PLL question, maximum freq multiplication?

[tony ennis]

I'm finally getting back to my eQCGB after a year of delays, unemployment, etc etc

I need a phased-locked-loop to multiply by 200 the frequency generated by an encoder (32 hole index?) on my spindle. I want the PLL to sync up pretty darned quick. The input frequency could be 1 Hz if the spindle is doing 1RPM and a relatively large TPI thread is being cut.

So, is 200 a pipe dream? I am inexperienced but it sounds unlikely to me unless I want to wait 15 seconds or something for the sync, lol.

Would I be better off designing a series of less ambitious PLLs?

[vger]

A higher count encoder will be your solution. I did some work with PLL's many years back and at low frequencies the lock in time is indeed long. Also, there is overshoot and ringing as it locks in so daisy chaining will cause all sorts of unpredictable issues.

Are you looking at to get 6400 ppr specificly? I've got some 5000 ppr encoders (used, working) if you are interested. Heck, I'd give you one if you'll pay shipping.

Steve

[tony ennis]

Just back to this. Thanks for your generous offer, but I gotta build this ;-)

I should have an experiment to learn more about PLLs.

[RomanLini]

You could use a high speed microcontroller to generate 200 output pulses for each input pulse. It would sync almost instantly (after 1 input pulse) but it is still only good for a tacho signal not for positioning as it requires a relatively constant inpput pulse train.

As Vger said the best solution is to use an encoder with more pulses.

[tony ennis]

pm sent.