Group: DynoMotion	Message: 12785	From: andysontag	Date: 2/16/2016
Subject: Need help with a Grounding and/or Noise Issue

Attachments :

Hi All,

We've been experiencing a persistent issue that will periodically cause our KFLOP to "go out to lunch". Basically we'll get an error saying there was a read error and then Kmotion and Kmotion CNC will typically be unresponsive. Power cycling the Kflop clears everything up.

Till now we've just dealt with the issue as it only occurred occasionally and restarting the flop, re-homing, and starting where we left off wasn't annoying enough to fix the issue. Although our three year old Huanyang VFD (cheapest/crap VFD ever) finally died and we replaced it with a Hitachi WJ200-022SF. The VFD works great although our KFLOP will now go out to lunch much more often. I actually haven't been able to complete a program as it'll go out just before the real cutting starts.

I'm hoping someone might have some ideas that will make our setup more bulletproof. I've attached a photo that shows our electronics box. I have a few thoughts on what I would probably try to change first, although I figured I might save some time by asking for help rather than guessing and checking. I'm open to trying anything that'll get the job done.

Andy

@@attachment@@

Group: DynoMotion	Message: 12786	From: Hardy Family	Date: 2/16/2016
Subject: Re: Need help with a Grounding and/or Noise Issue [1 Attachment]

Is it the turning on of the spindle which crashes it?

I don't think the kflop shield will do you much good there, since it is not a complete enclosure and will thus re-radiate whatever hits it. Looks like you have tried to eliminate ground loops, and you have ferrite cores on most of the wires, so that's the most obvious causes accounted for.

If it was me, I would get a 'scope and probe around all the kflop inputs to see where the noise was. Also check that there were no brown-outs on the supply. If it's really the VFD, try moving it well away in its own metal box. The modbus should work over long distance.

Regards,

SJH

On Tue, Feb 16, 2016 at 4:44 PM, andysontag@... [DynoMotion] <DynoMotion@yahoogroups.com> wrote:

[Attachment(s) from andysontag@... [DynoMotion] included below]
Hi All,

We've been experiencing a persistent issue that will periodically cause our KFLOP to "go out to lunch". Basically we'll get an error saying there was a read error and then Kmotion and Kmotion CNC will typically be unresponsive. Power cycling the Kflop clears everything up.

Till now we've just dealt with the issue as it only occurred occasionally and restarting the flop, re-homing, and starting where we left off wasn't annoying enough to fix the issue. Although our three year old Huanyang VFD (cheapest/crap VFD ever) finally died and we replaced it with a Hitachi WJ200-022SF. The VFD works great although our KFLOP will now go out to lunch much more often. I actually haven't been able to complete a program as it'll go out just before the real cutting starts.

I'm hoping someone might have some ideas that will make our setup more bulletproof. I've attached a photo that shows our electronics box. I have a few thoughts on what I would probably try to change first, although I figured I might save some time by asking for help rather than guessing and checking. I'm open to trying anything that'll get the job done.

Andy

Group: DynoMotion	Message: 12787	From: Andy Sontag	Date: 2/16/2016
Subject: Re: Need help with a Grounding and/or Noise Issue

Thanks for the reply SJH. It doesn't crash when the spindle turns on although I have had it crash when adjusting the spindle speed manually on VFD panel. The most recent crashes happened after the spindle had come up to speed, and the machine was moving to the position where it was going to start cutting. Basically the bit was within a few mm of cutting into the material.

One thing I'm not sure I have done right is grounding with my 220v line. The common ground point is on the PSU ground although the 220v earth ground isn't connected to the common ground point. Should I be jumping the earth ground to the common ground point?

I'll try moving the VFD as far away as I can and test. That should be fairly straightforward.

I can try a different PSU for the KFLOP supply although we've had brown outs on a couple smaller supplies before and moved up to the 3A supply we have now. It has worked well as far as I can tell in the past so I would be surprised if the KFLOP was crashing due to insufficient current.

If moving the VFD doesn't do it, what will I be looking for as I scope/probe the KFLOP? I have a pretty decent scope although I don't have a lot of experience using it. Which inputs are you referring to? +5V? USB? Inputs from the SnapAmps? Am I just looking for non-square signals?

Andy

On Tue, Feb 16, 2016 at 6:09 PM, Hardy Family hardy.woodland.cypress@... [DynoMotion] <DynoMotion@yahoogroups.com> wrote:

Is it the turning on of the spindle which crashes it?

I don't think the kflop shield will do you much good there, since it is not a complete enclosure and will thus re-radiate whatever hits it. Looks like you have tried to eliminate ground loops, and you have ferrite cores on most of the wires, so that's the most obvious causes accounted for.

If it was me, I would get a 'scope and probe around all the kflop inputs to see where the noise was. Also check that there were no brown-outs on the supply. If it's really the VFD, try moving it well away in its own metal box. The modbus should work over long distance.

Regards,
SJH

On Tue, Feb 16, 2016 at 4:44 PM, andysontag@... [DynoMotion] <DynoMotion@yahoogroups.com> wrote:

[Attachment(s) from andysontag@... [DynoMotion] included below]
Hi All,

We've been experiencing a persistent issue that will periodically cause our KFLOP to "go out to lunch". Basically we'll get an error saying there was a read error and then Kmotion and Kmotion CNC will typically be unresponsive. Power cycling the Kflop clears everything up.

Till now we've just dealt with the issue as it only occurred occasionally and restarting the flop, re-homing, and starting where we left off wasn't annoying enough to fix the issue. Although our three year old Huanyang VFD (cheapest/crap VFD ever) finally died and we replaced it with a Hitachi WJ200-022SF. The VFD works great although our KFLOP will now go out to lunch much more often. I actually haven't been able to complete a program as it'll go out just before the real cutting starts.

I'm hoping someone might have some ideas that will make our setup more bulletproof. I've attached a photo that shows our electronics box. I have a few thoughts on what I would probably try to change first, although I figured I might save some time by asking for help rather than guessing and checking. I'm open to trying anything that'll get the job done.

Andy

Group: DynoMotion	Message: 12788	From: Hardy Family	Date: 2/16/2016
Subject: Re: Need help with a Grounding and/or Noise Issue

Earth grounding should not make a difference, but it is worth trying if only for the safety improvement.

I was a bit surprised that your two 48V supplies were only putting out 73V. That's a fair drop from 96V, unless they are variable and you have deliberately adjusted them down.

As far as the scope probing, you want to clip the scope ground to the kflop ground, as close as possible to the kflop. If your scope ground is not floating, then don't connect the earth ground mentioned above. Then try out all of the input signals you can get to. You will see square wave signals. If significantly loaded, they might be rounded, but not by too much (maybe a few hundred ns). Look for noise riding on top of the signal, and compare the noise levels when running the spindle and when idle. If there's more than a few hundred mV peak to peak noise, then check out that signal's wiring and see if you can clean it up.

