Mesa 7i84 encoder addressing

Hello.

I am currently using a 7i84 in mode 2 (input pins 16/17/18/19 for encoder input).
The motion is a little jerky when jogging and thus am working on bringing in ilowpass (with reference to "jog.hal with ilowpass").

The current encoder used in hal from the 7i84 is addressed hm2_5i24.0.7i84.0.0.enc0.count

In summary, is there a syntax to obtain the A and B phases separately from the 7i84 encoder inputs other than using the (combined?) hm2_5i24.0.7i84.0.0.enc0.count address? Reason below:

Inputs for "jog.hal" encoder function calls for separate A and B phased encoder inputs:
# The MPG inputs
net mpg-a encoder.0.phase-A <= ???
net mpg-b encoder.0.phase-B <= ???

which I intend to set up only via phase A input using the total count thus:
# The MPG inputs
net mpg-a encoder.0.phase-A <= hm2_5i24.0.7i84.0.0.enc0.count
#net mpg-b encoder.0.phase-B <=

i.e. nothing for phase B.

I will be updating the HAL file tomorrow and hopefully this will work OK and a simple adjustment of ilowpass scale will make up for any lost pulses in the B phase. But if this is not successful (possibly with respect to defining direction?), is there a syntax to obtain the A and B phases separately from the 7i84 inputs to then net both A and B phases into hal as required above?

Many thanks,
Marty.

You don't need the A or B signals at all, only the count

(A and B are used in the jog example because a software encoder component was used, none is needed here)

Wonderful news, thanks Peter.

So just to confirm for the hal file, this would be correct, leaving the B phase out:
# The MPG inputs
net mpg-a encoder.0.phase-A <= hm2_5i24.0.7i84.0.0.enc0.count

as I can't see any other appropriate input pins for the encoder function.

Just re-read your reply and I don't think I thought about my reply enough.

Given what you've said, I think I can just take the encoder counts (hm2_5i24.0.7i84.0.0.enc0.count) and have it come directly into the ilowpass function, the encoder function (and therefore it's inputs of A and B phases) are not needed.

Am I on the right tack with that?

Yes, no encoder hal component or A/B signals are needed, those are only used
if no encoder hardware is available.

Thanks for the swift response.
Given it's 1.30am here, now it's time for bed.
Shall test out the modifications tomorrow.
Again, thanks Peter!

I have done something similar here:- forum.linuxcnc.org/38-general-linuxcnc-q...ing-a-mpg-to-a-7i85s

You have to be careful with the scales.

Thanks everyone. It is all incorporated now. Only fine tuning necessary was to reduce ilowpass gain to .0075 from the typical .01 to balance the acceleration settings of the machine.
Thanks for the input.