capacitance multiplier and Varicap diodes

[iainpunk]

so i was looking around on some websites and found out about varicap diodes! it is the coolest thing i ever heard of.
after 2 second of looking in to it, i found out that most of em have very small capacitances, most less than 100pF.
the first solution that came up in my head was capacitance multipliers. with voltage controlled, relatively large capacitance, the possibilities are endless.

https://en.wikipedia.org/wiki/Capacitance_multiplier

does anyone know what i can build with variable capacitors? i accidentally did an impulse purchase without thinking it through that well

thanks in advance,
Iain

[amptramp]

This kind of capacitance multiplier used to be popular in power supplies before three-terminal regulators took over.  The capacitance from base to ground was multiplied by the hfe + 1 of the transistor meaning a 100 µF capacitor connected to a transistor with an hfe of 50 would give you the effect of a 5050 µF capacitor.  It would also multiply the effect of the ESR, so you may not get something all that useful.  BTW some varicaps are designed for AM radio tuning and they are often 365 pF and sections of this varicap can be paralleled.

You could use this with some limitations but another choice would be Zener diodes operated below their conduction voltages.  Some people have built tuneable audio filters using Zener diodes as the variable capacitance element and in some cases, the capacitance can be up to 10 nF.

[Rob Strand]

The cap multipliers make the capacitance (to ground) at the Vi terminals look bigger than the actual cap C in the circuit.  The Vo terminal is only used to understand the circuit.   The whole idea is to pull more current down the opamp resistance than down the sensing resistor to the cap, making the cap look bigger.   The sensing resistor means the cap isn't a true cap it has a  series resistance.   For the multiplying effect to work the opamp cannot clip.

Normally varicap diodes are used for tuning for RF oscillators/tuners.   In the old days you had a variable capacitor which physically moved the cap plates.    That can be replaced by a varicap diode.   You vary the voltage across the diode the capacitance changes then that changes the oscillator frequency.   For that you can use a pot divider to set the voltage.  The next step is to use an D to A and digitally set the voltage.  There's also ways to use vari-cap diodes in auto-tuners and tracking tuners.

The fact the *Bias* voltage changes the capacitance means the signal level also changes the capacitance and that means the signal gets distorted.   For an oscillator you can handle some distortion but for processing a signal the distortion might be an issue and the only way around that is to keep the voltage swing across the cap low.

In order to reduce distortion the varicap diodes often come as a dual diode and you bias the center point.

Here's some common connections,


For the capacitance multiplier you need might need to separate the DC biasing of the opamp from biasing of the Varicap.     The cap in the top two circuits is an extra cap which you can use to decouple the DC levels at the diode and opamp. (ignore inductors etc.)   Any bias resistors are going to add resistance in parallel with the cap.   You might need to AC couple that at well in the audio path.

A last point, if you want to maintain a large range of capacitance adjustment you need to keep stray capacitance low.

If the vari-cap is varied from 2pF to 100pF you have 50:1 range.  However if you have 10pF stray capacitance the range is 12pF to 110pF which is only 9.2:1 range.

The biggest limitation trying to use these for audio is the distortion.

[iainpunk]

hey guys, thanks for the information,

i ordered BB112, those are 17,5pF to 440pF, but i can put a few in parallel for less stray capacitance effects.

i'm not afraid of nonlinearity, i have plans for a fuzz/auto wah in one kind of pedal/synth module, with a build in triple LFO and more general weirdness.

cheers,
Iain

[duck_arse]

if you find an SBLsomething-type dual rectifier is a PC switchmode supply, you might be able to use that for a high C value varicap. I was messing with one in a thermin oscillator, it worked ok on the breadboard.

[iainpunk]

if you find an SBLsomething-type dual rectifier is a PC switchmode supply, you might be able to use that for a high C value varicap. I was messing with one in a thermin oscillator, it worked ok on the breadboard.

interesting, ill open up some of my PC psu's to check for it, but there are already 10 BB112's in their way from chinkistan

[Rob Strand]

interesting, ill open up some of my PC psu's to check for it, but there are already 10 BB112's in their way from chinkistan

FYI, there was a site where a guy measured the vari-cap behaviour of a whole heap of different diodes.  (I need to check-out duck_arse's suggestion too.)

[Slowpoke101]

FYI, there was a site where a guy measured the vari-cap behaviour of a whole heap of different diodes.  (I need to check-out duck_arse's suggestion too.)

Rob, have a quick look at Hans' work here;
http://www.hanssummers.com/varicap.html

It may be the site that you were thinking of.

[Rob Strand]

It may be the site that you were thinking of.

Pretty sure it is because it had categories of diode types and some scanned hand written notes - just like that site.