need help using vsd-a with linear motor

[daedalus]

Hi,

Im having problems getting the VSD-A drive working with a small brushless linear motor. What i am unsure of is how to calculate the pole count / rev and cpr of the encoder. Basically i have an encoder that gives me one pulse per 0.5 um, and the magnetic poles are 1 1/8 inch apart (1 inch magnet + 1/8 inch gap between). The forcer is just 3 coils of wire.

I am assuming that in this application you would want to go for 2 pole, and set the encoder cpr to the number of encoder counts in 2 1/4 inch. The problem with doing so is that the software doesnt allow cpr to be set that high. Is it possible to set higher limits by issuing spi commands manually?

The other issue i am having is due to the design of the assembly the motor is in, as i only have just over 1 inch of travel, so phase search fails every time. Is there a way to limit phase search to less then one "rotation"? I can fix this issue by remounting the motor on a longer travel, but if theres a workaround for testing it would be very helpful. Also am i right in assuming that this issue would go away with a vsd-e as i could use hall effect sensors?

[daedalus]

OK, i managed to get the axis working for like 5 minutes, and in that time i was able to do the "test motor settings" microstepping option. The calculated encoder count per "rotation" is over the maximum value the software allows you to enter. Is this a limit of the software, or the drives firmware?

Also is there a document listing the SPI commands for VSD-A? I cant seem to find this info anywhere.

[Xerxes]

It looks like encoder resolution is higher than maximum supported by VSD-A. The maximum is 50000 counts per polepair (50000 counts for every 2 1/4 inches in your case).

I could compile you a special firmware that allows 16 times higher resolution. Changing UVW phase wiring order may help with phasing problem as phasing position would change (may need swapping encoder channels as well).

Are the lines or counts in 0.5µm spacing? It makes 4x difference.

[daedalus]

hi Tero,

0.5um is with x4 decoding, so 2 1/4 inches should work out around 50800 ticks, and this seems close to the measured tick counts i have seen. If you wouldnt mind doing me a special build, that would be awesome.

As for the phasing problem, now you mention it, i did have the windings the other way round when it was working, i will swap back and see if that helps. Eventually im going to remount everything to a longer axis, so it shouldnt be too much of a problem.

Thanks,
Tom

[Xerxes]

By my math it is 114300 counts/polepair. I'll try modifying firmware within a week.

[Xerxes]

Here is the special firmware for high resolution encoders. I haven't tested it myself so let me know if it works well.

It internally divides encoder by 16 for commutation, so divide your encoder value by 16 in GDtool. Set it to 114300/16=7144.

[daedalus]

Thanks, i will take it into work and give it a try later this week.

[veteq]

Hee,

How is it going, got it woring allready, how about some nice movie when its ready?

Regards,

roy B.