Hi Russ,
Sorry to hear you are having so much trouble.
Here are some thoughts:
#1 - you might verify that KFLOP is actually being Reset and it is not something else like a USB issue. If you turn off a KFLOP LED (Bit46) if KFLOP Resets it will turn it back on as an indication it rebooted.
#2 - a Mylar Cap may not be a good bypass capacitor. I think Mylar Caps are very accurate, stable, etc... but they may have too much inductance to reduce very high frequency noise (ie 10 MHz and above). I could be wrong but you might try a 0.01 uF Ceramic Cap instead to see if it helps.
#3 - You might try cutting the Reset wire at the JP4 connector to totally eliminate any possibility of noise coming in there
#4 - A cap on the Winford end of the cable is not likely to be of any value.
#5 - you might have grounding issues inducing a Reset. Are you powering KFLOP from an isolated supply? How is the MPG wired? Anything unusual?
HTH Regards TK
| Group: DynoMotion |
Message: 11371 |
From: Russ Larson |
Date: 4/13/2015 |
| Subject: Re: JP4 causes USB disconnects |
Tom, #1, Still need to run that experiment, but an interesting observation when turning the MPG and getting a USB disconnect the MPG still moves the mill. The displays do not update until you click the USB disconnect warning flag. #2, I replaced the Mylar cap with a 1500pF cap from AVX not sure if it was ceramic but it did improve things so at least the disconnect was not occurring every 20-30 seconds. It would sometimes go 2-4 minutes before a disconnect. #3, I did cut the forth wire in the ribbon so now I have less of an antenna to pick up noise. #4, I removed the Mylar cap at the Winford units since Pin #4 is no longer coming to that breakout board. #5, The supply powering KFLOP is a high quality Deltron multi-output supply. 5V@20A, 12V@5A, 15V@4A, each supply can be isolated from each other. I did tie the negatives together so they have the same reference. The 5V,12V is used for the KFLOP only, the 15V powers my optical homing sensors only. I also have an isolated 24V supply that powers the Panasonic amplifiers for my ac servo motors. MPG wiring is as follows: The A,B output of the MPG is converted into differential with small boards with driver IC. These differential outputs are fed into the Kanalog JP1 (pins 1-4). The switches on the MPG are all using the 3.3V available on JP4-3 and that is fed to through the various switches for X,Y,Z,A,X1,X10,X100 to the IO16 to IO IO22. I will order some ceramic caps from Digikey and get those installed to see if I can get this fixed. I also think I do have some kind of noise issue somewhere. I can disconnect the JP4 ribbon for the MPG and move the unit around with the arrow keys on the screen with no issues. I turn on the 7.5HP spindle and do the same thing with no issue so the special double shielded cable to the Spindle from the VFD appears to be working. The VFD is an Allen Bradley in a metal enclosure and the incoming power is also in shielded cable. The controls cable going to KFLOP is a double shielded cable carrying an analog voltage for speed and the output of two relays driven by the driver outputs of Kanalog. All of that stuff works perfect. The strange thing is when I put on a half inch carbide cutter and just used the jog keys to move the spindle cutting say 1/4" deep with 11K RPMs the wood cuts great but sudden the motion stops and I see a spinning wheel on Kmotion. I can never get it to respond after this happens and have to close Kmotion and start it again and it starts working again. I did not have much time this weekend to do additional experiments. Russ Larson From: DynoMotion@yahoogroups.com [mailto:DynoMotion@yahoogroups.com]
Sent: Friday, April 10, 2015 4:45 PM
To: DynoMotion@yahoogroups.com
Subject: Re: [DynoMotion] JP4 causes USB disconnects Sorry to hear you are having so much trouble. #1 - you might verify that KFLOP is actually being Reset and it is not something else like a USB issue. If you turn off a KFLOP LED (Bit46) if KFLOP Resets it will turn it back on as an indication it rebooted. #2 - a Mylar Cap may not be a good bypass capacitor. I think Mylar Caps are very accurate, stable, etc... but they may have too much inductance to reduce very high frequency noise (ie 10 MHz and above). I could be wrong but you might try a 0.01 uF Ceramic Cap instead to see if it helps. #3 - You might try cutting the Reset wire at the JP4 connector to totally eliminate any possibility of noise coming in there #4 - A cap on the Winford end of the cable is not likely to be of any value. #5 - you might have grounding issues inducing a Reset. Are you powering KFLOP from an isolated supply? How is the MPG wired? Anything unusual? I have KFLOP version 1.1, which does not have the caps on the reset lines for JP4 and JP6. I have a five inch ribbon cable connected to JP4 which goes to a Winford breakout board where my MPG is connected. With the machine sitting these doing absolutely nothing I get USB disconnects. I know this has to do with the JP4 connection and read many posts on this topic. I added a high quality mylar .1uF cap across Pin 4 (reset) to Pin 8 (ground) on the JP6 connector since it is not being used at all. I had these small press on pins so soldered two to the capacitor with short leads. This seemed to help slightly. I then also added another .1uF cap right at the Winford breakout board between pins 4 & 8. I still get these USB disconnects. The MPG works perfect until you get a disconnect. The machine moves fine all sensors work, etc. This last issue is driving me crazy. Any advise on how to get these resets to stop? |
