For those following: New Technics SP10 motor controller specification

[Steerpike]

More edits and improvements to the PWM controller circuit/schematic on my
PAGE 4

[Steerpike]

Tacho signal amp & pulse shaping circuit added to page 4.

[Brett]

Still following with great interest whilst I complete other projects. Then I'l start to tackle the SP10's.

[nigelwright7557]

Are you using a PID algorithm to keep the motor speed steady ?

[Steerpike]

Originally posted by nigelwright7557
Are you using a PID algorithm to keep the motor speed steady ?

Not decided yet, but most likely not PID (proportional /integral/derivative). I'm leaning more towards a state machine PLL with analogue lead/lag compensation after the PWM output.

Tacho preamp & shaper tested in hardware this evening, slight mod to circuit - I ALWAYS forget the op-amp DC bias path to ground!

(The symmetrical zener clamping might seem odd since this signal is going into a microprocessor, but it MAY also be used to drive an analog F/V converter to bring the PLL into lockable range - depending on which type of PLL I like best.)

[BrianL]

Even a PLL has some transfer function.

[nigelwright7557]

Originally posted by BrianL
Even a PLL has some transfer function.

What ever you use will be like a servo mechanism.
There will be an error signal being fed back. The gain of this is important if your not to get undershoot or oscilation. Its about getting the critical damping just right.

I have done quite a bit with PID and that gives very good results once you get the gain right.

If i remember correctly there is a PID example on the Microchip website www.microchip.com

[Steerpike]

I see you are all getting impatient for more maths!

My online write-up still stops short of describing the closed loop transfer function, which will incorporate both feedback loops (velocity and phase). And how & where the closed loop poles lie, and how the lead/lag compensator can be used to correct factors such as rise time, overshoot and steady state error, by adding poles & zeroes that shift the root locus to where one wants it to be.

As I mentioned before, I can't measure the root locus until I have a working DC motor, to do dynamics tests on. I'm making the PC boards that will give me that now.

Even a PLL has some transfer function.

Indeed it does.
This is exactly what I was writing software to simulate; I need to evaluage the linearity of the ASM.