Page 1 of 1

UC100 or other controller rigid tapping

PostPosted: Thu Jul 30, 2020 1:57 am
by infamous_panda
I am trying to re-purpose a router gantry frame from AVID CNC for a dedicated 4axis hole drilling and tapping station. I plan to drive a R8 spindle with an AC servo motor connected with a toothed XL belt at a 2:1 ratio. I am at a bit of a loss as to how to wire this together and the electronics I need. My other router setup uses a UC100 and I am guessing that the 5 input pins are maxed out with

1. Probe
2. XYZ home/limits
3. Y2 home for squaring
4. blank
5. blank

As I understand it, UCCNC needs three input signals for the spindle encoder? Is there a way to supplement the inputs of the UC100? If not which controller would be recommended. Also many of the threads I have read through use an external encoder than the one built into the servo or servo amplifier. Is this something that is really required? Any guidance or reference links for this sort of thing would be greatly appreciated.

Thanks.

Re: UC100 or other controller rigid tapping

PostPosted: Thu Jul 30, 2020 8:15 am
by dezsoe
You need one more input. UC400ETH has the double inputs and outputs as the UC100, AXBB-E has 6+5 inputs, UC300ETH has much more. I recommend AXBB-E, because that has some isolated inputs and outputs and also a safety charge pump driven output.

The separate encoder is needed, because the internal has a too high resolution. The problem is that the inputs have a maximum frequency and the high resolution encoder makes a too fast input. For tapping (and other synchronized jobs) a 100 CPR encoder is enough. (Also, because the index output of an encoder is gated with the A and B signals, the index pulse is so short with a high resolution encoder that the controller cannot detect it.)

Re: UC100 or other controller rigid tapping

PostPosted: Wed Sep 23, 2020 12:17 pm
by DavoH
dezsoe wrote:...

The separate encoder is needed, because the internal has a too high resolution. The problem is that the inputs have a maximum frequency and the high resolution encoder makes a too fast input. For tapping (and other synchronized jobs) a 100 CPR encoder is enough. (Also, because the index output of an encoder is gated with the A and B signals, the index pulse is so short with a high resolution encoder that the controller cannot detect it.)


I have a UC300ETH controller and want to do rigid tapping and have been looking at various incremental encoders. I already have a good spindle index pulse so I guess I just need the A and B inputs from an encoder.
My spindle's max speed is 6,000rpm so I need an encoder to handle the high speed as well as a lower speed for tapping.
You mentioned using a 100 CPR encoder. Am I correct in thinking that means an encoder of 400PPR? Everybody seems to list them as PPR in their specs.
Will 400PPR be OK for tapping and also speed regulation up to 6,000rpm?

Thanks,
Davo.

Re: UC100 or other controller rigid tapping

PostPosted: Wed Sep 23, 2020 12:38 pm
by dezsoe
Hi Davo,

Good question. 6000 RPM is 100 rev. per second. 400 PPR means 400 pulses per rev., but it contains A+B pulses, so it is 200 pulses per channel. 100*200 pulses per second is 20kHz, the limit on the inputs is about 50kHz, so it should be good. Read more details in this post.

Re: UC100 or other controller rigid tapping

PostPosted: Thu Sep 24, 2020 10:37 am
by DavoH
Hi deszoe, thanks for the reply and the link to the other post 8-) - I previously did a search for encoder but that post didn't come up.
So it looks like I'll be OK with 400PPR.
I have an AMT103 incremental encoder that I want to mount (somehow).

My existing index signal goes through a 2N2222A transistor base, the emitter is to ground and collector to +24V/UCBB input.
I assume the same input setup would work OK for the AMT103?

Many thanks,
Davo.

Re: UC100 or other controller rigid tapping

PostPosted: Thu Oct 08, 2020 8:07 am
by DavoH
OK, I have the AMT-103 encoder mounted, I tried connecting its index signal to the 2N2222A transistor instead of my original index signal but all I see in the software is a jumble of very large numbers so I don't think the input is seeing the signal correctly.
According to the encoder's electrical specifications the outputs are TTL square wave, so they should work OK on the 2N2222A? Somebody please tell me yes or no, I'm no electronics expert.

I see Note 1. on the spec sheet says "Some steppers may leak a magnetic field causing the Index signal not to work properly.' It's not on a stepper it's on the top of a DC motor but maybe that's the problem.
However I'm reluctant to connect the A and B signals until I'm sure I'm not going to damage anything, I can substitute another index signal if somehow I have damaged this one.
Please help.

Many thanks,
Davo.