[dkochan]

(Or "Spindle Upgrade: FAIL")

Ever think carefully about doing something for a long time? And then finally spend the money to do it? And then get kicked squarely in the boys for your trouble?

I'll put the bottom line at the top. For those of you with the series II controller/vfd, how is the speed accuracy? I.e., if you calibrated it at 500RPM, what's the speed at a commanded 1000RPM? 1500? 2000?

My series I VFD was actually very, very close all through the range. Damn close. Scary close for an analog VFD.

My upgrade? Not so much. Calibrated at 500RPM (and THAT took some doing - long story below) it's OK right there. Command 100RPM you get about 25RPM. Command 800RPM and you are at about 980. You reach max HZ on the VFD at about commanded 1450RPM, which nets me over 2K at the spindle. No change from 1450 to 2000. And no change from 100 to 200. Which is odd.

Is this normal? I can't think it's normal!

Anyone with a series II upgrade or machine willing to send me speeds through the range? Actually I don't have to have speeds - the frequency off the front of the VFD would be fine. S100, S500, S800, S1000, S1500, S2000? Would be much appreciated!

Long version:

I ordered the upgrade for my 117x SN'd series I (about 2.5 years old) and it arrived Thursday. Installed it Thursday night, installed the new sofware, etc. No workie. Checked all the wiring. Short version - installation was good to go. Checked out the drive on the 'net - got some technical docs. Bottom line: VFD DOA. Bad fan of all things.

Major unhappiness. I actually had some time to work in the shop this weekend. Had a proto that's been screaming to get out of my head and I finally had it CAM'd and ready to go. Wanted to figure out the T/C tapping head that also showed up Thursday.

So I called Tormach the next morning and sure enough, the hf29 trip code means a bad fan. They were going to send me a new drive. Ground. I really wanted to cut metal this weekend so I bumped it to overnight/Saturday. Actually, UPS bumped me. In the rear. I was willing to spend the $$$ to get the machine running this weekend, though. I'm crazy that way.

The nice lady in the brown truck brought the shiny new VFD yesterday morning. Ripped out the old one and stuffed the new one in. Checked the wiring at least three times.

Fired it up. The spindle turns! YES!

Began the calibration instructions. S500, check. 64.8Hz WTF? When you're looking for 35 (34.6) that's pretty far out of range. Stop. Check wiring. Fire it up. 64.8? WTF? That's not an error you can tune out of with XMLtweak/PCNCconfig.

This is getting long, so I'll spare you all the process. Eventually I figured out that the (also new) control board was sending 1.2-10.2 VDC to the drive for speed control. This was happening with the manual speed control and with the computer commanding speed. Apparently the range is supposed to be 0 (not really zero, but...) to 5VDC.

So anyone looking for a 10K RPM spindle, here's your answer. Just don't talk to me about bearing life. (No, the spindle would not have gone 10K, since the max Hz param in the VFD is 170.2...)

At this point I figured either the controller was bad, too, or the gain was just horribly misadjusted. I couldn't do anything about the first option, but lo and behold there was a VFD gain pot on the controller. Since Tormach didn't pick up the phone yesterday I gave it a go.

My best meter now tells me that the range is 0.072VDC to 5.00 VDC. Great!

I "recalibrated" the spindle speed. Tweaked it in, but it took several tries. Now, though, I am at the spot described at the top of this discourse. (I am really, really making an effort to not turn this into a rant. You probably can't tell from there, but it's not easy.) I uninstalled and reinstalled the control software. I tried everything I could think of here.

Since I can't call Tormach until tomorrow, I am just looking to see if this sort of non-linearity is normal? My sincere hope is that it's a bad board and that I have squashed the signal so badly with the gain reduction that I have changed the shape of the voltage curve, and thus the speed curve. Because if this is normal, well - it can't be normal.

Thoughts?

Related question - as if this isn't long enough already - How does Mach3 send spindle speed? Is it possible to do more than single point calibration with this sort of setup? I'm be happy to put in the time to measure and tweak if it's do-able.

Thanks,
Dave

[zephyr9900]

Dave, I'm sorry to read about your VFD problem. You've probably read the Series II troubleshooting guide already but I'll link it here for other people reading this thread because it has a VFD section.

Both the analog and digital VFD's are intended to be controlled with a 0-5VDC signal. The analog VFD drive calibration guide details a 2-point calibration (pretty much like setting up an analog voltmeter--range and zero) but I don't see anything in the Troubleshooting Guide about calibrating the digital VFD--only verifying the input signal voltage. It is funny, having read that, that your control board was putting out 0-10VDC. Hopefully the guys at Tormach can help you work through it.

Randy

[dkochan]

Thanks, Randy. Saw that - it led me to the voltage problem.

I think the controller is whacked. Here are a couple of graphs - the curves just can't be right.

Here's commanded vs. actual spindle speed:
(Sorry - left numbering off - each tick is 100 RPM)



Here's commanded speed vs. voltage:



Looks to me like the VFD is tracking right where it should be.

Dave

[dkochan]

Just an update: Bad controller board - replacement on the way!

Dave