|
| Group: DynoMotion |
Message: 11372 |
From: Tom Kerekes |
Date: 4/13/2015 |
| Subject: Re: JP4 causes USB disconnects |
Hi Russ,
Sorry you are having so much trouble. Things are not usually this sensitive.
Regarding:
#1 - That would indicate KFLOP is not being Reset. A more likely possibility would be a USB issue. How long is the USB cable and how is it routed? Please check the Shield continuity from one end to the other. It is supposed to be less than 1 ohm. Do you have it routed through other panel connectors? You might consider trying a USB isolator. See:
#5 - KFLOP doesn't use the +12V so you don't need this and might want to remove it.
If you tied the +15V GND and the +5V GND then you have defeated the opto isolation. I think you should disconnect that. I'm assuming the optical homing sensors are connected to Kanalog Opto inputs?
I'm not sure if the 24V supply needs to be made common GND with KFLOP/Kanalog or not. Is it? What signals are going between KFLOP/Kanalog and the Servo Amps?
Regarding cutting wood. I suppose that indicate that under load the VFD generates more noise and causes problems.
Regards TK
| Group: DynoMotion |
Message: 11375 |
From: Hardy Family |
Date: 4/13/2015 |
| Subject: Re: JP4 causes USB disconnects |
Don't forget that ferrite beads/toroids are also useful for noise suppression. There are clip-on units available for shielding ribbon or round cable. Also, putting 100 ohms resistance in series with the troublesome line can sometimes work miracles.
I have found that using a ceramic cap does not always do much good by itself, since these have very high Q which interacts with line inductance to cause ringing. Adding resistance (or a ferrite bead, which amounts to the same thing at high enough frequency) solves that problem.
Regards, SJH |
|
| Group: DynoMotion |
Message: 11376 |
From: Russ Larson |
Date: 4/13/2015 |
| Subject: Re: JP4 causes USB disconnects |
Tom, The USB cable is six feet long and is a shielded cable. I routes from the PC into the Metal Hoffman enclosure used as the control cabinet. The USB cable enters the cabinet alongside the low voltage control cable to the VFD. That cable handles the Direction switch signals and the analog voltage signal for speed. I will order a USB isolator as anything to make this system stable. The KFLOP is using power in the following manner. The 5V provides power to the KFLOP it is also used to power a small board I designed to take the Step & Direction outputs from KFLOP and converts them to Differential outputs to drive the Panasonic amplifiers for better noise immunity. The 12V is being used for the input on the Kanalog optical inputs for my homing sensors. The 15V is only being used to power the Omron Optical sensors that accept a metal flag to detect the axis has homed. I could have used 12V all the way to 24V with these sensors. The 24V is being used to provide the DC power to the Panasonic Amplifiers. This is also used to power the 24V relays that I used for switching the direction of the spindle with the VFD, as well as the 24V relay that enables all the amplifiers using the limit switches. If any limit switch is tripped the amplifiers are disabled. The use of 24VDC and a relay for this type enable is also pretty noise immune. I can try to remove the common ground between the supplies. The only thing that current ties the KFLOP/KANALOG to the Panasonic drivers is the step and direction outputs that get converted to differential signals prior to entering the amplifiers. Again, all control cable is has braided shield cable and routed carefully well away from any AC power. You are correct as the load on the spindle increases the noise generated must also be increasing. I can take a very shallow 1/16" cut and it behaves just like an air cut. When you take heavier cuts the current increases so noise must also increase causing me issues. Russ From: DynoMotion@yahoogroups.com [mailto:DynoMotion@yahoogroups.com]
Sent: Monday, April 13, 2015 11:58 AM
To: DynoMotion@yahoogroups.com
Subject: Re: [DynoMotion] JP4 causes USB disconnects Sorry you are having so much trouble. Things are not usually this sensitive.
