Re: [oc] PID controller in FPGA?

[Charles Lepple <clepple@ghz.cc>]

jwen@canada.com wrote:
> Hi all,
> 
> I'm interested in any suggestions or information on implementing a PID 
> controller in an FPGA.  This should be possible without the use of a 
> processor.  I know there is a way to do the integration using a 
> summation, but what about the differentiation?

Typically, differentiation is estimated with a simple difference between
the present error and the previous error. Depending on how noisy your
feedback signal is, you may not get good results from the
differentiation process.

One thing to watch out for when implementing the integrator is windup.
When implementing the summation, you will need a way to limit the value
of the integrator. If you can find an adder block that will let you do
saturation arithmetic, use it. There are other schemes out there;
searching the web for "integrator windup" should yield some helpful
suggestions.

-- 
Charles Lepple <clepple@ghz.cc>

--
To unsubscribe from cores mailing list please visit http://www.opencores.org/mailinglists.shtml