How to vary Capacitance using a variable resistor 1

[kps]

I have to adjust the integral gain of a PID controller. One would normally use a capacitor decade box externally and select the optimum capacitor value for stable operation and solder that value ultimately.

In my case, I have to use a microcontroller to automatically tune and stabilise the system. I can only use digital potentiometers for changing the Capacitance value.

Any ideas ?

 

[buzzp]

I guess when I designed a PID controller, I varied the resistance instead. I suppose you need an analog type of selection? In other words, you can't pick 4-5 cap values and use these. This would make tweeking extremely difficult. If you wanted to use the same cap box then you could select them using a switch with serial comm's. Another option (expensive and ugly) is to use a stepper motor to physically turn the variable cap. This is not desirable if this is a production design. Good luck.

 

[MMAshwin]

Hi,

A couple of thoughts.

1. You could use a varactor diode (although the capcitance values are small)

http://www.americanmicrosemi.com/tutorials/varactor.htm

for a preview

2. How about an analog switch which selects different capacitors

http://dbserv.maxim-ic.com/tarticle/view_article.cfm?article_id=90

for am introduction

Any help ?, yes no let me know.

Regards

 

[cbarn24050]

Hi, the simplest way to achieve this is to let the micro do integration.An intergrator is basicaly just a counter with is updated every time interval by an amount proprtional to the signal size and polarity.

 

[kps]

Thank you guys

Varacter diodes I ruled out because of too low a capacitance value

With analog switches I can only select a range but i can't continuosly vary the integral gain within that range.

I had long ago built a capacitance multiplier. The miller capacitance effect which is normally a nuisance, was effectively used in that circuit and was multipied by an op-amp. I built a PI controller using that and it worked beautifully. I could simulate high capacitance values using smaller value capacitors and vary the value with a pot.

I do not have that circuit with me. Any body who has done similar things ?

kps

 

[nbucska]

Hi KPS:

There is a simple controlled cap circuit which can reduce
the effective value of C from the nominal:

The value of C = max. required. Uses 2 VF Voltage followers
Wire list (one node per line):

Input --- C.1 --- VF1 inp
Potmeter.hi --- VF1 outp.
Pot.low --- gnd
pot.wiper---VF2 input
VF2 output --- C.2
--------------------------------------

Op-amp integrator can be adjusted in wide range

Digital control is simple and cheap. NEC gives away
development SW.

<nbucska@pcperipherals.com>

 

[MMAshwin]

Hi again,

So you blew out my varactor & switched capacitance ideas :-( .

Try

http://www.hep.upenn.edu/SNO/daq/parts/lm310.pdf

On page 12 is a variable capactor circuit using the multiplier principle.

or for millar intergrators see

http://lucas.elo.utfsm.cl/elo_b/an-32-2.pdf

Page 1

If you have any trouble with these pdfs I can send them to you if you wish. Any help ?, yes no let me know.

Regards

 

[kps]

Thank you Laffalot (by the way, is that your name or you Laugh_a_lot ?) I could download those pdfs. The variable capacitance multiplier of lm310.pdf may be just the thing I was looking for. I hope Capacitor C1 could be treated as a floating capacitor and included in the feedback circuit of the PI regulator op-amp.

Thank you nbucska.
For certain reasons I am not supposed to use digital control. The controller part of it has to be purely analog. But I am allowed to set all the parameters of the analog controller using digital pots. The microcontroller is used to change these pot settings under software control.

In a simple PI controller with one resistor and one capacitor in the feedback path, the resistor is varied to vary the P-Gain and the capacitor is varied to vary the integrating time. That is the reason why I was looking for a circuit idea which would enable me to vary the capacitance value with a pot.

I am also interested in PID controller circuits with multiple op-amps where each parameter can be independently varied without any interaction with other parameter.

Thank you guys for being so helpful.

KPS

 

[tehrani]

Variactors come in verity of sizes, I have used them for control circuits successfuly in the past.
I usually have a main capacitor and have one or two or many parallel variactors with it and can control the overall value of the cap with a D/A converter connected to the variactors.