Orientation problems

Hi,
I managed to setup sindle orientation and attached it to tool-change request . After long peods of time the orientation runs out by maybe 30 degrees . I'm asuming it is missing pulses . Can the mesa boards captre 2500 pulse encoder at 6000 revs? Should I look at wiring ? How can I force index enable on ?

6000 RPM is a 1 MHz count rate with a 2500 line encoder.
(A and B are 250 KHz square waves in quadrature)

This is possible to read with all Mesa FPGA encoder hardware but multiplexed
encoders need to have their filter bit turned off to resolve the 250 KHz signals well.
Also it requires good signal quality and may be at the edge of what most encoders
can do.

Possible solutions:

If you are using a Mesa daughtercard with multiplexed encoders
(7I48, 7I77, 7I52, 7I53, 7I85, 7I85S ), turn off the encoder filter bit

Make sure you are using differential signaling. TTL level signals
are much more susceptible to noise.

You can set the index enable with:

sets index-enable-signal

or logically using the tri-state component

The 7i33 was in differential mode . Is there a software filtering method ? We will try with a new cable . Also we're starting to suspect the encoder itself .

There is a filter enable bit for each encoder (man hostmot2), but with a 7I33 (and assuming a 5I20)
even with the filter on, the maximum count rate would be ~4MHz, much faster than your ~1 MHz count rate
with 2500 line encoder at 6000 RPM. So the encoder or wiring is suspect.

You might try checking the A/B signal with a scope (or HALScope if you turn it slowly) and check the symmetry and phase shift between A and B

When I checked it with the scope the edges were well defined but the top and bottom were hazy . But we have had spikes that don't show on the scope so I can't say that the edges are exactly right .

P.S. I checked the HAL and filtering is enabled .
Will check the encoder and cable when possible .

Hi,
I'd like to set index enable when m5 is encountered . When does motion.spindle-on turn off ? Does it turn off when M5 is seen or when zero speed is reached ? The inertion of the spindle should be enough to propperly see the index .