#1 - That would indicate KFLOP is not being Reset. A more likely possibility would be a USB issue. How long is the USB cable and how is it routed? Please check the Shield continuity from one end to the other. It is supposed to be less than 1 ohm. Do you have it routed through other panel connectors? You might consider trying a USB isolator. See: #5 - KFLOP doesn't use the +12V so you don't need this and might want to remove it. If you tied the +15V GND and the +5V GND then you have defeated the opto isolation. I think you should disconnect that. I'm assuming the optical homing sensors are connected to Kanalog Opto inputs?
I'm not sure if the 24V supply needs to be made common GND with KFLOP/Kanalog or not. Is it? What signals are going between KFLOP/Kanalog and the Servo Amps?
Regarding cutting wood. I suppose that indicate that under load the VFD generates more noise and causes problems. #1, Still need to run that experiment, but an interesting observation when turning the MPG and getting a USB disconnect the MPG still moves the mill. The displays do not update until you click the USB disconnect warning flag. #2, I replaced the Mylar cap with a 1500pF cap from AVX not sure if it was ceramic but it did improve things so at least the disconnect was not occurring every 20-30 seconds. It would sometimes go 2-4 minutes before a disconnect. #3, I did cut the forth wire in the ribbon so now I have less of an antenna to pick up noise. #4, I removed the Mylar cap at the Winford units since Pin #4 is no longer coming to that breakout board. #5, The supply powering KFLOP is a high quality Deltron multi-output supply. 5V@20A, 12V@5A, 15V@4A, each supply can be isolated from each other. I did tie the negatives together so they have the same reference. The 5V,12V is used for the KFLOP only, the 15V powers my optical homing sensors only. I also have an isolated 24V supply that powers the Panasonic amplifiers for my ac servo motors. MPG wiring is as follows: The A,B output of the MPG is converted into differential with small boards with driver IC. These differential outputs are fed into the Kanalog JP1 (pins 1-4). The switches on the MPG are all using the 3.3V available on JP4-3 and that is fed to through the various switches for X,Y,Z,A,X1,X10,X100 to the IO16 to IO IO22. I will order some ceramic caps from Digikey and get those installed to see if I can get this fixed. I also think I do have some kind of noise issue somewhere. I can disconnect the JP4 ribbon for the MPG and move the unit around with the arrow keys on the screen with no issues. I turn on the 7.5HP spindle and do the same thing with no issue so the special double shielded cable to the Spindle from the VFD appears to be working. The VFD is an Allen Bradley in a metal enclosure and the incoming power is also in shielded cable. The controls cable going to KFLOP is a double shielded cable carrying an analog voltage for speed and the output of two relays driven by the driver outputs of Kanalog. All of that stuff works perfect. The strange thing is when I put on a half inch carbide cutter and just used the jog keys to move the spindle cutting say 1/4" deep with 11K RPMs the wood cuts great but sudden the motion stops and I see a spinning wheel on Kmotion. I can never get it to respond after this happens and have to close Kmotion and start it again and it starts working again. I did not have much time this weekend to do additional experiments. Sorry to hear you are having so much trouble. #1 - you might verify that KFLOP is actually being Reset and it is not something else like a USB issue. If you turn off a KFLOP LED (Bit46) if KFLOP Resets it will turn it back on as an indication it rebooted. #2 - a Mylar Cap may not be a good bypass capacitor. I think Mylar Caps are very accurate, stable, etc... but they may have too much inductance to reduce very high frequency noise (ie 10 MHz and above). I could be wrong but you might try a 