > > > >
> > > > > > > > > Regards,
> > > > > > > > > Ray L.
> > > > > > > > >
> > > > > > > > > --- In DynoMotion@yahoogroups.com <mailto:DynoMotion%40yahoogroups.com> , "himykabibble" <jagboy@> wrote:
> > > > > > > > > >
> > > > > > > > > > Tom,
> > > > > > > > > >
> > > > > > > > > > At this point, there is not a doubt in my mind that v429z is broken. I have run dozens of tests, and v428 always works, v429z always fails. With v429z, the lower I set acceleration, or feedrate, in KMotionCNC, the worse it performs. With accel set to 25 in/sec^2, I get small, but absolutely consistent, position errors, and NO servo faults. With accel set to 0.1 in/sec^2, I get nearly constant servo faults, with attendant massive position loss. The exact same g-code will work perfectly time after time with v428, with acceleration set as low or as high as I choose, and with feed rates cranked up to near rapid speeds. Even worse, it will sometimes move the Z axis, and there is not a single Z-axis command anywhere in the G-code! On the crooked, non-straight segments, I can put my fingers on the ballscrews, and I can feel clear, smooth motion on the axis that is supposed to be stationary. No chance this is noise - these are deliberate,
> > constant speed
> > > > > > > moves, always in exactly the same place, and they ARE reflected in the DROs.
> > > > > > > > > >
> > > > > > > > > > I've posted the G-code I'm using in RayLs Files.
> > > > > > > > > >
> > > > > > > > > > Regards,
> > > > > > > > > > Ray L.
> > > > > > > > > >
> > > > > > > > >
> > > > > > > >
> > > > > > >
> > > > > >
> > > > >
> > > >
> > >
> >
> >
> >
> >
> >
> > {
> > {
> > }
> >
>
Group: DynoMotion Message: 4149 From: fireup_kev Date: 2/29/2012
Subject: Re: v429z Problem
It's just my preference but I like to have all the dll's needed for my application in the same directory or sub-directory. Also I like to make one simple setup file that the user don't have to install KMotion4xx.exe.

--- In DynoMotion@yahoogroups.com, Brad Murry <bradodarb@...> wrote:
>
> Kevin,
>
>
>
>
>
> It was recommended to leave the MainPathRoot out of the API and let cKMotionDLL.cpp define that.
>
>
>
> The underlying libraries expect a very specific folder structure. Even if you copied the files in the KMotion\Release things will likely become unstable.
>
>
>
> I have found it is reliable to leave the KMotion4.xx folder alone and just load the KMotion_dotNet.dll at runtime.
>
>
>
> I use the following code in MM. It loads a dll that provides interfaces of some basic machine objects. You can simplify it to only load the KMotion_dotNet.dll
>
>
>
> private void LoadMachineAssembly()
>
> {
>
> var machinedll = _AssemblyPath;
>
> if (File.Exists(machinedll))
>
> {
>
> var type = typeof(IMachineProvider);
>
> Assembly a = Assembly.LoadFrom(machinedll);
>
> var machineprovider = a.GetTypes().Where(p => type.IsAssignableFrom(p)).FirstOrDefault();
>
> if (machineprovider != null)
>
> {
>
> var instance = Activator.CreateInstance(machineprovider);
>
> _Provider = instance as IMachineProvider;
>
> }
>
> else
>
> {
>
> throw new Exception("No valid Machine Providers found; Library must contain at least one class that implements IMachineProvider.");
>
> }
>
> }
>
> else
>
> {
>
> throw new FileNotFoundException("Invalid Machine Provider File");
>
> }
>
> }
>
>
>
> The IMachineProvider interface can be as simple as you like and call it anything you want. Implement this interface with a class that houses an instance of KM_Controller.
>
>
>
> As I said, you could also just load KMotion_dotNet.dll with the above code. Your app already has a reference to the dll, just set it’s Copy Local to false and load the DLL like above.