Failing that, it is also possible for noise to couple through outputs, so they should be checked. They should be fairly clean since they are being driven by the kflop. If not, there might be some sort of conflict.

Don't forget to look at the USB +5V and kflop +3.3V power (pins on some of the kflop headers).

You are using a USB isolator, so it is probably not providing isolated power to the kflop USB, but check the kflop jumper setting for USB power to make sure there isn't a conflict (or maybe there is insufficient USB power - scope check should answer that).

Regards,

SJH

On Tue, Feb 16, 2016 at 6:51 PM, Andy Sontag andysontag@... [DynoMotion] <DynoMotion@yahoogroups.com> wrote:

Thanks for the reply SJH. It doesn't crash when the spindle turns on although I have had it crash when adjusting the spindle speed manually on VFD panel. The most recent crashes happened after the spindle had come up to speed, and the machine was moving to the position where it was going to start cutting. Basically the bit was within a few mm of cutting into the material.

One thing I'm not sure I have done right is grounding with my 220v line. The common ground point is on the PSU ground although the 220v earth ground isn't connected to the common ground point. Should I be jumping the earth ground to the common ground point?

I'll try moving the VFD as far away as I can and test. That should be fairly straightforward.

I can try a different PSU for the KFLOP supply although we've had brown outs on a couple smaller supplies before and moved up to the 3A supply we have now. It has worked well as far as I can tell in the past so I would be surprised if the KFLOP was crashing due to insufficient current.

If moving the VFD doesn't do it, what will I be looking for as I scope/probe the KFLOP? I have a pretty decent scope although I don't have a lot of experience using it. Which inputs are you referring to? +5V? USB? Inputs from the SnapAmps? Am I just looking for non-square signals?

Andy

On Tue, Feb 16, 2016 at 6:09 PM, Hardy Family hardy.woodland.cypress@... [DynoMotion] <DynoMotion@yahoogroups.com> wrote:

Is it the turning on of the spindle which crashes it?

I don't think the kflop shield will do you much good there, since it is not a complete enclosure and will thus re-radiate whatever hits it. Looks like you have tried to eliminate ground loops, and you have ferrite cores on most of the wires, so that's the most obvious causes accounted for.

If it was me, I would get a 'scope and probe around all the kflop inputs to see where the noise was. Also check that there were no brown-outs on the supply. If it's really the VFD, try moving it well away in its own metal box. The modbus should work over long distance.

Regards,
SJH

On Tue, Feb 16, 2016 at 4:44 PM, andysontag@... [DynoMotion] <DynoMotion@yahoogroups.com> wrote:

[Attachment(s) from andysontag@... [DynoMotion] included below]
Hi All,

We've been experiencing a persistent issue that will periodically cause our KFLOP to "go out to lunch". Basically we'll get an error saying there was a read error and then Kmotion and Kmotion CNC will typically be unresponsive. Power cycling the Kflop clears everything up.

Till now we've just dealt with the issue as it only occurred occasionally and restarting the flop, re-homing, and starting where we left off wasn't annoying enough to fix the issue. Although our three year old Huanyang VFD (cheapest/crap VFD ever) finally died and we replaced it with a Hitachi WJ200-022SF. The VFD works great although our KFLOP will now go out to lunch much more often. I actually haven't been able to complete a program as it'll go out just before the real cutting starts.

I'm hoping someone might have some ideas that will make our setup more bulletproof. I've attached a photo that shows our electronics box. I have a few thoughts on what I would probably try to change first, although I figured I might save some time by asking for help rather than guessing and checking. I'm open to trying anything that'll get the job done.

Andy

Group: DynoMotion	Message: 12789	From: Greg Carter	Date: 2/16/2016
Subject: Re: Need help with a Grounding and/or Noise Issue [1 Attachment]

Is the spindle motor cable shielded cable? You have marked "spindle motor cable shield", is that the shield or the ground. Either way the wire is way too long and it should terminate at the ground connection on the VFD and be no longer than you need. The shield (if you are using shielded motor cable) should terminate at both the VFD and the motor, to ground. How have you terminated the shield at the motor? Review section D2 in the WJ200 users manual.

One big no no is it looks like your motor cable is touching one of your encoder/limit switch cables, if not touching it looks to be running parallel to the low voltage cable, also bad. They should be separated by at least 12" and if they have to cross should cross at 90 degrees.

My suspicion is that the Hitachi VFD uses a higher switching frequency (less audible noise, more RFI noise) than the Chinese POS and is the reason you are having frequent problems. What is it set to? Parameter b083. Try lowering it.

Greg.

On 2/16/2016 7:44 PM, andysontag@... [DynoMotion] wrote:

Hi All,

We've been experiencing a persistent issue that will periodically cause our KFLOP to "go out to lunch". Basically we'll get an error saying there was a read error and then Kmotion and Kmotion CNC will typically be unresponsive. Power cycling the Kflop clears everything up.

Till now we've just dealt with the issue as it only occurred occasionally and restarting the flop, re-homing, and starting where we left off wasn't annoying enough to fix the issue. Although our three year old Huanyang VFD (cheapest/crap VFD ever) finally died and we replaced it with a Hitachi WJ200-022SF. The VFD works great although our KFLOP will now go out to lunch much more often. I actually haven't been able to complete a program as it'll go out just before the real cutting starts.

I'm hoping someone might have some ideas that will make our setup more bulletproof. I've attached a photo that shows our electronics box. I have a few thoughts on what I would probably try to change first, although I figured I might save some time by asking for help rather than guessing and checking. I'm open to trying anything that'll get the job done.

Andy

Group: DynoMotion	Message: 12790	From: Vlad O	Date: 2/16/2016
Subject: Re: Need help with a Grounding and/or Noise Issue

Andy,

1) Start with a single common ground point, not at one of your component, rather a common ground on chassis of the machine (true ground). It is a must ! Your wire from a PSU to the electrical feed would be a problem, and serve as a common parasitic resistor with has a voltage drop against ground.

2) Make an metal box for all electronic, to avoid a grounding and shielding differences. I use a SS box and carefully managed a single ground point at machine.

3) All cables select with shield, where shield grounded at one (source only) point. For instance, your step motor wire shield would grounded at random spot and that will cause a serious induced voltage on the connector spot area at your KFlop.

4) after you done all above, need to carefully measure with scope voltage amplitudes on all wire shielding against true ground mentioned above. There is no magic if you do it correctly. If you see some places voltage higher, need to find reason, and fix that to target any place voltage noise less then 5 milli volts (0.005V RMS) in all working/jogging/etc. regimes, where one or more steppers moving and stopping together or individually .... Your miles would vary, since you using VFD without filter and perhaps non-insulated or properly grounded motor.... It is a lot of learn there.

