Using quadrature encoders with PXI-6025

Question

Hi!,
We are using an encoder for position data. The encoder gives A and B pulses which are phase shifted by 90 deg.
Grating Pitch = 40 microns
We make the connection to our PXI-6025 as such:
Channel A pulse goes to PFI 8 (GPCTR_Source)
Channel B pulse goes to DIO6 (GP_Up_Down)
Ground goes to DGND
We are able to increment/decrement the counter on our 6025 for every 40 micron movement
However, we are unable to quad the input and increment/decrement for every 10 microns. How do we achieve this?
We have an electronic circuit which converts the 40 micron A and B pulse from the encoder into 4 pulses spaced at 10 microns. But, using this single pulse mode does not allow for up/down counting.
We are using LV-7.1
, NI-DAQ 7.2 and Win2000 prof.
Looking forward to some quick help.
Thanks,
Gurdas
Gurdas Singh
PhD. Candidate | Civil Engineering | NCSU.edu

Answer 1

There's one tedious way that might work if you only need to post-process the data. However there are several possible problems.
The idea is to perform buffered semi-period measurements with both the counters on the 6025. Once you've collected all the needed semi-period data, perform a cumulative sum of each array so you end up with timestamps for all the transitions. Then you'll need to go through them to distinguish (+) from (-) direction quadrature transitions and generate a cumulative position array.
I don't think there's another way -- only semiperiod measurements will react to both edges of an incoming pulse train. However, it *is* possible in principle to get your 4x factor in resolution. Now for the problem list:
1. Your 6025E only has 1 DMA channel. One of the counters will have to be configured to use interrupts, which in turn will limit the maximum encoder rate you can track. I wouldn't count on tracking better than single digits of kHz, and you may not even be able to achieve that much.
2. The last time I did semiperiods on an E-series board, there was no way to specify which edge would mark the first recognized transition. If the input was low when you started, the first transition recorded would be a rising edge; if high, then a falling edge.
Knowledge of the initial state of the inputs is critical for your software quadrature decoding. I think the solution I used was to parallel wire the A&B channels to a couple DIO pins which I inspected before any encoder motion started. If your encoder jitters or vibrates about its nominal position however, you're probably hosed.
Are you sure you can't use the LS7084 to replace your other external circuit? The alternatives are either going to be trouble-prone gimmicks like I described above or the purchase of new DAQ equipment.
If you have budget for new DAQ equipment from NI, I'd suggest you look into one of the following:
1. Low-cost M series multi-function board
These have 2 32-bit counters onboard that are capable of quadrature decode. Each counter gets its own DMA channel for buffered acquisition. Plus you get a bunch of analog in, timed digital i/o, and can get analog outs. The boards are $475 with 2 analog outputs, $375 with none.
2. dedicated counter/timer board
The 6601 is cheaper at $295, but is limited to 1 DMA channel and has a slower onboard timing clock. The 6602 costs about double that but gives you somewhat more than double capability.
-Kevin P.