Hardware PID Controller - Any advantage?

[David]

Hi all,
Does anyone have any experience in implementing a hardware based PID controller? I.e. using Op-amps? The advantage I see to it is reduced processor overhead and less complex algorithm programming, although I can see that fine tuning it (resistor values) may be quite cumbersome. Just evaluating a few of my options for a new design.

Dave

[KTP]

It is what the Gecko servo drive did in their G320 series of hobby drives.  They now are moving to a digital PID controller approach probably because it is easier to store/recall settings and debug.  The pots used in a analog PID controller will somewhat drift with age/temperature, not to mention vibration and this could be another disadvantage.

[tecman]

I used analog PID controllers for years.  Our company designed our own controllers and shipped thousands over the years.  Around 1990 we made the switch to digital and never looked back.  True it can take some cpu power to do it right, but the options available combined with no drift and perfect repeatability are big advantages.

There is a lot of good analog building blocks out today.  Biggest issue is part stability over time, parts matching and thermal effects.  Many of our analog controllers are still out there on a lot of machinery.  A solid analog controller can still outperform a digital one in certain cases.  Plus you rarely have "bugs" to worry about.  On the other side, more complex controllers require more complex circuitry and more pots to adjust.

Paul

[mikeselectricstuff]

CPU power is way cheaper than analog parts these days. Not to mention more stable, flexible, repeatable etc. etc.
Can't see any reason to use analogue these days.