0.01 uF Ceramic Cap instead to see if it helps. #3 - You might try cutting the Reset wire at the JP4 connector to totally eliminate any possibility of noise coming in there #4 - A cap on the Winford end of the cable is not likely to be of any value. #5 - you might have grounding issues inducing a Reset. Are you powering KFLOP from an isolated supply? How is the MPG wired? Anything unusual? I have KFLOP version 1.1, which does not have the caps on the reset lines for JP4 and JP6. I have a five inch ribbon cable connected to JP4 which goes to a Winford breakout board where my MPG is connected. With the machine sitting these doing absolutely nothing I get USB disconnects. I know this has to do with the JP4 connection and read many posts on this topic. I added a high quality mylar .1uF cap across Pin 4 (reset) to Pin 8 (ground) on the JP6 connector since it is not being used at all. I had these small press on pins so soldered two to the capacitor with short leads. This seemed to help slightly. I then also added another .1uF cap right at the Winford breakout board between pins 4 & 8. I still get these USB disconnects. The MPG works perfect until you get a disconnect. The machine moves fine all sensors work, etc. This last issue is driving me crazy. Any advise on how to get these resets to stop? |
|
| Group: DynoMotion |
Message: 11378 |
From: Russ Larson |
Date: 4/13/2015 |
| Subject: Re: JP4 causes USB disconnects |
Mr. Hardy, Good suggestion, I do have some ferrite Cores and will attempt to route the ribbon cable through the cores and see if things improve. That worked for my VFD control using RS485 on my small machine with a 4KW water cooled spindle. These constant disconnects really make the big machine totally unstable and to risky to attempt any jobs so far. Russ From: DynoMotion@yahoogroups.com [mailto:DynoMotion@yahoogroups.com]
Sent: Monday, April 13, 2015 1:39 PM
To: DynoMotion@yahoogroups.com
Subject: Re: [DynoMotion] JP4 causes USB disconnects Don't forget that ferrite beads/toroids are also useful for noise suppression. There are clip-on units available for shielding ribbon or round cable. Also, putting 100 ohms resistance in series with the troublesome line can sometimes work miracles. I have found that using a ceramic cap does not always do much good by itself, since these have very high Q which interacts with line inductance to cause ringing. Adding resistance (or a ferrite bead, which amounts to the same thing at high enough frequency) solves that problem. Regards, SJH On Mon, Apr 13, 2015 at 6:07 AM, 'Russ Larson' rdlarson@... [DynoMotion] <DynoMotion@yahoogroups.com> wrote: Tom, #1, Still need to run that experiment, but an interesting observation when turning the MPG and getting a USB disconnect the MPG still moves the mill. The displays do not update until you click the USB disconnect warning flag. #2, I replaced the Mylar cap with a 1500pF cap from AVX not sure if it was ceramic but it did improve things so at least the disconnect was not occurring every 20-30 seconds. It would sometimes go 2-4 minutes before a disconnect. #3, I did cut the forth wire in the ribbon so now I have less of an antenna to pick up noise. #4, I removed the Mylar cap at the Winford units since Pin #4 is no longer coming to that breakout board. #5, The supply powering KFLOP is a high quality Deltron multi-output supply. 