Hi Tapio,
Is this something that just started happening?
What Version of KMotion and Mach3 are you running?
Do you have Spindle On/Off commands? M3/M5? How are they configured?
Not aware of any other reports of this.
Regards TK
Group: DynoMotion |
Message: 9802 |
From: Tom Kerekes |
Date: 7/17/2014 |
Subject: Re: Mach dwell/KMotion feed hold? |
Hi Tapio,
That issue was regarding to the new "starvation protection" that automatically reduces the feed rate as the buffer starts to become empty. However at the end of a path the buffer really should run empty. So we turn this feature off by issuing a "FlushBuf" command to KFLOP. We forgot to do that in that one version for Mach3. That doesn't sound like your problem.
All patches are included in Version 4.33c.
Regards TK
Group: DynoMotion |
Message: 9808 |
From: Tapio Larikka |
Date: 7/18/2014 |
Subject: Re: Mach dwell/KMotion feed hold? |
Hi Tom,
This started rough two weeks ago. first
occasionally, then progressing to the present state where mach starts
dwell
soon as spindle is started either by screen button
or M3.
Today I noticed that this may have something to do
with Mach spindle delay settings. If delay is not zero mach dwells
forever.
With delay 0 mach starts spindle and start
executing but movement is app 0.01mm/sec instead of the commanded 0.15mm/rev at
500rpm.
I have had instance when after startup the movement
speed are not as high as they should be, but previously this was cured by simply
resaving motor tuning.
I also noticed that after the mach version
update/remove and reinstall of 042 I now have problems with commanded
movements.
Button jogs are fine, as crisp and stable as set
with KMotion step response, but G0 moves are not stable in speed.
This started with Mach 3.043.042/Kmotion431 and
remains with 043.066/433c and with 043.042/433c.
I think I will try a complete system reinstall,
starting with HDD formatting.
Can you think of any other possible cause than
somehow corrupted mach install or something malfunctioning on my
pc?
Rgds,
Tapio
----- Original Message -----
Sent: Friday, July 18, 2014 12:53
AM
Subject: Re: [DynoMotion] Mach
dwell/KMotion feed hold?
Hi Tapio,
Is
this something that just started happening?
What
Version of KMotion and Mach3 are you running?
Do
you have Spindle On/Off commands? M3/M5? How are they
configured?
Not
aware of any other reports of this.
Regards
TK
Group: DynoMotion |
Message: 9813 |
From: Tom Kerekes |
Date: 7/19/2014 |
Subject: Re: Mach dwell/KMotion feed hold? |
Hi Tapio,
There was a definite bug regarding Mach3 Dwells (G4 and Spindle On/off) in Version 4.33c again caused by the new buffer underrun protection. Here is a patch:
Please try it by copying it to the c:\KMotion433c\KMotion\Release directory. The Bug should hang consistently every time. I don't understand the intermittent or slow
motion behavior. That may be just not having consistent parameters in the Initialization C program?
Regards TK
From: "'Tapio Larikka' tapio.larikka@... [DynoMotion]" <DynoMotion@yahoogroups.com> To: DynoMotion@yahoogroups.com Sent: Friday, July 18, 2014 1:01 PM Subject: Re: [DynoMotion] Mach dwell/KMotion feed hold?
Hi Tom,
This started rough two weeks ago. first
occasionally, then progressing to the present state where mach starts
dwell
soon as spindle is started either by screen button
or M3.
Today I noticed that this may have something to do
with Mach spindle delay settings. If delay is not zero mach dwells
forever.
With delay 0 mach starts spindle and start
executing but movement is app 0.01mm/sec instead of the commanded 0.15mm/rev at
500rpm.
I have had instance when after startup the movement
speed are not as high as they should be, but previously this was cured by simply
resaving motor tuning.
I also noticed that after the mach version
update/remove and reinstall of 042 I now have problems with commanded
movements.
Button jogs are fine, as crisp and stable as set
with KMotion step response, but G0 moves are not stable in speed.
This started with Mach 3.043.042/Kmotion431 and
remains with 043.066/433c and with 043.042/433c.
I think I will try a complete system reinstall,
starting with HDD formatting.
Can you think of any other possible cause than
somehow corrupted mach install or something malfunctioning on my
pc?
Rgds,
Tapio
Group: DynoMotion |
Message: 9814 |
From: Tapio Larikka |
Date: 7/20/2014 |
Subject: Re: Mach dwell/KMotion feed hold? |
Hi Tom,
I will try the patch.
You wrote: "copy to ...Kmotion/release directory",
but the Dynomotion.dll from below link has file discription saying mach movement
plugin.
