Spindle and Stepper Motor Oscillation. Is it a Problem

[pzzamakr1980]

I have two Tormachs, both with the upgraded spindle drives. When running, regardless of cut depth, rpm, spindle load, any load, etc., the spindle will sound like its revving up and down a bit. This isn't a concern when I'm cutting air, but its more of a concern when I'm taking a heavy cut. I cannot actually tell if the spindle is oscillating (ie no abnormal finish, etc), but it sounds like it.

I also have the same issue with the steppers on an infrequent basis, it sounds like they also oscillate. I don't remember steppers doing this on my smaller personal mill.

Does anyone have any of the same issues or comments. Thx

[MarkWink]

Do a quick check of your power.

[pzzamakr1980]

Im getting a good 208 volts, stepped down from a 3 phase 208v system. There are no apparent issues.

[justgary]

Pzzamakr1980 -

What about the display on the VFD? If your spindle is changing speed, the display will show it. I have not peeked at mine in a while, but I think it tells you the frequency of the drive voltage. Drive frequency * 60 / Ratio = RPM. On the high speed pulley, use 2 for the ratio.

The fact that you have two machines that do the same thing would indicate some sort of power problem. Are both machines fed from the same phase of your mains? Is it possible to switch to another phase?

Regards,

- Just Gary

[pete from TN]

Buying one of these optical tachometers like I got from the internet and checking to see what the speed really is. Mine is what I used to get the actual spindle rpm values for my vector drive and it was only about $30.00... works great and is useful for other things as well....peace

[levelzero]

Did you ever figure out what it was?

[Don Clement]

I have a GenRad Strobotach that will show changes in spindle speed from constant.

[PinMachTool]

I've noticed the same thing with my Tormach. I to have the upgraded spindle drive. Just tonight I did notice the spindle come to an almost complete stop, then rev back up. So I
opened the cabinet to watch the frequency on the drive. This time it completely stopped and the frequency went to 0.0. Fortunately I was raised above the part. This has me concerned because I need this machine to be producing parts. Anyone have the same problem?

[bobeson]

I have had problems with my spindle speed changing unexpectedly, without command and without the spindle DRO changing. I verified the behaviour (clearly audible) with a strobe tachometer. I contacted Tormach about it, and they suggested reseating the ribbon cable that connects the machine control board to the DB25 parallel port connector bulkhead at the bottom of the control cabinet.

I'm not sure that this actually fixes the problem. It does seem that the combination of shutting down the machine & computer, opening up the cabinet, wiggling and reseating all the connectors I can find, and then restarting the machine, results in the problem usually no longer being apparent.

The Tormach tech I spoke with speculated that this problem may be caused by the nature of the spindle speed control mechanism, which uses a low-pass (or similar) filter to change a variable duty cycle digital output from the parallel port to a proportional analog control voltage for the VFD. I suppose an intermittent connection is being blamed in this instance, which presumably results in a lower duty cycle due to missing pulses. Given that I observed a rather consistent ratio of programmed to actual spindle speeds during the fault condition events, I'm still scratching my a head a bit over the loose connection theory. In any case I don't seem to encounter the problem much any longer; possibly since I last carefully reseated everything, or possibly by luck.

I will say, this and other parallel-port-related matters have me wishing for a control interface not based on such old-time bit-banging across a DB25. I don't know if Mach can decouple the real-time-sensitive motion control subsystem code easily, but I know that EMC's architecture allows for a completely independent embedded real-time hardware subsystem to handle all timing signal generation. Such a motion controller could exist entirely on the machine control board, and allow a huge expansion of I/O (axes, discrete control ports, more feedback, etc) and decouple the user interface PC (eg "troublesome" Windows machines) from causing timing issues in the machine control signals that are generated. I wince when I can hear my keyboard typing echo in the changing pitch of the stepper motors while my machine is operating. That just shouldn't happen! All it would take is an FPGA and some coding and debug time, and you'd get rid of so many parallel-port headaches. You could still use the parallel port, just change it to a non-real-time higher-level communication protocol to exchange motion control commands and status data, rather than discrete signaling. Or better yet, go with a dedicated USB bus, but that's all besides the point. IMHO, the motion control timing generation belongs inside the actual machine tool, not in the user interface PC.

Once you have that, then properly closing the loop on control matters like spindle speed gets trivial, and then we have a much more robust machine in general. In the meantime, I've broken several tools thanks to uncommanded slowing of my spindle during cutting operations. I'm crossing my fingers that it won't happen again.

BTW, despite this gripe-fest of a posting, I still love my PCNC1100, and am a big fan of Tormach in general. In fact, I'm going back out to the shop to use it right now...

[zephyr9900]

bobeson, there are already a number of USB-connected external step generators. SmoothStepper is one in particular that is designed to work with Mach. There is a SmoothStepper forum over at the Mach board. I was considering trying one of them until the new 3.042.029-based PCNC software came out last fall and works so much better on my machine than the old 2.xx-based PCNC.

And, of course, Tormach themselves now have a dedicated control box using embedded Windows OS rather than the general-purpose Windows.

Randy

[bobeson]

I realize that many variants of control products currently exist, but most of the existing ones have issues with the control system/UI/mach integration, and in any case they would all need proper integration with the tormach machine controller itself to really all make sense. Either the USB interface is a simple FTDI parallel port replicator, and you're just compromising the real-time fidelity of your PC timing control even further, or they are using a custom interface to a motion control timing generator, and most seem to rely on their own software interface, not mach or EMC. But like I said, the integration with the actual Tormach machine controller board is the key - close the loops on one board, and make the existing machine better.

[Tormachmaster]

I have not upgraded mine yet I am not sure it will make that big of a difference for me I cut alot of parts and have no problems so was the upgrade worth it? Are these problems you talk about here happening with all of the upgrades?