5V@20A, 12V@5A, 15V@4A, each supply can be isolated from each other. I did tie the negatives together so they have the same reference. The 5V,12V is used for the KFLOP only, the 15V powers my optical homing sensors only. I also have an isolated 24V supply that powers the Panasonic amplifiers for my ac servo motors. MPG wiring is as follows: The A,B output of the MPG is converted into differential with small boards with driver IC. These differential outputs are fed into the Kanalog JP1 (pins 1-4). The switches on the MPG are all using the 3.3V available on JP4-3 and that is fed to through the various switches for X,Y,Z,A,X1,X10,X100 to the IO16 to IO IO22. I will order some ceramic caps from Digikey and get those installed to see if I can get this fixed. I also think I do have some kind of noise issue somewhere. I can disconnect the JP4 ribbon for the MPG and move the unit around with the arrow keys on the screen with no issues. I turn on the 7.5HP spindle and do the same thing with no issue so the special double shielded cable to the Spindle from the VFD appears to be working. The VFD is an Allen Bradley in a metal enclosure and the incoming power is also in shielded cable. The controls cable going to KFLOP is a double shielded cable carrying an analog voltage for speed and the output of two relays driven by the driver outputs of Kanalog. All of that stuff works perfect. The strange thing is when I put on a half inch carbide cutter and just used the jog keys to move the spindle cutting say 1/4" deep with 11K RPMs the wood cuts great but sudden the motion stops and I see a spinning wheel on Kmotion. I can never get it to respond after this happens and have to close Kmotion and start it again and it starts working again. I did not have much time this weekend to do additional experiments. Russ Larson Sorry to hear you are having so much trouble. #1 - you might verify that KFLOP is actually being Reset and it is not something else like a USB issue. If you turn off a KFLOP LED (Bit46) if KFLOP Resets it will turn it back on as an indication it rebooted. #2 - a Mylar Cap may not be a good bypass capacitor. I think Mylar Caps are very accurate, stable, etc... but they may have too much inductance to reduce very high frequency noise (ie 10 MHz and above). I could be wrong but you might try a 0.01 uF Ceramic Cap instead to see if it helps. #3 - You might try cutting the Reset wire at the JP4 connector to totally eliminate any possibility of noise coming in there #4 - A cap on the Winford end of the cable is not likely to be of any value. #5 - you might have grounding issues inducing a Reset. Are you powering KFLOP from an isolated supply? How is the MPG wired? Anything unusual? I have KFLOP version 1.1, which does not have the caps on the reset lines for JP4 and JP6. I have a five inch ribbon cable connected to JP4 which goes to a Winford breakout board where my MPG is connected. With the machine sitting these doing absolutely nothing I get USB disconnects. I know this has to do with the JP4 connection and read many posts on this topic. I added a high quality mylar .1uF cap across Pin 4 (reset) to Pin 8 (ground) on the JP6 connector since it is not being used at all. I had these small press on pins so soldered two to the capacitor with short leads. This seemed to help slightly. I then also added another .1uF cap right at the Winford breakout board between pins 4 & 8. I still get these USB disconnects. The MPG works perfect until you get a disconnect. The machine moves fine all sensors work, etc. This last issue is driving me crazy. Any advise on how to get these resets to stop? |
|
| Group: DynoMotion |
Message: 11389 |
From: Russ Larson |
Date: 4/15/2015 |
| Subject: Re: JP4 causes USB disconnects |
Tom, I implemented what Sid Hardy suggested in a previous post while waiting on my USB isolator to arrive, which should be this Friday. I put Ferrite Core donuts on each end of the USB cable looping the cable through two full turns. Then I also put a Ferrite Core on the ribbon cable on J4, looping it one time. The Ferrite Cores act as resistors to EMI noise. I now have no USB disconnects even under heavy cutting loads and can continue to experiment with the machine. Thanks Sid.. Russ Larson From: DynoMotion@yahoogroups.com [mailto:DynoMotion@yahoogroups.com]
Sent: Monday, April 13, 2015 11:58 AM
To: DynoMotion@yahoogroups.com
Subject: Re: [DynoMotion] JP4 causes USB disconnects Sorry you are having so much trouble. Things are not usually this sensitive.