>
>
>
> -Brad Murry
>
>
>
>
>
> From: DynoMotion@yahoogroups.com [mailto:DynoMotion@yahoogroups.com] On Behalf Of fireup_kev
> Sent: Wednesday, February 29, 2012 9:52 PM
> To: DynoMotion@yahoogroups.com
> Subject: [DynoMotion] Re: v429z Problem
>
>
>
>
>
> I found another problem....
>
> When I move my application out to it's own directory, it gives me an unable to open file "\GCode Programs\emc.var" error. Last version I specified CoordMotion.Interpreter.MainPathRoot but 4.29z no longer has this value and CoordMotion.Interpreter.SetupFile does not seem to be the same thing.
>
> Thanks,
> Kevin
>
> --- In DynoMotion@yahoogroups.com <mailto:DynoMotion%40yahoogroups.com> , Tom Kerekes <tk@> wrote:
> >
> > Hi Kevin,
> > Â
> > Yes those should be the only changes.
> > Â
> > TK
> >
> >
> > ________________________________
> > From: fireup_kev <kliboon@>
> > To: DynoMotion@yahoogroups.com <mailto:DynoMotion%40yahoogroups.com>
> > Sent: Tuesday, February 28, 2012 7:33 AM
> > Subject: [DynoMotion] Re: v429z Problem
> >
> >
> > Â
> > Thanks Tom
> >
> > So this, and
> >
> > http://dynomotion.com/Software/Patch/FixTP429Arcs/CoordMotion.cpp
> >
> > http://dynomotion.com/Software/Patch/FixTP429Arcs/TrajectoryPlanner.cpp
> >
> > are the changes I need to make to the source to take care of the current bugs reported so far?
> >
> > Kevin
> >
> > --- In DynoMotion@yahoogroups.com <mailto:DynoMotion%40yahoogroups.com> , Tom Kerekes <tk@> wrote:
> > >
> > > Hi Kevin,
> > > ÂÂ
> > > Yes it looks like (KMotion_dotNet\KM_MainStatus.cs)
> > >  {
> > >    {
> > >    }publicintGetKFlopBitDirection(intindex)   if(index > 31)       returnGetValue(_BitsDirection[1], index - 31, 1);   else   {
> > >    }
> > > }       returnGetValue(_BitsDirection[0], index, 1);
> > > ÂÂ
> > > should be:
> > >  publicintGetKFlopBitDirection(intindex)if(index > 31)returnGetValue(_BitsDirection[1], index - 32, 1);else{
> > > }
> > > }returnGetValue(_BitsDirection[0], index, 1);
> > > ÂÂ
> > > Thanks
> > > TK
> > >
> > > From: fireup_kev <kliboon@>
> > > To: DynoMotion@yahoogroups.com <mailto:DynoMotion%40yahoogroups.com>
> > > Sent: Monday, February 27, 2012 2:18 PM
> > > Subject: [DynoMotion] Re: v429z Problem
> > >
> > >
> > > ÂÂ
> > > For me, MainStatus still as a bug in the upper bits.
> > >
> > > GetKFlopBitDiection does not return the correct values for bit 32 to 64
> > >
> > > Kevin
> > >
> > > --- In DynoMotion@yahoogroups.com <mailto:DynoMotion%40yahoogroups.com> , Brad Murry <bradodarb@> wrote:
> > > >
> > > > MainStatus works well.
> > > >
> > > >
> > > >
> > > > I have not tried the Connected() method with the new patches.
> > > >
> > > >
> > > >
> > > > -Brad
> > > >
> > > >
> > > >
> > > > From: DynoMotion@yahoogroups.com <mailto:DynoMotion%40yahoogroups.com> [mailto:DynoMotion@yahoogroups.com <mailto:DynoMotion%40yahoogroups.com> ] On Behalf Of himykabibble
> > > > Sent: Monday, February 27, 2012 11:59 AM
> > > > To: DynoMotion@yahoogroups.com <mailto:DynoMotion%40yahoogroups.com>
> > > > Subject: [DynoMotion] Re: v429z Problem
> > > >
> > > >
> > > >
> > > >
> > > >
> > > > Tom,
> > > >
> > > > Whatever that weirdness was yesterday, it did not repeat today, so I think we're OK.