For instance, if you step motor are not insulated from the body of gantry or table or what ever, and your wire with long loop coming to your KFlop or other controller whole system will be depended on how fast you moving that motor - at slow speed and slow stop all may work, but as soon as you run some what higher rate, a more electrical noise on the "common ground" point would spoil all, even such robust controller as kFlop.

So, do your 5miles turn. !

Vlad

On Tue, Feb 16, 2016 at 9:09 PM, Hardy Family hardy.woodland.cypress@... [DynoMotion] <DynoMotion@yahoogroups.com> wrote:

Is it the turning on of the spindle which crashes it?

I don't think the kflop shield will do you much good there, since it is not a complete enclosure and will thus re-radiate whatever hits it. Looks like you have tried to eliminate ground loops, and you have ferrite cores on most of the wires, so that's the most obvious causes accounted for.

If it was me, I would get a 'scope and probe around all the kflop inputs to see where the noise was. Also check that there were no brown-outs on the supply. If it's really the VFD, try moving it well away in its own metal box. The modbus should work over long distance.

Regards,
SJH

On Tue, Feb 16, 2016 at 4:44 PM, andysontag@... [DynoMotion] <DynoMotion@yahoogroups.com> wrote:

[Attachment(s) from andysontag@... [DynoMotion] included below]
Hi All,

We've been experiencing a persistent issue that will periodically cause our KFLOP to "go out to lunch". Basically we'll get an error saying there was a read error and then Kmotion and Kmotion CNC will typically be unresponsive. Power cycling the Kflop clears everything up.

Till now we've just dealt with the issue as it only occurred occasionally and restarting the flop, re-homing, and starting where we left off wasn't annoying enough to fix the issue. Although our three year old Huanyang VFD (cheapest/crap VFD ever) finally died and we replaced it with a Hitachi WJ200-022SF. The VFD works great although our KFLOP will now go out to lunch much more often. I actually haven't been able to complete a program as it'll go out just before the real cutting starts.

I'm hoping someone might have some ideas that will make our setup more bulletproof. I've attached a photo that shows our electronics box. I have a few thoughts on what I would probably try to change first, although I figured I might save some time by asking for help rather than guessing and checking. I'm open to trying anything that'll get the job done.

Andy

Group: DynoMotion	Message: 12791	From: Greg Carter	Date: 2/17/2016
Subject: Re: Need help with a Grounding and/or Noise Issue

Took a second look at your picture. That really doesn't look like shielded cable you have going from the VFD to the motor, and it appears you have connected the ground wire (which you have labeled "Spindle motor cable shield") going to that piece of aluminum between the VFD and the KFlop, which has no connection to the ground on the VFD. That is not good. See my comments in previous email.

Ideally you should be using shielded cable from your VFD to your motor, like this:
http://www.ecdcontrols.com/electrical-components.1322/flexible-vfd-motor-cable.aspx

In addition to the Hitachi manual, this guide from Belden shows how to properly terminate the shielded cable:
https://www.belden.com/resourcecenter/tools/installguides/upload/VFD_Cable_Termination_Guide.pdf

And just to clarify my statement below about the higher carrier frequencies: The Hitachi using higher frequencies isn't the cause of your problems, it's the poor ground and shielding, the higher frequencies just exacerbate that.

Here is what I would try:
- connect the ground wire from the VFD/Motor wire to the ground on the VFD, also make sure it is grounded on your spindle motor.
- route the motor cable away from any other low voltage cables.
- try to run
- if failing - lower carrier frequency
- if still fails: purchase proper shielded motor cable, install input line filter on VFD, move VFD, use proper shielded cable for your limit switches and encoders.

On 2/16/2016 11:06 PM, Greg Carter greg@... [DynoMotion] wrote:

Is the spindle motor cable shielded cable? You have marked "spindle motor cable shield", is that the shield or the ground. Either way the wire is way too long and it should terminate at the ground connection on the VFD and be no longer than you need. The shield (if you are using shielded motor cable) should terminate at both the VFD and the motor, to ground. How have you terminated the shield at the motor? Review section D2 in the WJ200 users manual.

One big no no is it looks like your motor cable is touching one of your encoder/limit switch cables, if not touching it looks to be running parallel to the low voltage cable, also bad. They should be separated by at least 12" and if they have to cross should cross at 90 degrees.

My suspicion is that the Hitachi VFD uses a higher switching frequency (less audible noise, more RFI noise) than the Chinese POS and is the reason you are having frequent problems. What is it set to? Parameter b083. Try lowering it.

Greg.

On 2/16/2016 7:44 PM, andysontag@... [DynoMotion] wrote:

Hi All,

We've been experiencing a persistent issue that will periodically cause our KFLOP to "go out to lunch". Basically we'll get an error saying there was a read error and then Kmotion and Kmotion CNC will typically be unresponsive. Power cycling the Kflop clears everything up.

Till now we've just dealt with the issue as it only occurred occasionally and restarting the flop, re-homing, and starting where we left off wasn't annoying enough to fix the issue. Although our three year old Huanyang VFD (cheapest/crap VFD ever) finally died and we replaced it with a Hitachi WJ200-022SF. The VFD works great although our KFLOP will now go out to lunch much more often. I actually haven't been able to complete a program as it'll go out just before the real cutting starts.

I'm hoping someone might have some ideas that will make our setup more bulletproof. I've attached a photo that shows our electronics box. I have a few thoughts on what I would probably try to change first, although I figured I might save some time by asking for help rather than guessing and checking. I'm open to trying anything that'll get the job done.

Andy

Group: DynoMotion	Message: 12792	From: Russ Larson	Date: 2/17/2016
Subject: Re: Need help with a Grounding and/or Noise Issue

Andy,

Several people have made some excellent observations, that cable leaving the VFD to the Spindle does not appear to be a shielded cable and does not appeared to be grounded on the VFD end. The second big issue is to try and make all your low voltage wiring is kept clear of the spindle cable. You are getting EMI interference, no question. As Greg points out of you have to run the Spindle cable or any AC power cable near lower voltage cable cross it at a 90 degree angle to reduce the amount it can impact the low voltage signals. You are using a USB isolator which is great as that is another area that killed me in my setup. If you could put the electronics in a metal enclosure that would also help. Some people actually put the VFD in a separate metal enclose to help shield the noise. I too have the Hunyang VFD which spews noise like crazy, but I also have an expensive Allen Bradley unit that also caused me issues on the KFLOP until I got all the grounds, shields, and USB isolator in place.

Good Luck

Russ

From: DynoMotion@yahoogroups.com [mailto:DynoMotion@yahoogroups.com]
Sent: Tuesday, February 16, 2016 11:07 PM
To: DynoMotion@yahoogroups.com
Subject: Re: [DynoMotion] Need help with a Grounding and/or Noise Issue

Hi All,

We've been experiencing a persistent issue that will periodically cause our KFLOP to "go out to lunch". Basically we'll get an error saying there was a read error and then Kmotion and Kmotion CNC will typically be unresponsive. Power cycling the Kflop clears everything up.