#1 - That would indicate KFLOP is not being Reset. A more likely possibility would be a USB issue. How long is the USB cable and how is it routed? Please check the Shield continuity from one end to the other. It is supposed to be less than 1 ohm. Do you have it routed through other panel connectors? You might consider trying a USB isolator. See: #5 - KFLOP doesn't use the +12V so you don't need this and might want to remove it. If you tied the +15V GND and the +5V GND then you have defeated the opto isolation. I think you should disconnect that. I'm assuming the optical homing sensors are connected to Kanalog Opto inputs?
I'm not sure if the 24V supply needs to be made common GND with KFLOP/Kanalog or not. Is it? What signals are going between KFLOP/Kanalog and the Servo Amps?
Regarding cutting wood. I suppose that indicate that under load the VFD generates more noise and causes problems. #1, Still need to run that experiment, but an interesting observation when turning the MPG and getting a USB disconnect the MPG still moves the mill. The displays do not update until you click the USB disconnect warning flag. #2, I replaced the Mylar cap with a 1500pF cap from AVX not sure if it was ceramic but it did improve things so at least the disconnect was not occurring every 20-30 seconds. It would sometimes go 2-4 minutes before a disconnect. #3, I did cut the forth wire in the ribbon so now I have less of an antenna to pick up noise. #4, I removed the Mylar cap at the Winford units since Pin #4 is no longer coming to that breakout board. #5, The supply powering KFLOP is a high quality Deltron multi-output supply. 5V@20A, 12V@5A, 15V@4A, each supply can be isolated from each other. I did tie the negatives together so they have the same reference. The 5V,12V is used for the KFLOP only, the 15V powers my optical homing sensors only. I also have an isolated 24V supply that powers the Panasonic amplifiers for my ac servo motors. MPG wiring is as follows: The A,B output of the MPG is converted into differential with small boards with driver IC. These differential outputs are fed into the Kanalog JP1 (pins 1-4). The switches on the MPG are all using the 3.3V available on JP4-3 and that is fed to through the various switches for X,Y,Z,A,X1,X10,X100 to the IO16 to IO IO22. I will order some ceramic caps from Digikey and get those installed to see if I can get this fixed. I also think I do have some kind of noise issue somewhere. I can disconnect the JP4 ribbon for the MPG and move the unit around with the arrow keys on the screen with no issues. I turn on the 7.5HP spindle and do the same thing with no issue so the special double shielded cable to the Spindle from the VFD appears to be working. The VFD is an Allen Bradley in a metal enclosure and the incoming power is also in shielded cable. The controls cable going to KFLOP is a double shielded cable carrying an analog voltage for speed and the output of two relays driven by the driver outputs of Kanalog. All of that stuff works perfect. The strange thing is when I put on a half inch carbide cutter and just used the jog keys to move the spindle cutting say 1/4" deep with 11K RPMs the wood cuts great but sudden the motion stops and I see a spinning wheel on Kmotion. I can never get it to respond after this happens and have to close Kmotion and start it again and it starts working again. I did not have much time this weekend to do additional experiments. Sorry to hear you are having so much trouble. #1 - you might verify that KFLOP is actually being Reset and it is not something else like a USB issue. If you turn off a KFLOP LED (Bit46) if KFLOP Resets it will turn it back on as an indication it rebooted. #2 - a Mylar Cap may not be a good bypass capacitor. I think Mylar Caps are very accurate, stable, etc... but they may have too much inductance to reduce very high frequency noise (ie 10 MHz and above). I could be wrong but you might try a 0.01 uF Ceramic Cap instead to see if it helps. #3 - You might try cutting the Reset wire at the JP4 connector to totally eliminate any possibility of noise coming in there #4 - A cap on the Winford end of the cable is not likely to be of any value. #5 - you might have grounding issues inducing a Reset. Are you powering KFLOP from an isolated supply? How is the MPG wired? Anything unusual? I have KFLOP version 1.1, which does not have the caps on the reset lines for JP4 and JP6. I have a five inch ribbon cable connected to JP4 which goes to a Winford breakout board where my MPG is connected. With the machine sitting these doing absolutely nothing I get USB disconnects. I know this has to do with the JP4 connection and read many posts on this topic. I added a high quality mylar .1uF cap across Pin 4 (reset) to Pin 8 (ground) on the JP6 connector since it is not being used at all. I had these small press on pins so soldered two to the capacitor with short leads. This seemed to help slightly. I then also added another .1uF cap right at the Winford breakout board between pins 4 & 8. I still get these USB disconnects. The MPG works perfect until you get a disconnect. The machine moves fine all sensors work, etc. This last issue is driving me crazy. Any advise on how to get these resets to stop? |
|
| Group: DynoMotion |
Message: 11391 |
From: Tom Kerekes |
Date: 4/15/2015 |
| Subject: Re: JP4 causes USB disconnects |
Hi Russ,
Wow that is interesting. Thanks for posting back.