> > > >
> > > > Any word on when we'll have a dotNet release where Connected works properly?
> > > >
> > > > Also, is MainStatus now working correctly in dotNet?
> > > >
> > > > Regards,
> > > > Ray L.
> > > >
> > > > --- In DynoMotion@yahoogroups.com <mailto:DynoMotion%40yahoogroups.com> <mailto:DynoMotion%40yahoogroups.com> , "himykabibble" <jagboy@> wrote:
> > > > >
> > > > > Tom,
> > > > >
> > > > > Today I ran a batch of the same parts that failed a few days ago, and that problem seems to be fixed. I was able to run one batch of parts that came out perfect.
> > > > >
> > > > > However, while setting up, I did have a different problem that seems might be related - When I first started up my app, and started doing the setup, I found the interpreter was hanging if I did any MDI command that involved axis motion. I have on-screen indicators that show whether the interpreter is running or not, and it would stay on after any MDI that required axis motion, and the motion would not occur. Any non-move commands, like M3, M7, G54, etc. worked fine. I tried KMotionCNC and got the same behavior. I went back to my app, loaded my program, and started it, and it executed to the first move, then hung. I stopped it, did a rewind, and this time single-stepped through the first few moves, and all seemed fine. So I started it running from the start, and the problem did not re-occur. I'll see if it repeats tomorrow.
> > > > >
> > > > > Regards,
> > > > > Ray L.
> > > > >
> > > > > --- In DynoMotion@yahoogroups.com <mailto:DynoMotion%40yahoogroups.com> <mailto:DynoMotion%40yahoogroups.com> , Tom Kerekes <tk@> wrote:
> > > > > >
> > > > > > Hi Ray
> > > > > > ÃÆ'Æ'‚
> > > > > > Here areÃÆ'Æ'‚ the sources changed from V4.29z:
> > > > > > ÃÆ'Æ'‚
> > > > > > http://dynomotion.com/Software/Patch/FixTP429Arcs/CoordMotion.cpp
> > > > > > ÃÆ'Æ'‚
> > > > > > http://dynomotion.com/Software/Patch/FixTP429Arcs/TrajectoryPlanner.cpp
> > > > > > ÃÆ'Æ'‚
> > > > > > Regards
> > > > > > TK
> > > > > >
> > > > > > From: himykabibble <jagboy@>
> > > > > > To: DynoMotion@yahoogroups.com <mailto:DynoMotion%40yahoogroups.com> <mailto:DynoMotion%40yahoogroups.com>
> > > > > > Sent: Saturday, February 25, 2012 2:35 PM
> > > > > > Subject: [DynoMotion] Re: v429z Problem
> > > > > >
> > > > > >
> > > > > > ÃÆ'Æ'‚
> > > > > > Tom,
> > > > > >
> > > > > > I was only able to run a quick test, but it appears to be working correctly now. I'll try to run some real parts tomorrow, once I get my machine back together (finally installing limit switches on X, Y and A axes).
> > > > > >
> > > > > > Can I get the modified source, as I need to make my tool length comp mods before I can run any of my existing G-code.
> > > > > >
> > > > > > Regards,
> > > > > > Ray L.
