Re: [CAD_CAM_EDM_DRO] Re: Tachs and encoders.

ron ginger wrote:

> I knoe everytime this topic comes up Jon can be counted on to beat his
> tach drum and tout the advantages at low speed. But it seems to me the
> tach output will drop off to zero at low speed. In fact, if you turn a
> motor slow enough its tach will generate no measureable voltage.

Yes, this is true. I get very good, clean tach output down well below
.001" per minute on my system. The stick-slip friction of the table
really dominates at .01 IPM, but the tach is STILL giving completely
valid and useable velocity feedback to the servo amp. Yes, I can't
MEASURE the voltage out of the tach, but I can EASILY see it is
still keeping the servo loop closed!

> If the motor is running so slow that the time between encoder steps is
> to long to adjust the servo loop then wouldnt the tach output also be
> zero?

No, clearly not true! You can't read it on a DVM, but the instrumentation
amp that picks up the tach signal and amplifies it a little before feeding
it into the velocity error amp processes it just fine. At .01 IPM, with an
encoder resolution of 20,000 counts/inch (5 TPI screw, 1000 cycle/rev
encoder in quardature) the encoder is only giving 3.333 counts/second.
But, the servo amp has a rolloff about 500 Hz in my system, and is still
supressing all internal and external perturbing forces, even if the encoder
never sees them. The tach voltage for a 7 V/ KRPM DC tach is
7 V/KRPM * .01 Inch/Min * 5 R/Inch * 1/1000 KRPM/RPM = 350 uV

> So, how does a tach help at low speed?

Analog electronics don't stop working below 5 Volts. And, in fact, good
analog electronics will work with signals that you can't detect with a
DVM, and can barely see a hint of on a good oscilloscope. 350 uV
sounds like an impossibly small signal for the servo amp, which is
also processing 70-100 V at several amps about 2 inches away,
but it DOES, indeed, work, and is stable, reliable, repeatable, predictable,

etc.

Jon