Quadrature Decoding

I spent some time last night thinking about quadrature decoding. I worked out a circuit which does it using two D flipflops, 3 XOR gates, two Schmitt trigger NOR's, and a quad two-input multiplexer.

This gives a clock pulse on every transition, using different pins for up pulses and down pulses. Is this something which is useful for RepRap?

The part count (5 ICs)is a little ridiculous, which is why I would implement something like this in a PLD. Either an SPLD (20V8, 22V10) or a CPLD (XC9500XL) would work. For the $1.07, that it costs, I'd go for the XC9536XL which provides the option of adding more logic. I think it should even be possible to program the CPLD through JTAG using the SPI port of the controller.

Should I write the code and/or post schematics, or is this something we don't even need?

I use quadrature output from the magnetic rotary encoder board to control my extruder, and I just wire it directly to two pins on my Arduino.

Of course, it's already a pretty clean signal, as it's produced by a micro in the first place. The Schmitt triggers in your setup might be useful for cleaning up opto outputs on a home built optical wheel, but I haven't tried that yet. I did use the guts from an opto-mechanical mouse wheel encoder originally, but it wasn't very sound mechanically (too much CAPA involved), so I switched to the magnetic encoder.

Right now the code I'm using is interrupt driven on the quadrature side, so a noisy signal could slow things down, but I don't think I'm seeing that yet. I didn't write the code though, so I probably don't know as much about it as I think I do; is there an advantage to doing the decoding with hardware? Either way it's two pins on the micro, which was my main concern.

Wade

It's only about 20 lines of C, including debouncing, so a hardware solution is only necessary if you have a very fast shaft, or lots of them.

---

[www.hydraraptor.blogspot.com]

Thanks, nophead. That's all I needed to know. I'll let that one drop and focus on something more useful to occupy myself.

Annirak

I am interested in your design. This is part of a subject I have been working on very slowly as a back ground task (in my head) for a while. The foreground being mostly retraining and getting a functioning Darwin built

I agree with the sentiment that as an item for what we currently have (Darwin)it would have perhaps limted value.

As a potential component for where we might go to next though I think it has great relevance.

There are some of us who are putting some design effort/thought into alternative actuator models for potential future machinery.

The hardware quadrature input would be of great importance/relevance to a hardware solution for an entirely digital mostly hardware servo controller.

Something that would be eminently suited to CPLD based development. Perhaps with a SPI interface...........

I am happy to talk about this further if you would like to.

Cheers

aka47

---

Necessity hopefully becomes the absentee parent of successfully invented children.

Hi Andy,
This is the core of it. Actually connecting it to a counter, etc. isn't so bad. The CPLD code, likewise, would be pretty easy to handle.

Let me know what you think!

Brendan