Mesa boards for BP Interact 1 Mk 2 conversion

I had just about decided to use a 5i20 + 7i33 + 7i37 but it appears that the 5i25 + 7i77 is a cheaper alternative and has 6 axes instead of four.
However the 7i77 is multiplexed and I was wondering if anyone had come across any performance issues with it?
Also the 5i20 is quite mature and well supported, is the 5i25 as well sorted?

Thanks guys.
BJT, I read your prior posts about your change from 5i20 to 5i25 so was curious to know whether you noticed any performance deterioration with the multiplexed 7i77. Presumably not, so I will order the 5i25/7i77 combo.

I'll be away from the shop for several months so I will just have the PC to play with.
Whilst I have no problems with the mechanics/electrics/electronics I really need to get a handle on the software. It is rather bewildering being faced with Linux, LinuxCNC, HAL, et al.
I hope I will be able to report some progress in a few months.

Yes the only real limitation of the muxed encoders on the 5I25/7I77 is maximum count rate. This count rate is 2.77 MHz on the 5I25/7I77 (with encoder filter off) and 555 KHz (with encoder filter on) These (with a safety margin added) limit you to about 12000 RPM and 2400 RPM max with a 2500 line (10000 count/turn) encoder. Usually the encoder itself will limit you to lower frequencies.

It is possible to get a little faster counting if needed by changing the mux rate from its default 33.333 MHz /8 to a higher speed. But this may limit the 5I25 --> 7I77 cable length.

PCW wrote:

Yes the only real limitation of the muxed encoders on the 5I25/7I77 is maximum count rate. This count rate is 2.77 MHz on the 5I25/7I77 (with encoder filter off) and 555 KHz (with encoder filter on) These (with a safety margin added) limit you to about 12000 RPM and 2400 RPM max with a 2500 line (10000 count/turn) encoder. Usually the encoder itself will limit you to lower frequencies.

It is possible to get a little faster counting if needed by changing the mux rate from its default 33.333 MHz /8 to a higher speed. But this may limit the 5I25 --> 7I77 cable length.

Peter, can you elaborate a bit more on the encoder filter? I have not seen the term used before and am curious about it. I have a 7i77 retrofit nearly done.

The hardware encoder has digital filters on its A,B and Z inputs. These filters work by counting up to maxcount (3 or 15) and down to 0 depending on the sampled input state. Only when the count is at its maxcount is the output of the filter set true, likewise, only when the count is 0 is the output of the filter set false. The filter is used to reject short noise pulses but let the slower encoder signals through.

The default maxcount is 3 sample times, but this can changed to 15 sampletimes by setting the filter bit true.
The sample time for non-muxed encoders is ClockLow (normally 33 or 50 MHz) but is quite a bit lower for muxed encoders, for example the 5I25 default encoder sample rate is 4.16 MHz, so with the filter off (3 samples needed)
you need 3 clocks or .72 usec minimum pulse widths to register (get through the filter). With the filter on the filter time is extended to 3.6 usec.

Gmouer,
Wow that would work out to 500mm/s with a 2:1 reduction and 5mm ballscrew.
More than adequate for a BP.

PCW,
does that mean that the digital filter averages the input over the number of counts?