Till now we've just dealt with the issue as it only occurred occasionally and restarting the flop, re-homing, and starting where we left off wasn't annoying enough to fix the issue. Although our three year old Huanyang VFD (cheapest/crap VFD ever) finally died and we replaced it with a Hitachi WJ200-022SF. The VFD works great although our KFLOP will now go out to lunch much more often. I actually haven't been able to complete a program as it'll go out just before the real cutting starts.

I'm hoping someone might have some ideas that will make our setup more bulletproof. I've attached a photo that shows our electronics box. I have a few thoughts on what I would probably try to change first, although I figured I might save some time by asking for help rather than guessing and checking. I'm open to trying anything that'll get the job done.

Andy

This email has been sent from a virus-free computer protected by Avast.
www.avast.com

Group: DynoMotion	Message: 12793	From: Tom Kerekes	Date: 2/17/2016
Subject: Re: Need help with a Grounding and/or Noise Issue

Hi Andy,

Another suggestion might be to use a small 10Watt switching power supply for KFLOP +5V and mount it a few inches from KFLOP rather than the wall wart plugged in far away with the DC wires running across the VFD. Maybe another for the encoder +5V?

Have you determined if the problem is caused by a USB communication fault? Or is KFLOP re-booting?

Regards
TK

On 2/17/2016 9:04 AM, 'Russ Larson' rdlarson@... [DynoMotion] wrote:

Andy,

Several people have made some excellent observations, that cable leaving the VFD to the Spindle does not appear to be a shielded cable and does not appeared to be grounded on the VFD end. The second big issue is to try and make all your low voltage wiring is kept clear of the spindle cable. You are getting EMI interference, no question. As Greg points out of you have to run the Spindle cable or any AC power cable near lower voltage cable cross it at a 90 degree angle to reduce the amount it can impact the low voltage signals. You are using a USB isolator which is great as that is another area that killed me in my setup. If you could put the electronics in a metal enclosure that would also help. Some people actually put the VFD in a separate metal enclose to help shield the noise. I too have the Hunyang VFD which spews noise like crazy, but I also have an expensive Allen Bradley unit that also caused me issues on the KFLOP until I got all the grounds, shields, and USB isolator in place.

Good Luck

Russ

From: DynoMotion@yahoogroups.com [mailto:DynoMotion@yahoogroups.com]
Sent: Tuesday, February 16, 2016 11:07 PM
To: DynoMotion@yahoogroups.com
Subject: Re: [DynoMotion] Need help with a Grounding and/or Noise Issue

Is the spindle motor cable shielded cable? You have marked "spindle motor cable shield", is that the shield or the ground. Either way the wire is way too long and it should terminate at the ground connection on the VFD and be no longer than you need. The shield (if you are using shielded motor cable) should terminate at both the VFD and the motor, to ground. How have you terminated the shield at the motor? Review section D2 in the WJ200 users manual.

One big no no is it looks like your motor cable is touching one of your encoder/limit switch cables, if not touching it looks to be running parallel to the low voltage cable, also bad. They should be separated by at least 12" and if they have to cross should cross at 90 degrees.

My suspicion is that the Hitachi VFD uses a higher switching frequency (less audible noise, more RFI noise) than the Chinese POS and is the reason you are having frequent problems. What is it set to? Parameter b083. Try lowering it.

Greg.

On 2/16/2016 7:44 PM, andysontag@... [DynoMotion] wrote:

Hi All,

We've been experiencing a persistent issue that will periodically cause our KFLOP to "go out to lunch". Basically we'll get an error saying there was a read error and then Kmotion and Kmotion CNC will typically be unresponsive. Power cycling the Kflop clears everything up.

Till now we've just dealt with the issue as it only occurred occasionally and restarting the flop, re-homing, and starting where we left off wasn't annoying enough to fix the issue. Although our three year old Huanyang VFD (cheapest/crap VFD ever) finally died and we replaced it with a Hitachi WJ200-022SF. The VFD works great although our KFLOP will now go out to lunch much more often. I actually haven't been able to complete a program as it'll go out just before the real cutting starts.

I'm hoping someone might have some ideas that will make our setup more bulletproof. I've attached a photo that shows our electronics box. I have a few thoughts on what I would probably try to change first, although I figured I might save some time by asking for help rather than guessing and checking. I'm open to trying anything that'll get the job done.

Andy

This email has been sent from a virus-free computer protected by Avast.
www.avast.com

Group: DynoMotion	Message: 12794	From: Andy Sontag	Date: 2/17/2016
Subject: Re: Need help with a Grounding and/or Noise Issue

Attachments :

Read Failed - Auto Disconnect-12794.JPG

Can't Connect-12794.JPG

Thank you for the great suggestions. I have started working on a few and will post the results as I work through each step.

The spindle cable is a shielded cable although it was homebuilt so it may need to be replaced with a proper cable such as the one Greg suggested. The ground/shield was connected to the Stepper PSU ground although I've now shortened the shield/ground wire and attached it to the VFD heatsink ground. This by itself didn't fix the issue(s).

The spindle cable is currently run through a cable track alongside all of the other stepper and sensor wires. It sounds like this is a big no-no. I'm curious how people avoid this. Every DIY machine I've seen on forums and youtube appears to run the spindle cable inside of cable tracks just as we have done. Any suggestions on how to get the spindle cable routed out of the way and not running parallel to all of the other wires?

Another question I have is with the common ground point. Vlad suggested grounding to a point on the chassis, which I assume means a point on the aluminum frame of the machine. So if I understand correctly the following should be done:

SnapAmp Grounds -> Chassis Ground

73V PSU (reduced from 96V) Ground -> Chassis Ground

Stepper Cable Shields -> Chassis Ground

Encoder and Limit Cable Shields -> Chassis Ground

Chassis Ground -> Earth Ground (VFD Heatsink Ground) ?

KFLOP 5V Ground -> Chassis Ground ?

I should note that our machine frame is made with 80/20 anodized aluminum extrusion, and so it doesn't necessarily have electrical conductivity with every metal component on the machine. If I use a point on the frame for a common ground location it might not be much different than a point on a nonconductive wood box for instance. Maybe someone could explain why a chassis ground is important so I can understand what needs to be done better. Like perhaps attaching short lengths of wire between each piece of the frame so that the entire frame is electrically connected.

Tom, is there a particular 5V PSU that you suggest? Would shortening the output wire on our switching 5V 3A wall wart PSU, and using an extension cord on the AC side accomplish what you had in mind?

I can't say for certain if the problem is caused by a USB communication fault or if the KFLOP is re-booting. However today when I was testing a couple of the suggestions our FLOP crashed again. I have attached two photos of the error pop-ups I received. During a prior attempt to improve things I added a resistor on the KFLOP although it didn't appear to have any effect. You should be able to see the resistor in electronics box photo.

