PMDX 107, PMDX 126, ESS, PWM signal blip

[yurtman]

Every now and then my spindle cuts out for a split second then resumes. This is only occurring once an hour but there is not pattern or regularity to it.  I noticed that the PWM led on the PMDX 107 board dims when the spindle cuts and so I believe that the PWM signal is the cause. I'm not sure where it originates, my ESS is mounted to and connected to the PMDX 126 with the short ribbon cable so I think it's unlikely to be electrical noise.  I did change the switch 5 of the 107 board which slows the respond to the PWM input, this did change the nature of the cutouts from being a quick slowdown and acceleration to a slower deceleration and speed up.  Can anyone help with this or suggest a way of diagnosing where the PWM signal may be slipping?

[Bob at PMDX]

We had someone report this kind of behavior once before (several years ago), with what I think is the same configuration you have (ESS/PMDX-126/PMDX-107).  I don't remember the solution, and a quick search of my emails hasn't turned it up yet.  I'll keep looking.

Meanwhile - I need a bit more information.  Are you using Mach3 or Mach4?

If you are using Mach3, please send us the XML file for your profile.  This will be in your main Mach3 installation directory (usually C:Mach3).  If your profile is called "MyMill", then the file will be C:Mach3MyMill.xml.

If you are using Mach4, please create a profile package and send that to me.  See the "BUT AN EVEN BETTER WAY" section in this FAQ for instructions on creating the profile package:
    http://www.pmdx.com/PMDX-Forums/index.php?topic=74.msg224#msg224

I also need to know the DIP switch settings on the PMDX-126 and the PMDX-107.

Bob

[yurtman]

Hi Bob,

Thanks for looking into this for me. I have attached pics of the 126 and 107 dip switch positions and my mach 3 xml

[thosj]

Did this ever get resolved? The thread just died on 8/12!

I'm pretty sure I'm the guy with the original issue a couple years ago! Never fixed it, it still randomly stops the VFD and starts back up. It's very random, nowhere near once per hour, but less often. Might be a week. It's always when it's running along just fine and the spindle slows. Once it stopped and never did restart before I hit eStop to stop the crunching cutter/part!! Most times it winds down, never completely stopping, then winds back up and continues along just fine assuming some tiny endmill didn't break. When this happens, the readout on the Hitachi VFD showing Hz, goes to zero, and when it starts back up, it goes back to whatever is/was correct.

Other than that, the 126 Rev A/107 combo works just fine on a BP clone belt head on Mach 3.

If you guys solved this, via email or something, post it up here!!

Try thosj1@gmail.com or thosj@new.rr.com if you're rooting around in email. This WAS before this forum existed, pretty sure

[Steve Stallings]

Sorry, but we have not reproduced the problem here and cannot figure it
out. Fingers want to point at some kind of strange interaction between
Mach3 and the SmoothStepper, but I have no evidence other than the
fact that the PWM light on the PMDX-107 is going dim which is caused
by the PWM coming into the PMDX-126 changing. I seem to remember
that you had a scope and could look at signals that way. If so you may
be able to put the scope on the pin 16 PWM signal present on J6 of the
PMDX-126 board and try to see if the duty cycle of the PWM signal really
has gone wrong when you hear the spindle slow down. I know that
catching it in the act will be tough, but that is the only thing that I can
think of.

[Bob at PMDX]

If so you may
be able to put the scope on the pin 17 PWM signal present on J6 of the
PMDX-126 board and try to see if the duty cycle of the PWM signal really
has gone wrong when you hear the spindle slow down.

That should be "Pin 16 PWM" signal :-)

Or Pin 17 if you want to look at the charge pump signal, which if *IT* had a temporary drop out could also cause that kind of behavior, though the charge pump dropping out would also cause a very brief drop out in the step pulses - which may or may not be noticeable by listening to the sound of the motors.

[thosj]

If I put a scope on Pin 17 and managed to catch it in the act and saw , what would this indicate, the ESS?

[Steve Stallings]

As Bob said, PWM is on pin 16 and Charge Pump is on pin 17. If either of
these drop out (assuming Charge Pump is enabled on the PMDX-107)
then the VFD would get signals that would cause it to lose speed.

The source of the problem could be either Mach or the SmoothStepper.
You might ask Andy at Warp9TD if he could repeat your experiment
using a scope to narrow down the source of the signal dropout.

[thosj]

Steve/Bob,

I've contacted Andy at Warp9 and we're 'working on it.'