> > > > > >
> > > > > > --- In DynoMotion@yahoogroups.com <mailto:DynoMotion%40yahoogroups.com> <mailto:DynoMotion%40yahoogroups.com> , Tom Kerekes <tk@> wrote:
> > > > > > >
> > > > > > > Ray,
> > > > > > > ÃÆ'Æ'Æ'ÃÆ'¢â‚¬Å¡ÃÆ'Æ'‚
> > > > > > > Please try this patch to fix GCode Arcs for V4.29z.ÃÆ'Æ'Æ'ÃÆ'¢â‚¬Å¡ÃÆ'Æ'‚ Copy to the C:\Kmotion429z\Kmotion\Release directory
> > > > > > > ÃÆ'Æ'Æ'ÃÆ'¢â‚¬Å¡ÃÆ'Æ'‚
> > > > > > > http://dynomotion.com/Software/Patch/FixTP429Arcs/GCodeInterpreter.dll
> > > > > > > ÃÆ'Æ'Æ'ÃÆ'¢â‚¬Å¡ÃÆ'Æ'‚
> > > > > > > Regards
> > > > > > > TK
> > > > > > > ÃÆ'Æ'Æ'ÃÆ'¢â‚¬Å¡ÃÆ'Æ'‚
> > > > > > >
> > > > > > > From: himykabibble <jagboy@>
> > > > > > > To: DynoMotion@yahoogroups.com <mailto:DynoMotion%40yahoogroups.com> <mailto:DynoMotion%40yahoogroups.com>
> > > > > > > Sent: Thursday, February 23, 2012 2:25 PM
> > > > > > > Subject: [DynoMotion] Re: v429z Problem
> > > > > > >
> > > > > > >
> > > > > > > ÃÆ'Æ'Æ'ÃÆ'¢â‚¬Å¡ÃÆ'Æ'‚
> > > > > > > Tom,
> > > > > > >
> > > > > > > Well, how many people really NEED to do arcs anyway.... :-)
> > > > > > >
> > > > > > > Regards,
> > > > > > > Ray L.
> > > > > > >
> > > > > > > --- In DynoMotion@yahoogroups.com <mailto:DynoMotion%40yahoogroups.com> <mailto:DynoMotion%40yahoogroups.com> , Tom Kerekes <tk@> wrote:
> > > > > > > >
> > > > > > > > Hi Ray,
> > > > > > > > ÃÆ'Æ'Æ'ÃÆ'†'ÃÆ'Æ'¢ÃÆ'¢â€šÂ¬ÃÆ'…¡ÃÆ'Æ'Æ'ÃÆ'¢â‚¬Å¡ÃÆ'Æ'‚
> > > > > > > > I think I see what is going on.ÃÆ'Æ'Æ'ÃÆ'†'ÃÆ'Æ'¢ÃÆ'¢â€šÂ¬ÃÆ'…¡ÃÆ'Æ'Æ'ÃÆ'¢â‚¬Å¡ÃÆ'Æ'‚ I re-plotted your data and uploaded the Excel Sheet.ÃÆ'Æ'Æ'ÃÆ'†'ÃÆ'Æ'¢ÃÆ'¢â€šÂ¬ÃÆ'…¡ÃÆ'Æ'Æ'ÃÆ'¢â‚¬Å¡ÃÆ'Æ'‚ ÃÆ'Æ'Æ'ÃÆ'†'ÃÆ'Æ'¢ÃÆ'¢â€šÂ¬ÃÆ'…¡ÃÆ'Æ'Æ'ÃÆ'¢â‚¬Å¡ÃÆ'Æ'‚ Looks like there is a problem with Arcs getting out of order with Line segments due to the Grouping into blocks of segments that we did in V4.29 to improve USB traffic.ÃÆ'Æ'Æ'ÃÆ'†'ÃÆ'Æ'¢ÃÆ'¢â€šÂ¬ÃÆ'…¡ÃÆ'Æ'Æ'ÃÆ'¢â‚¬Å¡ÃÆ'Æ'‚ Everything works fine for Albert (and me) because we weren't doing any Arcs.ÃÆ'Æ'Æ'ÃÆ'†'ÃÆ'Æ'¢ÃÆ'¢â€šÂ¬ÃÆ'…¡ÃÆ'Æ'Æ'ÃÆ'¢â‚¬Å¡ÃÆ'Æ'‚ Give me some time to come up with a fix.
