A-B Quad encoder input module


Thread Starter

Michael Little

Does anybody know if the is a free-standing module that takes an A-B quad encoder signal and outputs a pulse and direction signal? It looks
like it will be too much work to configure the DAQ card I'm using to handle the A-B encoder signal, and I don't really want to use just one
phase of it if I can get the 4x multiplier.

Mike Little
Application Engineer
Compact Air Products
864-647-5454 x318

Steve Bailey

Many of the Micro PLCs can be configured to accept a quadrature pulse train and also have PWM and frequency output modes for their DC outputs. Try GE Fanuc, Cutler-Hammer, Automation Direct, A-B, Omron, etc.

Johan Bengtsson P&L Automatik AB

Are you sure you have to do anything at all?
If you really need the resolution it is of course needed to do something but if it is good enough to count once per full sequence it is quite simple (see below).

It might be quite tricky to fix what you need if you want pulse and direction and full resolution since that would rise other problems like what signal reaches the input first and so on (see below the simple solution).

Simple solution:

I suppose your signals looks like this:

_____ _____ __
A __| |_____| |_____|
_____ _____
B _____| |_____| |_____

If you go backwards the signals are changed following the same sequence backwards.

If you simply take oen of the phases as pulse and the other as direction and study the waveform carefully you will see that:

If you chose A as pulse and B as direction:
Every time A goes from low to high going forward the B signal is low, and if you go backward it is high.

If you chose B as pulse and A as direction:
Every time B goes from low to high going forward the A signal is high, and if you go backward it is low.

This effectively means it simply works without anything extra If the counter is counting the wrong way reverse the wires.

Hard solution (full resolution):
There is no way to see if the direction is reversed until one of the phases changes. In that instant you have to change the direction pin. It is not a good solution to immediately send a pulse to since you have to wait to make sure the
direction pin is going to be read correctly. this rises some timing issues possibly limiting the top pulse speed more than needed (not necesarily of course). If there is a device doing this (and of course there is somewhere) then this have to be
something that takes this into consideration.
If a suitable delay and pulse width can be decided that is low enough to ensure max frequency and still high enough to ensure that the signals will arrive in the correct order it would not be that hard to fix it with for instance standard logic circuits and a RC-link to provide the necesaty delays (please reply again
if you want more help with that).

An easier solution is if you could accept to use half the max resolution and a half pulse hysteresis, this could be quite easily done with standard logic circuits and without further delays. (reply if you are interested in more help with this).

/Johan Bengttson

P&L, the Academy of Automation
Box 252, S-281 23 H{ssleholm SWEDEN
Tel: +46 451 49 460, Fax: +46 451 89 833
E-mail: [email protected]
Internet: http://www.pol.se/
Johan's "simple" solution will work for 1 pulse per line provided that you don't care that your "up" pulses occur half a cycle away from your down pulses.

At the chip level, Agilent's (ex HP) HCTL-2020 quadrature-decoder-counter IC, which I designed a dozen years ago, outputs a pulse-and-direction signal pair with 4 pulses per line. Its main purpose is to accumulate in a counter, but I had a couple of extra pins, so I threw this feature on, not knowing whether it would be of any real use. All of Johan's concerns about timing and reversal are taken care of in the IC design. All it really needs is a 5V supply and an input clock frequency of a few MHz.

Curt Wilson
Delta Tau Data Systems

Setbacken, Bob

If you don't mind doing a little breadboarding, the LSI 7084 will give you what you want. You can get info from them at 516.271.0400.
Or, their address is 1235 Walt Whitman Rd, Melville, NY. 11747


Bob Setbacken
RENCO Encoders, Inc.
U.S. Digital at www.usdigital.com has a broad line of converters available in module form. The PC6 module would work for your conversion.

I particularly like the PC3 module, which is useful to translate TTL logic to 24vdc levels, say for connection to A-B's MicroLogix PLCs. It
includes a 24v/5v converter and hooks up with Molex connectors.


Tom Rhodes
Pyxis Engineering Corp., Redwood City, CA
e: [email protected] v: 650/365-2694 f: 650/365-2695
"Industrial Electrical Engineering Services"
Wandering off the topic a little . . . I have been surprised at how good the USDigital stuff works for how inexpensive it is. We did an app where we took single ended hall outputs for a brushless servo motor and converted them to differential signals using a product from
USDigital. The part lives in an environment of about 100 deg C and works very well for less than $20 for all three channels.

Ken Brown
Applied Motion Systems, Inc.