[dkaustin]

Decided to check my actual spindle speeds (PCNC 1100 w/vectorless drive upgrade) against the input value, and this is what I got. It's not BAD (less than 5% error over most of the range), but I'm wondering if it could be better. Is this comparable to what the rest of you are seeing?

The frequency value at 500 RPM input is 34.6, exactly what it should be per the Tormach calibration procedure. I checked the Emerson Commander SK manual for tips, but nothing jumps out at me.

Is there a way to massage this curve, or should I just nudge the input value on critical jobs to get the correct actual RPM?

TIA

[BAMCNC.COM]

I would prefer it just like your graph shows actually, atleast its not the other way around.

[SCzEngrgGroup]

Don't know why you'd care about a few percent, especially since it's a few percent high. Getting it close will gain nothing.

Regards,
Ray L.

[gjvander]

Decided to check my actual spindle speeds (PCNC 1100 w/vectorless drive upgrade) against the input value, and this is what I got. It's not BAD (less than 5% error over most of the range), but I'm wondering if it could be better. Is this comparable to what the rest of you are seeing?

Thanks for sharing. Can you give us a little more info on how the data was acquired - specifically, how did you measure the spindle rpm?

Regards,

Geo

[philbur]

Correct cutting speed is a long, long way from an exact science. Your graph as is looks way better than good enough, even for the most critical of the most critical operations.

Phil

PS: What do you think the correct rpm would be before the advent of the VFD.

[dkaustin]

Yes, I agree that, as a practical matter, the error is unimportant as far as milling cutters are concerned. I checked the speeds because I am working up the nerve to try some tapping, and I thought accurate speeds would be critical for that.

Geo, I checked the actual RPM using a hand-held laser tach. Of course, I'm assuming the tach is accurate

[dbrija]

Yes, I agree that, as a practical matter, the error is unimportant as far as milling cutters are concerned. I checked the speeds because I am working up the nerve to try some tapping, and I thought accurate speeds would be critical for that.

Geo, I checked the actual RPM using a hand-held laser tach. Of course, I'm assuming the tach is accurate

A tension/compression head takes that worry away, especially when actual vs. commanded is that close. I tap on my 1100 all the time, no issues so far.

[Al_The_Man]

vectorless drive upgrade)
TIA

Sensorless Vector?

The only way to get closer with a VFD is with an encoder option.
V/F has no feedback, Sensorless Vector relies on the controller electronically sampling the current/voltage/frequency, so what you are seeing is pretty good, IMO.
Al.

[TurboStep]

The spindle speed can be calibrated using the Tormach PCNC Config utility. Mine was about 12% too fast at the top end and I saw no reason NOT to adjust it.
Step

[LRF]

I agree with dbrija, its not that significant an issue. I have checked my 1100 at low and high speed. I was at the time also concerned about tapping. My spindle spped at 400 setting is 389 so if you program 411 you get 400. Frankly I see no difference between the 2 settings. The T/C takes care of the difference.
At a 4500 I get actual of about 4800. Never measured the speed when set at 5000 but maybe i'm getting more then the 5000 max. Doesn't seem to matter one way or the other.

[philbur]

The spindle speed will vary depending on the magnitude of the applied load in any case.

Phil

[Zetopan]

Decided to check my actual spindle speeds (PCNC 1100 w/vectorless drive upgrade) against the input value, and this is what I got. It's not BAD (less than 5% error over most of the range), but I'm wondering if it could be better.

I did a calibration on my series 1 spindle speed after upgrading to the new
series 2 & 3 spindle drive.

Based on my measurements you could get the actual unloaded RPM to be
within +/-3% of the commanded RPM everywhere. I will need to redo all of
this after finishing my upgrade to the Series 3 steppers and drivers since
the control board is also being replaced as part of that upgrade. YMMV.

See the two attached for the actual data plots. Note that the actual speed
is slightly parabolic rather than exactly linear.

[SCzEngrgGroup]

Unloaded spindle speed is really a pretty useless number. What you really should care about is the spindle speed under load, which WILL be less than the unloaded speed, unless your spindle has a closed-loop speed control. Even then, a few percent is of no consequence. Reminds me of an old saying: "Measure it with a micrometer, mark it with chalk, cut it with an axe". Speeds and Feeds are approximations, with FAR more than a few % error inherently built into them, due to the many uncontrollable factors involved. A spindle that is off by 5% will have no measurable negative effect on machine operation, and a spindle that is accurate under all conditions to within 0.001% will have no measurable benefit to machine operation.

Regards,
Ray L.

[MichaelHenry]

I checked the speeds because I am working up the nerve to try some tapping, and I thought accurate speeds would be critical for that.

For tapping, it might be best for accuracy to be greatest at low speeds, though the T-C heads probably eliminate most concern there. I generally tap at 500 rpm for 1/4" and under holes and have had no problems there. I generally use SpruutCAM to create the code and the post defaults to a 10% feed adjustment which has been working fine. That is much greater than the speed error you documented so I would be inclined to leave your setup alone.

Mike

[BAMCNC.COM]

Great thread here, and MichaelHenry brings an excellent point. Tapping at a lower speed is the Only way to "Safely Assume" you are ok without knowing the exact spindle speed.

[sansbury]

I also use a T/C head, and I've had it work fine with the FRO accidentally set to 80%. While I wouldn't recommend it, so long as you're within the excursion of the head, it can soak up an awful lot of error.