I'm curious did you ever check the USB shield continuity from one end of the USB cable to the other before adding the Ferrite cores? One common thing we have run into is that often USB cables will not have a good shield connection and show a high resistance. But then if one pulls, bends, or especially twists the cable near the connectors the shield will then have a good connection and low resistance.
Regards TK
| Group: DynoMotion |
Message: 11392 |
From: Russ Larson |
Date: 4/15/2015 |
| Subject: Re: JP4 causes USB disconnects |
Tom, I ran out and used my Fluke 87 multimeter and measured 1.2 ohms between the shields ends of the six foot USB cable. The cable I am using appears to be very high quality. It works on everything I have ever plugged in like printers, scanners, etc. with no issues. I had similar issues with my little cnc machine that uses a Chinese water cooled spindle which had lots of EMI noise. The VFD had a RS485 interface for modbus, but it would lose connection all the time. I added the Ferrite Cores to both ends of that RS485 cable and it allow the PC to talk to that VFD. This is strange as I took so much effort to ensure I used shielded cables and double shielded cables, and star ground inside the metal cabinet. I even have EMI filters for incoming AC power to the power supplies. The VFD is located about 10 feet away from the control box so, I am not sure where the EMI is running which drives these issues. The JP4 disconnects apparently had nothing to do with the capacitor on that connector. However, the disconnects were occurring without the VFD or spindle powered at all. Very strange stuff. The speed the signals coming through the JP4 inputs were nothing, this switch closures on the MPG. The actual A,B signals were going directly into the differential inputs of the Kanalog. Still scratching my head over these issues. Russ From: DynoMotion@yahoogroups.com [mailto:DynoMotion@yahoogroups.com]
Sent: Wednesday, April 15, 2015 2:16 PM
To: DynoMotion@yahoogroups.com
Subject: Re: [DynoMotion] JP4 causes USB disconnects Wow that is interesting. Thanks for posting back. I'm curious did you ever check the USB shield continuity from one end of the USB cable to the other before adding the Ferrite cores? One common thing we have run into is that often USB cables will not have a good shield connection and show a high resistance. But then if one pulls, bends, or especially twists the cable near the connectors the shield will then have a good connection and low resistance.
