PMDX: DEVICE ERROR 'SYSTEM ERROR(18): Packet timeout, communications lost'

[Mark3796]

Hi there,
Me again. You replaced my PMDX 411 recently and it's been good as gold. I was running as a Mach4 3 axis router and although I haven't done any very big jobs, everything was running smoothly and as expected.
I recently added a fourth axis. I created a new profile on Mach4 using Mach4 mill. I copied most of the configuration settings across from the Mach4 Router profile and then played around getting the A axis revolving around X. After a bit of fine tuning I ran a few test cuts. everything seems OK except I get some weird crashes. Here is the history from the most recent few runs.
Any thoughts?

PMDX: Communication status 'Sync'd to Device'
Home switch Z Home tripped!
Home switch X Home tripped!
Home switch Y Home tripped!
Current tool == Selected tool so there is nothing to do
Current tool == Selected tool so there is nothing to do
PMDX: DEVICE ERROR 'SYSTEM ERROR(18): Packet timeout, communications lost'
PMDX: Communication error 'Lost Sync'
PMDX: Communication status 'Sync'd to Device'
Home switch Z Home tripped!
Home switch X Home tripped!
PMDX: Error 7 (Communication error) queuing HOME pkt for axis 'Y' motor 'Master' (id 1), aborting HOME cmd
Home switch Z Home tripped!
Home switch X Home tripped!
Home switch Y Home tripped!
Cycle Stopped
Home switch Z Home tripped!
Home switch X Home tripped!
Home switch Y Home tripped!
Current tool == Selected tool so there is nothing to do
Cycle Stopped
Home switch Z Home tripped!
Home switch X Home tripped!
Home switch Y Home tripped!
Current tool == Selected tool so there is nothing to do
Current tool == Selected tool so there is nothing to do
PMDX: Communication error 'Lost Sync'
PMDX: Communication status 'Sync'd to Device'
PMDX: DEVICE ERROR 'SYSTEM ERROR(18): Packet timeout, communications lost'

Cheers
Mark

[12strings]

This may be a "call/email tech support" kind of problem.  But lets see.

What are you using between the PMDX-411 and your motor drivers?  Some break-out board?  G540? PMDX-340?

When you added the 4th axis, did you add or change the motor power supply?

Do you get the error even if you never move the (new) 4th axis?

What if you try "air cutting" with your router un-powered?  Do you still get the "packet timeout" errors?

Check your machine wiring and try to keep al limit/home switch wiring as far from the motor/power wiring as possible.

[Mark3796]

Thanks 12strings, but no joy with any of that.

What are you using between the PMDX-411 and your motor drivers?  Some break-out board?  G540? PMDX-340?
G540

When you added the 4th axis, did you add or change the motor power supply?
No. I reconfigured my machine. When I first made this machine I had two motors on the y-axis; motor "A" was slave to "Y". I was getting annoyed with the gantry having to be set up for square every time I wanted to use it. I connected a shaft between both sides of "Y" so that both sides of the gantry are driven from one motor. So now I had an extra motor. This is what I used to create the new "A, rolls around X" configuration.

Do you get the error even if you never move the (new) 4th axis?
OK, so this was a really good question. I got to thinking all sorts of ways about the problem. I tried an old computer I had in the spare room. It's an XP machine. I uninstalled just about every program on it and set it up with the PMDX driver and Mach4. I was able to plug the PMDX straight into the XP computer without using a USB extension cable. Mach 4 runs fine. I unplugged the new "A" axis and tried the 3 axis router profile. Unfortunately, things got worse. Even just jogging the machine without powering up the router I get "PMDX: DEVICE ERROR 'SYSTEM ERROR(18): Packet timeout, communications lost' and "PMDX: Communication error 'Lost Sync'" and even the old "PMDX: Could not disable the PMDX-SmartBOB-USB, error 'Cannot communicate with device'".

What if you try "air cutting" with your router un-powered?  Do you still get the "packet timeout" errors?
Yes, see above.

Check your machine wiring and try to keep al limit/home switch wiring as far from the motor/power wiring as possible.
I don't think that is the problem. I haven't changed the wiring much at all and it was going really well for a while. The "Y" limit switches wiring path is well away from the stepper motor wiring.

[Mark3796]

Update.
The idea of using the old XP machine was to test out a couple of doubts that were nagging me.
1. Was the laptop not synching with the PMDX? That is, busy doing background tasks like checking wifi, antivirus, online accounts like OneDrive, etc. The XP machine was stripped right back with no antivirus, no internet connection, as few background processes as possible. But it was worse!
2.  Was there a problem with my USB connections? I connected the PMDX straight to the USB port on the XP machine, but again it was worse.
I removed the "A" axis entirely. Then I went back to using my laptop and the 3-axis router profile. I had a couple of errors--
I had Mach 4 open and PMDX plugged in, but I had the machine powered off. When I powered it up I got this:
E-Stop is in effect!  Clear the E-Stop condition.
E-Stop cleared!
PMDX: Communication error 'No Device Found'
PMDX: Could not disable the PMDX-SmartBOB-USB, error 'Cannot communicate with device'
PMDX: Communication status 'Sync'd to Device'
PMDX: DEVICE ERROR 'SYSTEM ERROR(18): Packet timeout, communications lost'

I restarted everything and had a good run. After a bit of time I got a homing error--
PMDX: Error 7 (Communication error) queuing HOME pkt for axis 'X' motor 'Master' (id 0), aborting HOME cmd
PMDX: Error 7 (Communication error) queuing HOME pkt for axis 'Y' motor 'Master' (id 1), aborting HOME cmd

After this I had quite a good run with no problems.