> > > > > > > > ÃÆ'Æ'Æ'ÃÆ'†'ÃÆ'Æ'¢ÃÆ'¢â€šÂ¬ÃÆ'…¡ÃÆ'Æ'Æ'ÃÆ'¢â‚¬Å¡ÃÆ'Æ'‚
> > > > > > > > Looks like it wasn't noise :}
> > > > > > > > ÃÆ'Æ'Æ'ÃÆ'†'ÃÆ'Æ'¢ÃÆ'¢â€šÂ¬ÃÆ'…¡ÃÆ'Æ'Æ'ÃÆ'¢â‚¬Å¡ÃÆ'Æ'‚
> > > > > > > > Regards
> > > > > > > > TK
> > > > > > > >
> > > > > > > > From: himykabibble <jagboy@>
> > > > > > > > To: DynoMotion@yahoogroups.com <mailto:DynoMotion%40yahoogroups.com> <mailto:DynoMotion%40yahoogroups.com>
> > > > > > > > Sent: Thursday, February 23, 2012 12:05 PM
> > > > > > > > Subject: [DynoMotion] Re: v429z Problem
> > > > > > > >
> > > > > > > >
> > > > > > > > ÃÆ'Æ'Æ'ÃÆ'†'ÃÆ'Æ'¢ÃÆ'¢â€šÂ¬ÃÆ'…¡ÃÆ'Æ'Æ'ÃÆ'¢â‚¬Å¡ÃÆ'Æ'‚
> > > > > > > > BTW - I can duplicate this with two different KFlops, and two different PCs....
> > > > > > > >
> > > > > > > > --- In DynoMotion@yahoogroups.com <mailto:DynoMotion%40yahoogroups.com> <mailto:DynoMotion%40yahoogroups.com> , "himykabibble" <jagboy@> wrote:
> > > > > > > > >
> > > > > > > > > OK, three spreadsheets uploaded to RayLs Files. The v428 trace looks the same, regardless of parameter setting. The v429 trace gets goofier the lower acceleration is set. Two captures, for accel = 30.0, and accel = 0.10 are given. All captures were recording Dest on every 50th pass of my Init program. First column is time, second thru fourth are X, Y, Z position.
> > > > > > > > >
> > > > > > > > > Regards,
> > > > > > > > > Ray L.
> > > > > > > > >
> > > > > > > > > --- In DynoMotion@yahoogroups.com <mailto:DynoMotion%40yahoogroups.com> <mailto:DynoMotion%40yahoogroups.com> , "himykabibble" <jagboy@> wrote:
> > > > > > > > > >
> > > > > > > > > > Tom,
> > > > > > > > > >
> > > > > > > > > > I'm posting some captures of ch[0-2].Dest for versions v428 and v429z. My machine was not even turned on for this. The problem is quite evident.
> > > > > > > > > >
> > > > > > > > > > Regards,
> > > > > > > > > > Ray L.
> > > > > > > > > >
> > > > > > > > > > --- In DynoMotion@yahoogroups.com <mailto:DynoMotion%40yahoogroups.com> <mailto:DynoMotion%40yahoogroups.com> , "himykabibble" <jagboy@> wrote:
> > > > > > > > > > >
> > > > > > > > > > > Tom,
> > > > > > > > > > >
> > > > > > > > > > > At this point, there is not a doubt in my mind that v429z is broken. I have run dozens of tests, and v428 always works, v429z always fails. With v429z, the lower I set acceleration, or feedrate, in KMotionCNC, the worse it performs. With accel set to 25 in/sec^2, I get small, but absolutely consistent, position errors, and NO servo faults. With accel set to 0.1 in/sec^2, I get nearly constant servo faults, with attendant massive position loss. The exact same g-code will work perfectly time after time with v428, with acceleration set as low or as high as I choose, and with feed rates cranked up to near rapid speeds. Even worse, it will sometimes move the Z axis, and there is not a single Z-axis command anywhere in the G-code! On the crooked, non-straight segments, I can put my fingers on the ballscrews, and I can feel clear, smooth motion on the axis that is supposed to be stationary. No chance this is noise - these are deliberate,
> > > constant speed
> > > > > > > > moves, always in exactly the same place, and they ARE reflected in the DROs.
> > > > > > > > > > >
> > > > > > > > > > > I've posted the G-code I'm using in RayLs Files.
> > > > > > > > > > >
> > > > > > > > > > > Regards,
> > > > > > > > > > > Ray L.
> > > > > > > > > > >
> > > > > > > > > >
> > > > > > > > >
> > > > > > > >
> > > > > > >
> > > > > >
> > > > >
> > > >
> > >
> > >
> > >
> > >
> > >
> > > {
> > > {
> > > }
> > >
> >
>