I implemented what Sid Hardy suggested in a previous post while waiting on my USB isolator to arrive, which should be this Friday. I put Ferrite Core donuts on each end of the USB cable looping the cable through two full turns. Then I also put a Ferrite Core on the ribbon cable on J4, looping it one time. The Ferrite Cores act as resistors to EMI noise. I now have no USB disconnects even under heavy cutting loads and can continue to experiment with the machine. Sorry you are having so much trouble. Things are not usually this sensitive. #1 - That would indicate KFLOP is not being Reset. A more likely possibility would be a USB issue. How long is the USB cable and how is it routed? Please check the Shield continuity from one end to the other. It is supposed to be less than 1 ohm. Do you have it routed through other panel connectors? You might consider trying a USB isolator. See: #5 - KFLOP doesn't use the +12V so you don't need this and might want to remove it. If you tied the +15V GND and the +5V GND then you have defeated the opto isolation. I think you should disconnect that. I'm assuming the optical homing sensors are connected to Kanalog Opto inputs? I'm not sure if the 24V supply needs to be made common GND with KFLOP/Kanalog or not. Is it? What signals are going between KFLOP/Kanalog and the Servo Amps? Regarding cutting wood. I suppose that indicate that under load the VFD generates more noise and causes problems. #1, Still need to run that experiment, but an interesting observation when turning the MPG and getting a USB disconnect the MPG still moves the mill. The displays do not update until you click the USB disconnect warning flag. #2, I replaced the Mylar cap with a 1500pF cap from AVX not sure if it was ceramic but it did improve things so at least the disconnect was not occurring every 20-30 seconds. It would sometimes go 2-4 minutes before a disconnect. #3, I did cut the forth wire in the ribbon so now I have less of an antenna to pick up noise. #4, I removed the Mylar cap at the Winford units since Pin #4 is no longer coming to that breakout board. #5, The supply powering KFLOP is a high quality Deltron multi-output supply. 5V@20A, 12V@5A, 15V@4A, each supply can be isolated from each other. I did tie the negatives together so they have the same reference. The 5V,12V is used for the KFLOP only, the 15V powers my optical homing sensors only. I also have an isolated 24V supply that powers the Panasonic amplifiers for my ac servo motors. MPG wiring is as follows: The A,B output of the MPG is converted into differential with small boards with driver IC. These differential outputs are fed into the Kanalog JP1 (pins 1-4). The switches on the MPG are all using the 3.3V available on JP4-3 and that is fed to through the various switches for X,Y,Z,A,X1,X10,X100 to the IO16 to IO IO22. I will order some ceramic caps from Digikey and get those installed to see if I can get this fixed. I also think I do have some kind of noise issue somewhere. I can disconnect the JP4 ribbon for the MPG and move the unit around with the arrow keys on the screen with no issues. I turn on the 7.5HP spindle and do the same thing with no issue so the special double shielded cable to the Spindle from the VFD appears to be working. The VFD is an Allen Bradley in a metal enclosure and the incoming power is also in shielded cable. The controls cable going to KFLOP is a double shielded cable carrying an analog voltage for speed and the output of two relays driven by the driver outputs of Kanalog. All of that stuff works perfect. The strange thing is when I put on a half inch carbide cutter and just used the jog keys to move the spindle cutting say 1/4" deep with 11K RPMs the wood cuts great but sudden the motion stops and I see a spinning wheel on Kmotion. I can never get it to respond after this happens and have to close Kmotion and start it again and it starts working again. I did not have much time this weekend to do additional experiments. Sorry to hear you are having so much trouble. #1 - you might verify that KFLOP is actually being Reset and it is not something else like a USB issue. If you turn off a KFLOP LED (Bit46) if KFLOP Resets it will turn it back on as an indication it rebooted. #2 - a Mylar Cap may not be a good bypass capacitor. I think Mylar Caps are very accurate, stable, etc... but they may have too much inductance to reduce very high frequency noise (ie 10 MHz and above). I could be wrong but you might try a 0.01 uF Ceramic Cap instead to see if it helps. #3 - You might try cutting the Reset wire at the JP4 connector to totally eliminate any possibility of noise coming in there #4 - A cap on the Winford end of the cable is not likely to be of any value. #5 - you might have grounding issues inducing a Reset. Are you powering KFLOP from an isolated supply? How is the MPG wired? Anything unusual? I have KFLOP version 1.1, which does not have the caps on the reset lines for JP4 and JP6. I have a five inch ribbon cable connected to JP4 which goes to a Winford breakout board where my MPG is connected. With the machine sitting these doing absolutely nothing I get USB disconnects. I know this has to do with the JP4 connection and read many posts on this topic. I added a high quality mylar .1uF cap across Pin 4 (reset) to Pin 8 (ground) on the JP6 connector since it is not being used at all. I had these small press on pins so soldered two to the capacitor with short leads. This seemed to help slightly. I then also added another .1uF cap right at the Winford breakout board between pins 4 & 8. I still get these USB disconnects. The MPG works perfect until you get a disconnect. The machine moves fine all sensors work, etc. This last issue is driving me crazy. Any advise on how to get these resets to stop? |
|
| | | | | |