Thanks again for all of the suggestions.

Andy

On Wed, Feb 17, 2016 at 10:25 AM, Tom Kerekes tk@... [DynoMotion] <DynoMotion@yahoogroups.com> wrote:

Hi Andy,

Another suggestion might be to use a small 10Watt switching power supply for KFLOP +5V and mount it a few inches from KFLOP rather than the wall wart plugged in far away with the DC wires running across the VFD. Maybe another for the encoder +5V?

Have you determined if the problem is caused by a USB communication fault? Or is KFLOP re-booting?

Regards
TK

On 2/17/2016 9:04 AM, 'Russ Larson' rdlarson@... [DynoMotion] wrote:

Andy,

Several people have made some excellent observations, that cable leaving the VFD to the Spindle does not appear to be a shielded cable and does not appeared to be grounded on the VFD end. The second big issue is to try and make all your low voltage wiring is kept clear of the spindle cable. You are getting EMI interference, no question. As Greg points out of you have to run the Spindle cable or any AC power cable near lower voltage cable cross it at a 90 degree angle to reduce the amount it can impact the low voltage signals. You are using a USB isolator which is great as that is another area that killed me in my setup. If you could put the electronics in a metal enclosure that would also help. Some people actually put the VFD in a separate metal enclose to help shield the noise. I too have the Hunyang VFD which spews noise like crazy, but I also have an expensive Allen Bradley unit that also caused me issues on the KFLOP until I got all the grounds, shields, and USB isolator in place.

Good Luck

Russ

From: DynoMotion@yahoogroups.com [mailto:DynoMotion@yahoogroups.com]
Sent: Tuesday, February 16, 2016 11:07 PM
To: DynoMotion@yahoogroups.com
Subject: Re: [DynoMotion] Need help with a Grounding and/or Noise Issue

Is the spindle motor cable shielded cable? You have marked "spindle motor cable shield", is that the shield or the ground. Either way the wire is way too long and it should terminate at the ground connection on the VFD and be no longer than you need. The shield (if you are using shielded motor cable) should terminate at both the VFD and the motor, to ground. How have you terminated the shield at the motor? Review section D2 in the WJ200 users manual.

One big no no is it looks like your motor cable is touching one of your encoder/limit switch cables, if not touching it looks to be running parallel to the low voltage cable, also bad. They should be separated by at least 12" and if they have to cross should cross at 90 degrees.

My suspicion is that the Hitachi VFD uses a higher switching frequency (less audible noise, more RFI noise) than the Chinese POS and is the reason you are having frequent problems. What is it set to? Parameter b083. Try lowering it.

Greg.

On 2/16/2016 7:44 PM, andysontag@... [DynoMotion] wrote:

Hi All,

We've been experiencing a persistent issue that will periodically cause our KFLOP to "go out to lunch". Basically we'll get an error saying there was a read error and then Kmotion and Kmotion CNC will typically be unresponsive. Power cycling the Kflop clears everything up.

Till now we've just dealt with the issue as it only occurred occasionally and restarting the flop, re-homing, and starting where we left off wasn't annoying enough to fix the issue. Although our three year old Huanyang VFD (cheapest/crap VFD ever) finally died and we replaced it with a Hitachi WJ200-022SF. The VFD works great although our KFLOP will now go out to lunch much more often. I actually haven't been able to complete a program as it'll go out just before the real cutting starts.

I'm hoping someone might have some ideas that will make our setup more bulletproof. I've attached a photo that shows our electronics box. I have a few thoughts on what I would probably try to change first, although I figured I might save some time by asking for help rather than guessing and checking. I'm open to trying anything that'll get the job done.

Andy

This email has been sent from a virus-free computer protected by Avast.
www.avast.com

@@attachment@@

Group: DynoMotion	Message: 12795	From: Tom Kerekes	Date: 2/17/2016
Subject: Re: Need help with a Grounding and/or Noise Issue [2 Attachments]

Attachments :

GND Loop 750-12795(1).png

Hi Andy,

I don't think you should have all those chassis ground connections. Do you understand the concept of a ground loop? Its basically providing a path for earth ground currents to flow through DC ground wiring. Its common to have grounds connected in ways that you didn't realize were connected. ie through a controller, frame, power supplies, shields, etc... So I like to remove all the connections between grounds and verify that there is in fact no connections that I'm not aware of by using an ohm meter. The resistance between all combinations of separate grounds should be all 100K Ohms or higher. Then if you decide to connect two different grounds you can do it one place of your choosing.

SnapAmp is not an isolated amplifier. Therefore SnapAmp DC GND, the 73V PSU DC GND, KFLOP GND, USB GND, and the USB Shield are all one single node. Having more than one connection from this node to Earth GND will create a ground loop. See attached diagram (although your situation should be different with a USB isolator).

Reducing the wall wart wiring length can't hurt, but I prefer something with known specifications. I've had good luck with these:
http://www.mouser.com/search/ProductDetail.aspx?R=0virtualkey0virtualkeyRS-15-5

To determine if KFLOP is rebooting (any loss of communication will result in those same errors) a trick I use is to turn off one of the KFLOP LEDs (Bit46). If KFLOP reboots it will default back to on.

I think you meant to say you added a capacitor between pins 4 and 8 on the connector? I can't quite tell if its on the right pins in the photo.

Regards
TK

On 2/17/2016 4:32 PM, Andy Sontag andysontag@... [DynoMotion] wrote:

Thank you for the great suggestions. I have started working on a few and will post the results as I work through each step.

The spindle cable is a shielded cable although it was homebuilt so it may need to be replaced with a proper cable such as the one Greg suggested. The ground/shield was connected to the Stepper PSU ground although I've now shortened the shield/ground wire and attached it to the VFD heatsink ground. This by itself didn't fix the issue(s).

The spindle cable is currently run through a cable track alongside all of the other stepper and sensor wires. It sounds like this is a big no-no. I'm curious how people avoid this. Every DIY machine I've seen on forums and youtube appears to run the spindle cable inside of cable tracks just as we have done. Any suggestions on how to get the spindle cable routed out of the way and not running parallel to all of the other wires?

Another question I have is with the common ground point. Vlad suggested grounding to a point on the chassis, which I assume means a point on the aluminum frame of the machine. So if I understand correctly the following should be done:

SnapAmp Grounds -> Chassis Ground

73V PSU (reduced from 96V) Ground -> Chassis Ground

Stepper Cable Shields -> Chassis Ground

Encoder and Limit Cable Shields -> Chassis Ground

Chassis Ground -> Earth Ground (VFD Heatsink Ground) ?

KFLOP 5V Ground -> Chassis Ground ?

I should note that our machine frame is made with 80/20 anodized aluminum extrusion, and so it doesn't necessarily have electrical conductivity with every metal component on the machine. If I use a point on the frame for a common ground location it might not be much different than a point on a nonconductive wood box for instance. Maybe someone could explain why a chassis ground is important so I can understand what needs to be done better. Like perhaps attaching short lengths of wire between each piece of the frame so that the entire frame is electrically connected.

