[CAD_CAM_EDM_DRO] Re: Major encoder trouble, need advice.

Hi Randy,
I think you are missreading my replies, or I'm missreading yours....

I've tried with two OTHER, different encoders (one AMT and one Hengstler)
and it works with BOTH of them. And I've tried TWO E7P's and it does NOT
work with ANY of them, all else being equal. I've also tried two different
motors and two different drives, the only time it does not work is when
there is an USDigital E7P in the mix - that is why it "has to be" the
encoder.

Now, if I look at the encoder signals with the motor power supply OFF, it
looks just as it's supposed to (at least to me it does). The complementary
pairs are always the inverse of each other, rise and fall times are WELL
within specs and so is the "high" voltage at ~3.5V. The "low" voltage is
0.45V while the specs says it should be <0.4V but I doubt that's the problem
sice it's a differential interface.

So, the encoder isn't faulty per se. It's just way more sensitive to noice
than the other two are. Or, at least that's my theory anyway....

Thanks!
/Henrik.

-----Original message-----

As I recall, you have tried at least two encoders with, more or less, the
same results. Is it possible the problem isn't with the encoder at all,
but
the control or driver?

If I read your reply correctly the cable is quite short, only 300 mm or
around 12 inches, that shouldn't be long enough for any real cable problems
in any case. I have run cables for 5 volt encoders for well over 50 feet
with no problems most of the time.

In some cases where I was not getting the voltage swing from zero to a good
positive voltage, I used either pull up resistors, 10K or even in some cases
as small as 1K. One of the best solutions I use when I have a problem with
the voltage is to fabricate a small interface board with simple NPN bipolar
transistors, such as a 2n2222 or 2n2222a to clean up the signal and drive
the square waves to the full positive voltage. It isn't a fancy integrated
circuit, just some transistors and resistors, but it works. This way, if
you have any noise on the cable this simple circuit will pretty much clean
that up as long as it isn't enough to falsely trigger the transistors.

In a perfect situation, you would have some spare encoders and
driver/controllers to swap out and make a definite diagnosis as to which is
the problem.

If the signals you are getting at the driver/control end are clean, and I
think you indicated they were, then it should work and that would also tend
to eliminate any problem with the cables.

Randy Abernathy
Industrial woodworking machinery installation, design, repair and servicing
Over 24 year experience with factory training
Fax: 770-974-5295
Phone: 770-974-5295