AXBB-E encoder problem

[drumsticksplinter]

Hello,

I am retrofitting a Denford Triac with a AXBB-E and UCSB using UCCNC. I have managed to get on quite well so far, but I'm having a bit of an issue with encoder feedback. I am using the 0-10v analog output, which is working great! I am using a HEDS optical encoder for spindle speed feedback, which outputs 5v line signal on A,B and Z. I have this connected to the inputs on my UCSB which works up to around 1000rpm, after that the signal completely drops out. I was wondering if this might be because its only 5v signals and the inputs on the boards are ideally for 12-24v? If this is the case, does anybody have any suggestion on how I might be able to get my 5v encoder pulses to 24v please? optocoupler maybe?

Thanks

[cncdrive]

What is the encoder resolution? The controller can read upto about 50kHz encoder signal. If the frequency goes higher can cause the RPM to dropout, because the signal can't be read properly by the controller.
The UCSB is sure not the bottleneck, because it has very high speed optocouplers, they can throughput about 5MHz signals.

[Spuds]

Please excuse this interruption for a quick question.
Why would you use a UCSB breakout board with the AXBB-E controller that has integral breakout terminals ?
Aren't you running the signals through dual isolation ?
Just curious.
Thanks,

[cncdrive]

The AXBB-E has port#3 which is a IDC26 port with LPT port pinout, it is an extension port and is not isolated. An UCSB or a 3rd party LPT port compatible breakout board can be connected to it.
He probably needed more I/Os is why he used a UCSB board on port#3.

[drumsticksplinter]

Hi, the encoder resolution is 1000PPR, so if I'm correct: 1000PPR @ 1000rpm would be 33.3KHZ ish?? so that would make sense that at 1500 rpm it would be 50KHZ and the signal drops out.... My spindle goes to around 4000RPM, so 133.3KHZ would be needed to read my current encoder setup. I have the option of trying to find another encoder wheel to change the resolution to around 200PPR or I have another encoder, but with only 50PPR, not sure if this would be enough to try and achieve a setup capable of rigid tapping? The second encoder is capable of being supplied with 24v though instead of 5v, so may provide a more noise resilient signal.

Yes I am using the USCB for additional inputs as I will use all of the available ones on the AXBB-E

Thanks

[Robertspark]

Cd4017 decade counter as a frequency divider circuit

[drumsticksplinter]

ooh, I hadn't thought of that option actually, I wonder if there is a pre built module available?

Thanks

[Robertspark]

It's only 1 component per channel so you need 2, one for the A and a second for the B encoder channels. The index Z remains the same.... One pulse per revolution...

2 X cd4017
A small Veri board or prototype board
An some short lengths of solid core network cable to act as jumpers.
You could add on some push in cable terminals or screw terminals 2.54mm pin spacing


One thing to check..... Is your encoder a single ended or differential output type?

Does it have A+ A- B+ B- Z+ Z- (differential output) or A B Z (single ended output)?

Where in the world are you located?

[drumsticksplinter]

My encoder is single ended output so just A B Z outputs.

Thanks for the advice on the kit needed, you wouldn't happen to have some sort of schematic to hand at all would you?

I'm actually just thinking about this some more, I'd like to get the spindle up and running counting correctly etc for now. However, I don't think my drive and motor setup will provide tight enough control for rigid tapping. The motor is a permanent magnet DC brushed motor, of which I have some experience with before as this spindle motor was very similar if not the same model as the DC servo's I retrofitted on a cnc Bridgeport machine some years ago. However, its current setup with a sprint 1200 DC drive only uses tacho feedback. I may be able to (further down the line) use a DG4S-16035 DC servo drive to add encoder feedback into the motor drive closed loop and feed this back to the motion control for a tighter control loop?

I'm in the UK, Staffordshire btw

Thanks

[drumsticksplinter]

Hmm, maybe my understanding of analog servo control is flawed actually.... if anyone could put me right on this please? I believe my spindle motor: SEM MT30U4-36 was designed as a servo, but doesn't have encoder feedback for position on my machine. I have a sprint 1200 drive which is quoted as a speed or torque controller, feedback is via tacho. Would I be able to use this effectively enough for rigid tapping at all if I used my encoder feedback through my AXBB-E and UCCNC with analog output? I still think I could use the motor with an analog servo drive, but i'd prefer to leave as is if it would work?

Thanks.