Tom, is there a particular 5V PSU that you suggest? Would shortening the output wire on our switching 5V 3A wall wart PSU, and using an extension cord on the AC side accomplish what you had in mind?

I can't say for certain if the problem is caused by a USB communication fault or if the KFLOP is re-booting. However today when I was testing a couple of the suggestions our FLOP crashed again. I have attached two photos of the error pop-ups I received. During a prior attempt to improve things I added a resistor on the KFLOP although it didn't appear to have any effect. You should be able to see the resistor in electronics box photo.

Thanks again for all of the suggestions.

Andy

On Wed, Feb 17, 2016 at 10:25 AM, Tom Kerekes tk@... [DynoMotion] <DynoMotion@yahoogroups.com> wrote:

Hi Andy,

Another suggestion might be to use a small 10Watt switching power supply for KFLOP +5V and mount it a few inches from KFLOP rather than the wall wart plugged in far away with the DC wires running across the VFD. Maybe another for the encoder +5V?

Have you determined if the problem is caused by a USB communication fault? Or is KFLOP re-booting?

Regards
TK

On 2/17/2016 9:04 AM, 'Russ Larson' rdlarson@... [DynoMotion] wrote:

Andy,

Several people have made some excellent observations, that cable leaving the VFD to the Spindle does not appear to be a shielded cable and does not appeared to be grounded on the VFD end. The second big issue is to try and make all your low voltage wiring is kept clear of the spindle cable. You are getting EMI interference, no question. As Greg points out of you have to run the Spindle cable or any AC power cable near lower voltage cable cross it at a 90 degree angle to reduce the amount it can impact the low voltage signals. You are using a USB isolator which is great as that is another area that killed me in my setup. If you could put the electronics in a metal enclosure that would also help. Some people actually put the VFD in a separate metal enclose to help shield the noise. I too have the Hunyang VFD which spews noise like crazy, but I also have an expensive Allen Bradley unit that also caused me issues on the KFLOP until I got all the grounds, shields, and USB isolator in place.

Good Luck

Russ

From: DynoMotion@yahoogroups.com [mailto:DynoMotion@yahoogroups.com]
Sent: Tuesday, February 16, 2016 11:07 PM
To: DynoMotion@yahoogroups.com
Subject: Re: [DynoMotion] Need help with a Grounding and/or Noise Issue

Is the spindle motor cable shielded cable? You have marked "spindle motor cable shield", is that the shield or the ground. Either way the wire is way too long and it should terminate at the ground connection on the VFD and be no longer than you need. The shield (if you are using shielded motor cable) should terminate at both the VFD and the motor, to ground. How have you terminated the shield at the motor? Review section D2 in the WJ200 users manual.

One big no no is it looks like your motor cable is touching one of your encoder/limit switch cables, if not touching it looks to be running parallel to the low voltage cable, also bad. They should be separated by at least 12" and if they have to cross should cross at 90 degrees.

My suspicion is that the Hitachi VFD uses a higher switching frequency (less audible noise, more RFI noise) than the Chinese POS and is the reason you are having frequent problems. What is it set to? Parameter b083. Try lowering it.

Greg.

On 2/16/2016 7:44 PM, andysontag@... [DynoMotion] wrote:

Hi All,

We've been experiencing a persistent issue that will periodically cause our KFLOP to "go out to lunch". Basically we'll get an error saying there was a read error and then Kmotion and Kmotion CNC will typically be unresponsive. Power cycling the Kflop clears everything up.

Till now we've just dealt with the issue as it only occurred occasionally and restarting the flop, re-homing, and starting where we left off wasn't annoying enough to fix the issue. Although our three year old Huanyang VFD (cheapest/crap VFD ever) finally died and we replaced it with a Hitachi WJ200-022SF. The VFD works great although our KFLOP will now go out to lunch much more often. I actually haven't been able to complete a program as it'll go out just before the real cutting starts.

I'm hoping someone might have some ideas that will make our setup more bulletproof. I've attached a photo that shows our electronics box. I have a few thoughts on what I would probably try to change first, although I figured I might save some time by asking for help rather than guessing and checking. I'm open to trying anything that'll get the job done.

Andy

This email has been sent from a virus-free computer protected by Avast.
www.avast.com

@@attachment@@

Group: DynoMotion	Message: 12796	From: andysontag	Date: 2/17/2016
Subject: Re: Need help with a Grounding and/or Noise Issue

Tom,

Thanks for the PSU suggestion. I'll order one and see if it makes any difference.

Yes, I meant to say I added a capacitor. It is connected to pins 4 and 8.

I tried your trick of turning off an LED on the KFLOP and the LED stayed off after the "crash". This time I received a "G Code error GCode Aborted" popup from KMotion. I clicked ok on the popup although Kmotion and Kmotion CNC remained unresponsive. After forcing both programs to quit I opened them up again. Without power cycling KFLOP Kmotion says "disconnected". Power cycle KFLOP and it connects as usual and the LED turns on. At the moment it doesn't appear to reboot after the errors I'm experiencing.

I believe I understand the concept of ground loops, although I'll admit I'm feeling my way around in the dark when it comes to electrical noise, grounding, and how best to set things up. I'll try to keep the "KFLOP ground node" separate from possible earth grounds and test with the ohm meter. It sounds like all shields should be connected to the earth ground on the VFD. As long as the shields aren't somehow connected to the KFLOP ground node. Am I on the right track?

Andy

Group: DynoMotion	Message: 12797	From: Greg Carter	Date: 2/18/2016
Subject: Re: Need help with a Grounding and/or Noise Issue

Andy,

I just noticed that your KFlop-PC USB cable is touching the VFD-Motor cable. I would try rerouting it so that it doesn't come into contact with the VFD-Motor cable and is far away as you can get it.

Just out of curiosity, how did you make your "home built" shielded cable? And is the shield connected to ground on the spindle?

This is reaching a bit but figured I would mention it:
Is your spindle motor a no name Chinese variant? They have been known to NOT have the ground pin connected to the actual spindle ( http://www.cnczone.com/forums/spindles-vfd/271440-chinese-2-2kw-spindle-anyone-else-get.html ). While only having the shield connected at the VFD is still beneficial, standard practice now is to have it connected at both the VFD and the motor. Also you would want the spindle grounded for safety reasons.
Some more information on why it's good to connect the shield at both ends:
http://www.practicalmachinist.com/vb/robots-and-automation/vfd-power-cable-shield-grounding-question-299703-post2495602/#post2495602

Did you change anything else (wires, wire routes, etc) when you changed the VFD?

Might help to post pictures of your setup after you make changes.

On 2/17/2016 10:36 PM, andysontag@... [DynoMotion] wrote:

Tom,

Thanks for the PSU suggestion. I'll order one and see if it makes any difference.