Would it be possible for either of you to read this entire thread and see what YOU think?

http://warp9td.com/index.php/kunena/7-general-discussion/6442-mach3-ess-pwm-spindle-problem?limitstart=0

Thanks,

Tom

[Steve Stallings]

Hi Tom,

Wow, that is quite a thread.

I am hoping that your digital scope is dual trace or at least that you
have two probes and can trigger from one signal and capture the
other.

First, make sure that the analog ground that is connected from the
VFD to the PMDX-107 is isolated and not connected to the mains
voltage inside your VFD. Most modern VFDs do isolate this signal.
This is important because we want you to measure one signal
referenced to that ground and another referenced to the logic
ground of your PC. The two scope probe grounds will be connected
together inside the scope on almost all scopes.

If the grounds are OK, please set up one probe so that it triggers
when the analog control voltage drops. The probe input would
connect to the Aout of the PMDX-107 and the ground for this
probe would connect to the Agnd of the PMDX-107.

Set up the other probe to look at the Pin 16 signal on J6 of the
PMDX-126 and uses the PCgnd for reference. This is the same
signal that goes into the PMDX-107 to control speed. You will
want to set the scope for single event capture so that the traces
stay on the screen until you manually clear them and re-arm the
trigger.

Now try turning your spindle on and off. The off transition should
trigger the scope and you should see the analog voltage signal
drop on one trace of the scope. Because this event was intentional,
you should also see the PWM signal from Pin 16 disappear because
that is what stopped the spindle.

If you can get the scope showing things as described above, then
set your trap. When the analog signal drops when it should not,
then look at the captured PWM signal to see if it stopped or did
anything funny.

Regards,
Steve Stallings

[thosj]

Here are 4 captures of when the spindle slows and speeds back up and when it stops.

02_01 and 02_02 the spindle slowed and the scope triggered. Looks like the PWM is consistent thru there.

02_03 I don't know, looks like the 0-10v went UP then down, triggering on the down as it's set.

02_04 the spindle STOPPED and never restarted. On that one the PWM looks to have dropped along with the 0-10v.

So what have I got going on here? Any ideas?

[Steve Stallings]

The main discussion of this issue is happening over on the SmoothStepper forum and
you can track it here:

http://warp9td.com/index.php/kunena/7-general-discussion/6442-mach3-ess-pwm-spindle-problem?limitstart=0

Results will be summarized here later.

[thosj]

I'm going to dredge this instead of starting over.

I'm on Mach4 now and still having an issue.

Another guy over on Warp9 had this problem and changing PWM from 25Hz to 100Hz  in the ESS plugin config. solved his. I've tried 20 to 500 and it seems I have another problem now. My spindle stops/starts, like stop for a second, stopped dead, then starts. Sometimes it'll run for half an hour then start this stopping/starting. Sometimes it'll sag off and speed back up, sometimes it'll speed up and slow back down. Scope on Pin 16 shows PWM going while the spindle is stopping. VM on 0-10v out of 107 goes to zero when the spindle stops.

Something else bothers me, but perhaps it's nothing. I reached in the box and touched the back of the 107, side opposite the terminal strip, and the spindle speeds up!! Seems to take 3 fingers, one won't do it! Is this normal, probably is, but I'm wondering if I have some sort of connection problem somewhere that this does this!

I've revived the linked thread over on Warp9, so hopefully between you guys and Andy I can get this stopping/starting fixed. I actually started to like Mach4 and now it's almost unusable because of this spindle thing.

Thanks,

Tom Hesselman

[thosj]

Caught on tape.

https://www.dropbox.com/s/boy6iy15hss3aqe/2018-02-03%2015.48.44.mov?dl=0

Red is 0-10v from the PMDX-107

Yellow is PWM from PMDX-126 Pin 16.

I tried PWM 20Hz, 50Hz, 100Hz, and per Andy 300Hz, same thing.

It seems obvious to the unwashed that when the spindle stops the PWM marches on, the 0-10v drops to zero. Not as obvious is why?

I cross posted on Warp9 in hopes someone has a suggestion.

Tom

[thosj]

I don't see a way to edit my post so I'll add here.

Not being able to make any parts I fired up Mach3 on a different computer hooked to the machine and had the same stop/start issue with the spindle immediately. So it ain't the computer, it ain't Mach3/4 or bad profiles or any of that. With the PWM going I don't think it's the ESS. It HAS to be hardware or something between the PMDX-126/107 and the VFD, doesn't it? This is a big, big mystery to me!!

I'm going to set the PMDX-126 and 107 to not use the chargepump just for something to do in the meantime!!