So now I'm beginning to think that there might be some setting that could be tweaked. The rotating "A" axis adds a level of complexity to the control. When running the simpler 3 axis profile with relatively simple G-code (repetition drilling), everything seems fine.

Any suggestions appreciated.

[12strings]

Most of the time when I've seen these kind of errors, it was electrical noise somehow interfering with the USB signals (either conducted or radiated).  This is kind-of born out by your experiment where the G540 (I presume) and rest of the system was powered down, Mach4 was running and talking to the 411.  Then you powered on the system and lost comms.

I think the "adding 4th axis" is a partial red herring here.  Since you were before using 4 motors, with 2 driving a single axis, you haven't really added a motor, just repurposed it.  That said, maaaaaaaaayyyyyyybe, with one motor now driving the gantry and the other driving the rotary table (holding it still even if it is not rotating), perhaps the total system load current is different enough to cause different noise.  Geeze that is really grasping at straws.

Is the laptop plugged into wall power? If so, try running the laptop from battery and see if it makes a difference.  If so, there may be a grounding issue.  Or, maybe adding a ferrite choke around the USB cable might help.  I had forgotten that PMDX sells those until just now.  Look for "noise filters" on the left hand menu.

[Rubes]

wanted to add my experience here for future searchers.
MINE WAS DUE TO A BAD POWER SUPPLY IN THE PC!!!!!!

Went thru three days of troubleshooting for this same thing. Except that my system was working fine last week, then all of a sudden started having this issue. Worse yet, I would get the error immediately when I hit the enable button (spindle still off of course). Well, maybe not "worse" since it wasn't ruining a huge job.
I turned off the power supply for the motors...same thing.
Disconnected all the connectors from the PMDX-416 to the stepper controllers...same.
Isolated all the grounds...same
Redid all the grounds in the control box to a star (which should have been done from the get go)...same
totally removed the PMDX-416 and plugged it into a different USB port with NOTHING else on...same (kind of indicated not random electrical noise??)

Tried my laptop, and all was good (yep...maybe it is a weird electrical noise thing...but why all of a sudden??)
hmmm...I have an exact duplicate Dell series 3000 machine in the "office"...lets try that...load up Mach4, PMDX driver...an voila...perfect!!
I swapped the power supplies from the office machine into the CNC controller and low and behold back in business!!!
And the office PC ain't complaining 'bout the "noisy" supply yet either...but I suspect its on its last legs.

Well, at least it forced me to blow out all the dust from inside the machines...sheesh what a mess.

[Rubes]

well...nevermind...was good for awhile, but its back to its ole self again.
As soon as I enable MACH4, the PMDX looses communication.
For some reason this particular PC just wont work with this SmartBOB anymore.
Another identical Dell Series 3000 machine seems to work OK...I give up. Time to move on.

[Mark3796]

OK, I admit I was sceptical.

I had learned about building a CNC machine by building an x-carve. I had no problems with noise, but I did use capacitors to filter noise on the limit switches following Ben Harper's suggestion here: https://www.instructables.com/id/End-Stop-Limit-Switch-Problems/
So when I built my new machine, I thought it was just a matter of scaling up: Bigger, stronger aluminium extrusions, proper linear guides, nice big stepper motors, beefy timing belts and a smarter interface than Arduino, hence the Gecko 540 and the PMDX 411.

Steve from PMDX said it was likely noise.

12strings, you were right! It's taken me a while, but I have completely rewired my machine and the problems are all solved.

Here's what I did:
1. replaced all stepper wiring with 4-core shielded microphone cable
2. replaced all limit switch wiring with 2-core shielded microphone cable
3. grounded all cable shielding to the machine frame in such a way that all the machine rails and gantry are connected to earth and all the cable shields are grounded at one end
4. inside the control box I moved all the AC wiring over to one side away from the power supply to the Gecko and any Gecko wiring
5. took particular care to keep the AC supply to the router separated from all the other wiring and laid it against the edge of the metal control box (kind of tucked into a recess). The control box is also connected to earth.
6. kept the AC to the router outside the cable ducts, allowing further separation.
7. where I couldn't use shielded cable, for example, the last little bit to the steppers, I used twisted pair wiring

In short, I did everything I could to reduce/isolate any potential sources of noise and to shield out as much noise as I could from the control side.
And it works a treat now!!!!

Thanks for all your help.