Yes, I meant to say I added a capacitor. It is connected to pins 4 and 8.

I tried your trick of turning off an LED on the KFLOP and the LED stayed off after the "crash". This time I received a "G Code error GCode Aborted" popup from KMotion. I clicked ok on the popup although Kmotion and Kmotion CNC remained unresponsive. After forcing both programs to quit I opened them up again. Without power cycling KFLOP Kmotion says "disconnected". Power cycle KFLOP and it connects as usual and the LED turns on. At the moment it doesn't appear to reboot after the errors I'm experiencing.

I believe I understand the concept of ground loops, although I'll admit I'm feeling my way around in the dark when it comes to electrical noise, grounding, and how best to set things up. I'll try to keep the "KFLOP ground node" separate from possible earth grounds and test with the ohm meter. It sounds like all shields should be connected to the earth ground on the VFD. As long as the shields aren't somehow connected to the KFLOP ground node. Am I on the right track?

Andy

Group: DynoMotion	Message: 12799	From: andysontag	Date: 2/19/2016
Subject: Re: Need help with a Grounding and/or Noise Issue

Attachments :

Greg,

One of the many things I tried over the last couple days was rerouting the USB cable. Unfortunately the KFLOP still crashes. Tom, I've been keeping an eye on the LED and I still haven't seen a reboot.

The spindle cable is basically three main conductors of ~14 awg wire and one bare copper wire wrapped in aluminum foil. Heat shrink was then applied to the entire length of the cable (with manageable overlapping pieces). The bare copper wire and shield is connected to the spindle case instead of the connector pin which isn't connected to anything. I do indeed have a cheap water cooled 2.2 kw chinese spindle. It works great!

The VFD was the only thing I can remember changing. I have since made a bunch of changes to routing and connections, although except for a brief reprieve, the same behaviour of crashing is still occurring. I've attached a new photo that shows some of my "progress".

I tried using the scope although I felt like a third grader mucking around with a graduate student's thesis project. I didn't seem to find any smoking guns but I'm not sure I would have been able to identify them. I'd be happy to give it another shot, I just might need more detailed instructions/guidance.

I think the next step will be to physically move the VFD away from the KFLOP and see if anything changes.

Andy

Group: DynoMotion	Message: 12800	From: Russ Larson	Date: 2/19/2016
Subject: Re: Need help with a Grounding and/or Noise Issue

Attachments :

Andy,

It sounds like you are getting EMI interference coming in via the USB connection. The spindle cable should have a braided shield, not just a foil shield to help block EMI. See the photo below which is not spindle cable but low voltage connection cable but it has braided shield on top of the foil. You can find this stuff on ebay, in fact I have used some high end servo cables with 14 AWG for spindle cables, the link I provided from ebay is an example of an Allen Bradley Servo cable that uses 14 AWG and braided shield.

http://www.ebay.com/itm/Allen-Bradley-2090-CPWM7DF-14AA05-Servo-Drive-Cable-Assembly-Motor-Power-Cable-/262295673536?hash=item3d120a5ac0:g:Ha4AAOSwX~dWq5kV

The second recommendation is to make sure your USB cable is also shielded, again get the braided shield type as you see in the photo below. In addition to using a USB isolator I found that I needed to use a Ferrite Core to loop the USB cable through. Many USB cables come with one similar to the one plugged in the top position in the picture below, but I put the donut type on both ends of the cable. These filter block the EMI, I have posted some videos on cnczone of someone who explains all the details and shows the results with a scope, they actually work and are pretty cheap on ebay. Save yourself a lot of headaches. :)

Russ

From: DynoMotion@yahoogroups.com [mailto:DynoMotion@yahoogroups.com]
Sent: Friday, February 19, 2016 3:58 AM
To: DynoMotion@yahoogroups.com
Subject: Re: [DynoMotion] Need help with a Grounding and/or Noise Issue

Greg,

I think the next step will be to physically move the VFD away from the KFLOP and see if anything changes.

Andy

This email has been sent from a virus-free computer protected by Avast.
www.avast.com

Group: DynoMotion	Message: 12801	From: Vlad O	Date: 2/19/2016
Subject: Re: Need help with a Grounding and/or Noise Issue

Attachments :

Russ,

cable it is must !

Problems with distributed grounding - multiple grounding points ! Need to be fix first

Vlad

On Fri, Feb 19, 2016 at 8:45 AM, 'Russ Larson' rdlarson@... [DynoMotion] <DynoMotion@yahoogroups.com> wrote:

Andy,

It sounds like you are getting EMI interference coming in via the USB connection. The spindle cable should have a braided shield, not just a foil shield to help block EMI. See the photo below which is not spindle cable but low voltage connection cable but it has braided shield on top of the foil. You can find this stuff on ebay, in fact I have used some high end servo cables with 14 AWG for spindle cables, the link I provided from ebay is an example of an Allen Bradley Servo cable that uses 14 AWG and braided shield.

http://www.ebay.com/itm/Allen-Bradley-2090-CPWM7DF-14AA05-Servo-Drive-Cable-Assembly-Motor-Power-Cable-/262295673536?hash=item3d120a5ac0:g:Ha4AAOSwX~dWq5kV

The second recommendation is to make sure your USB cable is also shielded, again get the braided shield type as you see in the photo below. In addition to using a USB isolator I found that I needed to use a Ferrite Core to loop the USB cable through. Many USB cables come with one similar to the one plugged in the top position in the picture below, but I put the donut type on both ends of the cable. These filter block the EMI, I have posted some videos on cnczone of someone who explains all the details and shows the results with a scope, they actually work and are pretty cheap on ebay. Save yourself a lot of headaches. :)

Russ

From: DynoMotion@yahoogroups.com [mailto:DynoMotion@yahoogroups.com]
Sent: Friday, February 19, 2016 3:58 AM
To: DynoMotion@yahoogroups.com
Subject: Re: [DynoMotion] Need help with a Grounding and/or Noise Issue

Greg,

One of the many things I tried over the last couple days was rerouting the USB cable. Unfortunately the KFLOP still crashes. Tom, I've been keeping an eye on the LED and I still haven't seen a reboot.

The spindle cable is basically three main conductors of ~14 awg wire and one bare copper wire wrapped in aluminum foil. Heat shrink was then applied to the entire length of the cable (with manageable overlapping pieces). The bare copper wire and shield is connected to the spindle case instead of the connector pin which isn't connected to anything. I do indeed have a cheap water cooled 2.2 kw chinese spindle. It works great!

The VFD was the only thing I can remember changing. I have since made a bunch of changes to routing and connections, although except for a brief reprieve, the same behaviour of crashing is still occurring. I've attached a new photo that shows some of my "progress".

I tried using the scope although I felt like a third grader mucking around with a graduate student's thesis project. I didn't seem to find any smoking guns but I'm not sure I would have been able to identify them. I'd be happy to give it another shot, I just might need more detailed instructions/guidance.

