Machsupport Forum
Mach Discussion => General Mach Discussion => Topic started by: Overloaded on February 21, 2014, 03:39:18 PM
-
Hey folks,
I have a new problem.
Been threading with this set-up for several years, thousands of identical parts.
I'll have to retrieve the version, but think its 22.
Suddenly, the rpm is bouncing around +/- 4 onn the Mach screen and about half that on the VFD readout.
This of course ruins a 48 pitch thread (5 passes).
It used to only fluctuate 1 rpm.... maybe 2, but mostly 1 and made acceptable threads. Sometimes 1600 pcs. non stop.
Quite by accident, I noticed that closing Mach, then restarting the rpm was stable again.
Then I checked the diagnostics, not sure though if it was fluctuating or not at the time. But, here is what it reported:
Time in int. 4 > 11
CPU speed 2018.0
Pulse Freq. 44023
The kernel is set a 25
This is PPort, PMDX 122 BoB - PMDX 106 Spindle control.
I'm trying to narrow it down to .... the PC ? the 106 ? ... or the inverter ?
We opened the pc case and blew it out (was very dirty) thought the cpu might be overheating .. didn't help.
Never know when it might happen, sometimes instantly ... might be good for a day or two. Regardless, much too frequent now.
Any help troubleshooting would be much appreciated,
Thanks,
Russ
-
One other little anomaly.
More frequently than before, Mach will hank up when clicking on certain buttons. One being Load Code or sometimes (iirc) Edit Code. But LOAD for sure.
Requires a manual kill and reboot. No other way.
Russ
-
Russ,
I would go with at least version 3.042.033. or .034.
+- 1 rpm is ok, but +-4 is not good as mach can't account for an increase in spindle speed ......adjustment for rpm slow down only if i remember correctly.
Maybe remove spindle control and set manualy ( which should be done anyway ) and see what the rpm readout is giving as that may give you a clue on what is the culprit.
RICH
-
Good points Rich !
I hadn't thought of bypassing the spindle control for a test.
That would certainly narrow it down.
Will check and try the newer versions.
Will be back at the machine Monday or tuesday. I'm making a list. Might take along a new VFD and a newer PC too.
Thanks a bunch Rich,
Russ
-
Time in int. 4 > 11
CPU speed 2018.0
Pulse Freq. 44023
The kernel is set a 25
Would look there first.
Hood
-
Thank you Hood.
What do you think might cause this to suddenly happen to a machine that has run very well and basically untouched for several years ?
And, what to do to possibly fix it ?
I don't have a scope, but plan to try to check the PWM out from the pc with an analog meter. Not expecting that to work very well though.
Also planning to check the analog IN to the drive (from the pc > 106) to see if it is wavering.
Will be back where the machine is Monday or Tuesday.
Thanks everyone,
Russ
-
If the pulse frequency is way out compared to the kernel and it has just recently happened, then I would say hardware could be the likely culprit.
Look at the mobo and see if any caps are bulging, look in the BIOS and see if the voltages and temps are close to where they should be. If you are having crashing problems with the computer (not sure if you are or not) then could be memory.
Hood
-
OK Hood, thanks.
I'll prepare another pc, just in case.
Russ
-
Russ,
I would go with at least version 3.042.033. or .034.
RICH
Hi Rich,
My lathe here is running 3.043.022, not sure which "22" is on the other machine yet.
Would you kindly verify that the 3.042. stated above is correct ? Just to be sure.
Thanks Rich,
Russ
-
Russ,
3.042 is corrrect.
RICH
-
Thanks Rich !
Roger wilco.
Russ
-
Russ, you could also try deleting the driver from device manager and then running the driver test to reinstall it, just in case the driver is somehow corrupt, my vote is still hardware though.
Hood
-
Added to my list, thanks Hood.
You mentioned memory. Does the memory fail ? Or go bad ?
Not sure how much is there, but been fine for a long time.
Thanks,
Russ
-
I dont think the memory would really have an effect on the pulse frequency being wildly out but it could cause other issues and possibly corruption.
I recently had to look at a computer for a friend and it was a bad stick of memory, it had been great for 6 years but one day it just started crashing and all sorts of errors, I had a look and thought it could be memory so hauled one stick out and it was the same so hauled the other out and it has been fine since, so one stick had gone bad.
Hood
-
OK, thanks, I'll look into that also.
This never really crashes, just hangs up to where the mouse wont move the cursor, device mgr. wont open, totally stuck (or does that qualify as a crash). Only one way out ... kill it.
The code always runs fine. Never a hitch, except for the rpm during threading.
The PWM out must be wavering, will check that the best I can.
There are few and only slightly varying S calls in the code, so I might be able to do away with the pwm output / 106 board all together and just use the spindle relay on the 122, setting the single desired speed in the VFD itself. If the rpm stability returns, I can go that route until I get a permanent fix.
Thanks again Hood,
regards,
Russ
:)