Encoder to DC tach circuit Wanted.

[Al_The_Man]

Anyone used or have details of LSI or Pic method of bi-polar DC Tach signal derived from A & B quad encoder pulses?
I need a working circuit rather than re-invent the wheel.
Al.

[NC Cams]

How about a tachometer IC???

You could AND the output signals and feed them to an LM2917 D-A converter chip via a buffer IC and some diodes for appropriate isolation. Easy and cheap way to get a voltage proportional to output RPM.

Sadly, this wouldn't tell you if you're running CW or CCW as I suspect you may have an interest to do.

Another trick might be to check out Red Lion Controls. They have some ready made RPM and direction sensing devices thay you might be able to plug and play with.

[Al_The_Man]

Thanks, I have a raft of circuits for uni-polar DC but cannot find any for Bi-polar (decided by the A & B pulse dir.)
AMC and Copley use a circuit in their drives but I have not been able to reverse engineer them. I need a compact LSI solution rather than off shelf hardware.
Al.

[unterhaus]

us digital sells a box that does this
obviously, the trick is to get the sign correct. I know the guys at New Micros http://www.newmicros.com/ have used the timer inputs as encoder inputs on motorola chips -- they say the regular timers work ok, and you don't have to use the special encoder inputs available on a lot of chips.

This article looks promising http://www.mcmanis.com/chuck/robotic...uadratrak.html
I didn't read it too far though.

For a tach, you'd have to keep track of the velocity instead of the position. I'd keep a running average like this: vel = (xnew - xold)/delta_time
then vel_filt = (7*vel_old+vel)/8
output vel_filt
vel_old = vel_filt
repeat
Outstanding issues are how to find the delta time and how to drive a pwm pin to give an analog out.

[Al_The_Man]

I will look into those articles, it seems that the US digital devices do not do exactly what I want to do which is to produce a Max ±60vdc analogue signal derived from a quad encoder.
IOW to replicate a DC tach from A&B pulses.
I have searched the web for solutions and am suprised that there does not seem to be any articles out there for this app.
Al.

[smarbaga]

hello al.
if you have the time and the programming experience you could use a pic or whatever to:
1. - count the encoder pulses.
2. - convert the encoder pulse count to set a digital pot.
3. - voltage amplify the pots output to a scale factor of your choice.
if you want me to draw you a schematic , i will,

[Al_The_Man]

Thanks, I will see how far I get and if I get stuck I will take you up on the offer.
One of the tricks is getting the Analogue polarity to match direction.
The biggest enemy is TIME
Al.

[pastera]

http://www.fpga4fun.com/QuadratureDecoder.html

Gives some logic for a Quadrature to count and direction interface.
Hook up the count signal to a frequecy/voltage converter chip (LM2907) invert based on direction and send to a HV push-pull stage.

Take a look at Apex Microtechnology's high voltage line amplifier line for an off the shelf output. EDIT: ouch - looked at the prices

take a look here: http://www.edn.com/article/CA484492.html


Aaron
Damn EE's always like complicating things! LOL

[One of Many]

This is as close as I had found. These will put out up to +/-10v for direction tach velocity.

For $173 it could shorten time to run once pumped to +/-60? USDigital has quite a variety of encoder converters.

http://www.usdigital.com/products/etach2/

http://www.genesisautomationonline.c...hp?cat=1&sub=0

I have yet to see a direct single chip quadrature to analog tach conversion. Could this be that it is not a directly stable linear conversion to simulate a tach analog output? Obvioulsy it is done, but I'd wonder if there are compensations involved based on encoder counts/rev, pulse width etc. seems like a direct freq/voltage conversion.

DC

[kenski]

http://www.avtron.com/pdf/encoders/k661.pdf

[Al_The_Man]

Thanks everyone for help, I think I have found something in my old analogue records that should be able to put together with 4 IC's. I will post the schematic if it works out.
Al.

[maquiza]

Did it WorK?
I am looking for one that can generate 6V/1000 rpms.



Regards,


Jorge