I think the next step will be to physically move the VFD away from the KFLOP and see if anything changes.

Andy

This email has been sent from a virus-free computer protected by Avast.
www.avast.com

Group: DynoMotion	Message: 12802	From: andysontag	Date: 2/19/2016
Subject: Re: Need help with a Grounding and/or Noise Issue

Attachments :

Extra Ferrite on USB-12802.JPG

Russ,

I think you might have hit the nail on the head. I ordered new shielded VFD cable earlier today although I figured I'd try adding ferrite cores to the USB cable for kicks. I added two cores on the kflop side and one on the PC side. I've attached a photo that shows the extra ferrite. After the change, Bertha successfully completed a gcode program I've been using as a test. It's still too early to say if the ferrite actually solved my problem but it seems promising.

As an aside, I needed to disconnect the stepper motor shields from earth ground for some reason. Either the SnapAmp or the KFLOP would fall into what appeared to be a loop, trip a following error, and make a strange knocking sound on the X axes when the shields were connected.

I'll know more tomorrow but this is the most significant improvement so far. I may need to cancel the cable order...

Andy

@@attachment@@

Group: DynoMotion	Message: 12803	From: Russ Larson	Date: 2/20/2016
Subject: Re: Need help with a Grounding and/or Noise Issue [1 Attachment]

Andy,

Glad to help. The Ferrite cores really do a nice job of cutting down EMI. I would not cancel my spindle cable order, honestly having a cable with a nice braided shield is very important. The noise generated between the VFD and Spindle is pretty bad, and if you have totally shielded cables for both the stepper motors, Spindle Cable, and low voltage signal cables you can avoid future headaches. I have my VFD cable in the same cable tray as my limit switch, and other servo cables and since they are all shielded with braided shields and grounded probably they play nice together.

On you setup you said you have to disconnect the stepper motor shields. Those shields should only be connected on the stepper driver end, not on the motor end. This way they form what is called a ferrite cage. If you connect both ends to ground then you will get a ground loop which you have already discovered. You are headed down the right path and soon will have a machine that is realiable.

Russ

From: DynoMotion@yahoogroups.com [mailto:DynoMotion@yahoogroups.com]
Sent: Saturday, February 20, 2016 1:00 AM
To: DynoMotion@yahoogroups.com
Subject: RE: [DynoMotion] Need help with a Grounding and/or Noise Issue [1 Attachment]

[Attachment(s) from andysontag@... [DynoMotion] included below]

Russ,

I'll know more tomorrow but this is the most significant improvement so far. I may need to cancel the cable order...

Andy

This email has been sent from a virus-free computer protected by Avast.
www.avast.com

Group: DynoMotion	Message: 12811	From: Dan W	Date: 2/20/2016
Subject: Re: Need help with a Grounding and/or Noise Issue

Attachments :

20160218_203106-12811.jpg

What Russ just said about grounding one end is something I learned to do a short while ago and it seems to work well.

I started grounding one end after finding a paper on grounding Industrial controls. The paper made said to only connect the ground at the source end of the electrical signal.

I have done this with my KANALOG install. The analog signals to the servo drives are grounded at the KANALOG. The analog load signals coming out of the the servo drives are grounded at the drive. This has worked very well for me so far.

I have many more wires in what appears to be a smaller space than you installation and I have had no know problems with noise.

-Dan

Group: DynoMotion	Message: 12821	From: andysontag	Date: 2/22/2016
Subject: Re: Need help with a Grounding and/or Noise Issue

Dan,

I love your control cabinet. Very nicely organized and contained. Looks like you have a pretty impressive machine. VMC rebuild? It's really helpful to know that if done correctly you can have everything tightly packed together without issues.

I started with grounding the stepper shields at both ends a while back, although the more I read it appeared the right way to go was to disconnect the motor side. I might have a short somewhere along the stepper motor cables because it doesn't make sense that disconnecting the shields at the source side would improve things.

In any event it appears that adding a few cheap ferrite cores to the USB cable solved all of my issues. I used the machine all weekend without even the smallest hiccup. A first since we originally built everything up three years ago. (The machine was completely disassembled and relocated around 6 months ago.)

Using proper shielded VFD cable, shielded stepper cables and grounding everything except the VFD Shield and ground on the source side sounds like the right recipe. If I had done all of this to begin with I may not have needed to add ferrite to the USB cable.

A very sincere thank you to everyone for their help, especially Russ for providing the magic bullet.

Andy

Group: DynoMotion	Message: 12825	From: Vlad O	Date: 2/22/2016
Subject: Re: Need help with a Grounding and/or Noise Issue

Dan,

I would suggest to keep VFD close to motor, so cable inductive would be minimized.

VFD inside the control box, by their output wires, would be induce high frequency, high current spikes at the tool hit the material. I have see bursting, which increase ground loop noise.

Also, by splitting part of control functions to the front panel, increase amount of wiring (red/orange bunch). If you make a control box wider and mount some components vertically (not flat) to the bottom, you will save tons of wiring.

Otherwise all slick !

Andy, People telling you do shielding of control cables to VFD - it is only partially true. Also need to make a high current cables In/Out from/to VFD route far away from any control circuits if you expected high performance output from VFD.

If you going even to extreme, you need line In/Out filters with proper grounding to respect a single point of grounding.

Problems starts in power wiring and especially a distributed main motor grounding - so they need to be treated there, as a preliminary response on noise.

Even all starts to move gantry or table and you can cut low chip rate, as soon as you increase chip rate or drive speed or install a heavy blanks on the table - all could be appear in much disaster form, where is not missing steps but rather some fast advancement on directions or depth - could be catastrophic ! Been there, done that!

Be careful !

Vlad

On Mon, Feb 22, 2016 at 3:00 PM, andysontag@... [DynoMotion] <DynoMotion@yahoogroups.com> wrote:

Dan,

I love your control cabinet. Very nicely organized and contained. Looks like you have a pretty impressive machine. VMC rebuild? It's really helpful to know that if done correctly you can have everything tightly packed together without issues.

I started with grounding the stepper shields at both ends a while back, although the more I read it appeared the right way to go was to disconnect the motor side. I might have a short somewhere along the stepper motor cables because it doesn't make sense that disconnecting the shields at the source side would improve things.

In any event it appears that adding a few cheap ferrite cores to the USB cable solved all of my issues. I used the machine all weekend without even the smallest hiccup. A first since we originally built everything up three years ago. (The machine was completely disassembled and relocated around 6 months ago.)

Using proper shielded VFD cable, shielded stepper cables and grounding everything except the VFD Shield and ground on the source side sounds like the right recipe. If I had done all of this to begin with I may not have needed to add ferrite to the USB cable.

A very sincere thank you to everyone for their help, especially Russ for providing the magic bullet.

Andy