Should the patch go to Kmotion/release or to
Mach/Plugins? Please confirm
I'm not sure, but it can be that the movement speed
problem has something to do with Mach driver. First possible driver
corruption,
now the driver is not installed at all, because I
was under impression that with external motion engines it was not needed.
Then again Mach documentation says that installing
without driver sets mach in simulation mode.
I will reinstall mach with driver to see if that
makes any difference.
Rgds,
Tapio
Group: DynoMotion |
Message: 9815 |
From: TK |
Date: 7/20/2014 |
Subject: Re: Mach dwell/KMotion feed hold? |
Hi Tapio,
Oops. Yes copy to Mach3/Plugins directory.
Do not install the Parallel Port Driver. It is not needed or used when using KFLOP.
Regards TK
Hi Tom,
I will try the patch.
You wrote: "copy to ...Kmotion/release directory",
but the Dynomotion.dll from below link has file discription saying mach movement
plugin.
Should the patch go to Kmotion/release or to
Mach/Plugins? Please confirm
I'm not sure, but it can be that the movement speed
problem has something to do with Mach driver. First possible driver
corruption,
now the driver is not installed at all, because I
was under impression that with external motion engines it was not needed.
Then again Mach documentation says that installing
without driver sets mach in simulation mode.
I will reinstall mach with driver to see if that
makes any difference.
Rgds,
Tapio
Group: DynoMotion |
Message: 9830 |
From: Tapio Larikka |
Date: 7/22/2014 |
Subject: Re: Mach dwell/KMotion feed hold? |
Hi Tom,
I tried the new plugin but that did not
help.
-All axes jog as they are supposed
-X works OK with MDI commad
-Y from MDI has unstable/uneven speed, as if on mid
half of the movement the slide were too tight, causing excessive load.
deceleration to target
is much longer than set in parameters.
Also 2. consecutive MDI command does not produce movement, unless I hit Reset
first.
-Z I did not test
Good news is that I got the machine back to
production by
Complete HDD format and system install with Mach
3.043.066 and KMotion431 and
FixMach3ThreadingHangV431V432/DSPKFLOP.out
First I got Data starvation error, but increasing
buffer from 2 to 3 seconds started working, so I will stick with this setup for
a while now.
If I get any mor odd behavior i report. I soon
start the threading too so, if there comes any hangs will know that
too.
I don't know if this is related, but my other mill
with Mach 3.043.066 and KMotion431 and standard V431 DSPKFLOP.out
runs flawlessly 10 hours a day nonstop, with buffer
time set to 1.
Rgds,
Tapio
Group: DynoMotion |
Message: 9842 |
From: Tom Kerekes |
Date: 7/22/2014 |
Subject: Re: Mach dwell/KMotion feed hold? |
Hi Tapio,
So I understand you to say that with Version 4.33c some MDI command causes the Y axis to move in an irregular manner?
What MDI command were you
using?
Could you have an incorrect Trajectory Planner Setting for Y?
Regards TK
Group: DynoMotion |
Message: 9847 |
From: Tapio Larikka |
Date: 7/23/2014 |
Subject: Re: Mach dwell/KMotion feed hold? |
Hi Tom,
I'm running with Mach3 so there is no separate
trajectory planner settings?
I was using "G0 Y30" and "G0 Y-30" as MDI
command.
Now with Mach 3.043.066 and KMotion431 and
FixMach3ThreadingHangV431V432/DSPKFLOP.out
The above works as it shoud.
When I wrote about the other mill that runs happily all day I no longer
remembered 100% correctly.
Originally it also had a dwell/stop button related problem.
When Stop button was configured as Stop/Standard code 3 it caused the same
KMotion Feed hold/persistent Mach dwell/Mach starting dwell if
just rewinding the g-code and pressing cycle start.
After poking around for a while I noticed that closing and reloading G-code
removed the dwell, so finally
I got around this by changing the stop button to vb command with
DoOEMButton(1003)
Code"M30"
Now the Stop button also terminates any possible scripts that might be
running when stop is pressed.
For some reason closing and reloading g-code did not help when
KMotion433c was installed.
Would it be of any help/give any extra info to run the
"ShowThreadingValues2.c" if the mill that had this problem starts acting up
again?
Out of curiosity: which Kmotion release introduced the data starvation
protection?
Rgds,
Tapio
----- Original Message -----
Sent: Wednesday, July 23, 2014 9:59
AM
Subject: Re: [DynoMotion] Mach
dwell/KMotion feed hold?
Hi Tapio,
So
I understand you to say that with Version 4.33c some MDI command causes the Y
axis to move in an irregular manner?
What
MDI command were you using?
Could
you have an incorrect Trajectory Planner Setting for Y?
