Synchronized PWM

[moonrock]

Hello. It's my first appearence on english-speaking forum, so sorry for my

english.

I'm experiment with my own stepmotor driver and i came to conclusion, that

chopper current stabilization is too noisy, even through i'm set chopper off-time

about 15-25 uSec and smaller.

I'm read at this forum, that geckos(maybe another drives) using another method -

synchronized PWM with constant switching period.
it's good method to eliminate motor noise, but when i model this current

controller, i find out intresting thing in current behavior.
After current was rise and reach comparator level(marked with mag. line),
it may osscilate on different ways(pictures 1,2,3), depending on where fisrt

switching occured(red dot) relative to PWM cycle. Current looks like "double

peaks". In different cases current swing may differ twice, but it's not good.
i want to force current to look like on picture 4, where swing is minimal.

Any ideas?

[Mariss Freimanis]

Synchronized PWM is an over-constrained feedback loop and it results in 'double pulsing'. Add some feed-forward compensation to your feedback loop (sum a positive slope to your sensed phase current). I cannot get into specifics because it is proprietary information.

Mariss

[moonrock]

Thanks, Mariss, for your answer.
It's clear about PWM. I'll continue my experiments.

I have another question. As i see, gecko dont using slow decay during motor rotation, only when it stopped. But using mixed decay results lower motor heating. Why you using fast decay only? Correct me if i wrong.
In my 4axis driver i had to set decay to 35/65 (fast/slow) to reach comprehensible heating of motor.

By the way, taking an opportunity, I invite you to visit our forum www.cnczone.ru