411 Issues and firmware weirdness

[atom]

Have been getting some really weird behavior the last few days from my 411. Seemed to run fine for the last few months, so not sure what has changed. Losing communications in the middle of large jobs with errors like "internal Device Error 110" and the normal "Cannot communicate with device".

Then this morning after trying to recover a really important job that it died on last night, I got some crazy firmware behavior. When launching Mach it prompted me to update the firmware. I didn't catch the version numbers it listed. After upgrading it, the machine was completely unusable. I would cycle start a job, the spindle would fire up for about 2 seconds, then Mach would display "Invalid or unknown PMDX-SmartBOB-USB device found". After launching Mach again, the firmware box came back up and said the current version was 0.48.165 and the available version was 0.45.155. This didn't make much sense to me having the version number go backwards.

After that second firmware update (backdate), I could run the machine again. Although 15 minutes into the job, it then lost communications again.

-Adam

[Bob at PMDX]

That *is* some bizarre behavior.  The "Invalid or unkown PMDX-SmartBOB-USB device found" should only be displayed when you try to enable Mach4 (by clicking on the "Enable" button).  As best I can tell it should never be generated while running a GCode file.  And with the two firmware versions begin reported, it almost sounds like you have 2 installations of Mach4 each with a different version of our plug-in.

FYI, when you get a "cannot communicate with device" error, the most reliable recovery process is to shut down Mach4, unplug the SmartBOB from the USB (or for the PMDX-424, unplug the AC/DC power supply.  Then reconnect the SmartBOB, wait 15 seconds and then restart Mach4.

So, here is what I would like  you to do.  Start Mach4 and enable our plug-in's debug log file as described in item (5) of the SmartBOB FAQ (http://www.pmdx.com/PMDX-Forums/index.php?topic=75.0).  Exit Mach4, then restart it.  Try running your GCode file and see if you can replicate any of these problems.  When you can replicate a problem, disable the plug-in's debug log file as per the FAQ, then create a "profile package" as per the FAQ and either post that here or email it to me at bob at this domain.

Bob

[atom]

Thanks for the reply Bob. I am trying to replicate it now with debugging turned on. Right after I posted, I air ran my file twice and it lost communications both times. One time was very early on in the job (10-15 min), and the second was about 1.5 hours in.

The current file I'm working with is very large at about 36MB and just under 2 million lines. I have seen this problem on smaller files as well, but just can't replicate as easily.

Another thing to note is this afternoon I went ahead and finished installing a new PC I have had waiting. I wanted to rule out that side if possible and this was a good excuse to go ahead and spend the time to get it loaded.

At a high level the problem screams noise, but the weird firmware behavior and randomness with no environment changes have me puzzled.

I will have some more info over with the debug files if this current run hoses up on me.

-Adam

[Bob at PMDX]

At a high level the problem screams noise, but the weird firmware behavior and randomness with no environment changes have me puzzled.

Definitely a possible cause.  And I realized that I glassed over the "start the spindle" phrase in your first post.  Usually that is a red flag for me.  What kind of spindle motor and control are you using?  How are you turning the spindle on and off (and controlling direction, if you do that)?  I see in your earlier posts that you are (or were) using a PMDX-126 and G540.  Is that still the case?

Bob

[atom]

Hey Bob-

Here is more more detailed information for you. I have also emailed you my debug and gcode file.

Current Setup:

PMDX-411 > PMDX-126 > G540

Spindle- PMDX-107. Hitachi X200 VFD and a no-brand 3HP spindle

Windows 7 x64 Enterprise
Mach Build 2914
SmartBob Plugin version- 0.41.205
SmartBob bootloader version- 1.8.64
SmartBob firmware version- 0.48.165

After getting the new PC installed last night, I started going through more runs to create the debug logs. I still got weird firmware behavior at first on the new PC. At first launch it asked me to update the firmware. After relaunching Mach4, I would get different errors, but could never get a job to start. A new one I got was "System Error(18): Packet timeout, communications lost".

At this point I went into the plugin config and told it to update the firmware. The update screen showed current and available firmware version as 0.48.165. I told it to download and did the normal relaunch of Mach. At this point I was back to my normal behavior of the PMDX-411 losing communications during the job.

I was still swearing it was noise from the VFD, so I did a run of the job with power completely disconnected from the VFD. At first it seemed to confirm the VFD noise because it went for almost two hours, but then lost communications around line 400,000. If it is of any value, this is the longest the machine has ran on the particular job before failing. With the VFD on the furthest it got was around line 370,000. This test run with the VFD off is what the debug files came from that I emailed over.

The true source of my issues is most likely that I am supposed to leave the state for a month in 5 days and am supposed to deliver the job I've been trying to complete :)

Thanks-

Adam

[Bob at PMDX]

To close out this thread, there appeared to be 2 issues (which Atom did most of the work figuring out):

(1) Interference from a combination of power line Ethernet adapters and wireless access points in use in the house.

(2) Electrical noise from the VFD

Bob