Regards
TK
Group: DynoMotion |
Message: 9872 |
From: Tapio Larikka |
Date: 7/28/2014 |
Subject: Re: Mach dwell/KMotion feed hold? |
Hi Tom,
I got some progress with this dwell/feed hold
problem of mine where movement stops and spindle keeps
on endlessly.
It appears that the reason is hitting the soft
limit :-(
Since I don't use them I never given them any
thought and just left them at default 1e009. With my 14400ppr spindle encoder this makes ~69400revs/87minutes at
800rpm which is pretty accurately what has been going on.
When I hit thi limit today I changed the limit to
1e019 and the machine continued to work again. After couple of runs I changed
thelimit back to 1e009 and
got an immediate limit error.
For some reason, probably because the spindle
channel is not included in coordsystem, I dont see any message of the limit
error. this may also be the reason
that the spindle is left running.
At the beginning of this thread you asked if this
just started to happen and now I think it started when I upgraded to Kmotion431
while we were hunting the lathe threading bug. I just did not notice this being
so busy with the lathe side. I checked today and my lathe init.c does not have
softlimits defined.
Second problem at the beginning of this thread I
mentioned was that the feedrates change from cycle to cycle. On Friday I
clocked ~20 consecutive workcycles and got times from 50sec to over
180secs. After weekend of wondering why my other machine runs steadily I
realized that the steady machine runs
open loop/contactor driven spindle. So I changed
from G95 to G94 and started getting cycle times repeatedly +/-
1sec.
For some reason feed/rpm is not consistent, but
why?
The above is with Mach
3.043.066 and KMotion431standard. Same bahavior was with
FixMach3ThreadingHangV431V432/DSPKFLOP.out
Rgds,
Tapio
----- Original Message -----
Sent: Wednesday, July 23, 2014 9:59
AM
Subject: Re: [DynoMotion] Mach
dwell/KMotion feed hold?
Hi Tapio,
So
I understand you to say that with Version 4.33c some MDI command causes the Y
axis to move in an irregular manner?
What
MDI command were you using?
Could
you have an incorrect Trajectory Planner Setting for Y?
Regards
TK
Group: DynoMotion |
Message: 9882 |
From: Tom Kerekes |
Date: 7/30/2014 |
Subject: Re: Mach dwell/KMotion feed hold? |
Hi Tapio,
The Soft Limits should not be used for the spindle and they default to +1e30 and -1e30. Soft Limits generate a Feed Hold and do not generate a console message. So I think your explaination is correct.
Regarding variable feed rates: With Threading Operations like G32/G76 the feed rate should vary proportional to the spindle speed, but I wouldn't expect to have such a variation in spindle speed. The actual Spindle Speed is displayed on the Mach3 Screen. How much variation do you have in that value?
Regards TK
Group: DynoMotion |
Message: 9884 |
From: Tapio Larikka |
Date: 7/30/2014 |
Subject: Re: Mach dwell/KMotion feed hold? |
Hi Tom,
I'll go and set the soft limits manually to 1e30
just to be sure.
The feed rates:
The problem I have now is with milling feed rates.
I normally use G95 feed/rev mode on all my machining. Spindle speed is
constant and at the
commanded rate according to external tachometer. The TrueRpm DRO might show an occasional change of
1Rpm.
Problem is that the same part program with F0.06
mm/rev at 800rpm the work cycle time varies from 50 to 180secs or so on
consecutive executions.
When I change to G94 feed/min mode F48 at 800rpm
makes consistent cycle times of 53 sec from cycle to cycle.
I know this well worked before I updated to Mach 3.043.066 in an attempt to get rid of the odd dwell/feed
hold, which as I have now convinced myself was
caused by my misunderstanding of the
softlimit.
I rolled mach back to 3.042.22 that I had before I
started poking it.
By Friday I should have some more statistic if
I have G95 feed/rev mode consistent again. I'll post a
follow-up.
Rgds,
Tapio
PS: In about two weeks I'll have to start the
thread cutting again, and any I'll findings, good or bad, to the threading hang
discussion.
----- Original Message -----
Sent: Wednesday, July 30, 2014 5:31
PM
Subject: Re: [DynoMotion] Mach
dwell/KMotion feed hold?
Hi Tapio,
The Soft Limits
should not be used for the spindle and they default to +1e30 and -1e30.
Soft Limits generate a Feed Hold and do not generate a console message.
So I think your explaination is correct.
Regarding variable feed
rates: With Threading Operations like G32/G76 the feed rate should vary
proportional to the spindle speed, but I wouldn't expect to have such a
variation in spindle speed. The actual Spindle Speed is displayed on the
Mach3 Screen. How much variation do you have in that
value?
Regards TK
| | | | | | | | | | | |
| | | | | | | | | | | |