Compressor using a varactor diode in a capacitive voltage divider?

Hey guys,
I've had an idea I'd like your input on.

A while back there was a thread where someone was asking about passive compressor designs and more specifically possible compression using capacitors. In this thread Jakob hinted that there was no passive way of doing this, but that compression should be possible using varactors in an RF-circuit. Now that idea kind of stuck with me and I started looking into RF-modulation/demodulation and how varactors are integrated into oscillators tank circuits, but I couldn’t really find a way to make a compressor work in this domain, at least nothing, that was actually making use of these techniques.

Nevertheless when browsing through mic schematics and specifically pads in mic designs it hit me: shouldn’t it be possible to use a varactor in conjunction with another capacitor to form a capacitive voltage divider and achieve compression this way?

So I thought maybe something like this would get me started: a simple sidechain like Fred Forssell’s opto-sidechain with an added DC-offset in the end to lower the varactor’s capacitance. Then the control voltage has to be negative so that it subtracts from the positive DC-offset, decreasing the width of the depletion region and upping the capacitance. Since afaik the capacitive voltage divider is calculated as C1/(C1+C2), the AC input signal should be lowered when C2’s capacitance rises, which in this case would be formed by varactor and a bypass cap in series.

I attached a mockup of the basic idea in Eagle and just wanted to know if this would work in principle or if I'm just missing something. Is there a reason I've never seen this before?

Best
Jannis

I don't know anything about varactor diodes but from a quick check on Mouser it looks like they're < 100p. So the impedance of the load would have to be >10M ohms to get low-frequency audio. Maybe if you use a capacitance multiplier you can then drop the impedance and use two dual packages like BB201 which is ~4 * 89p.



In theory it could work. Not sure how practical it is if you can do the same or better with a JFET. But I like little theory circuits like this.

Thank you so much for your insight and the idea to use a capacitive mulitiplier!

Yes, I thought that might be a problem and was also thinking about using multiple varactors in parallel to overcome this. Also at the expense of gain reduction it would be possible to bias the varactor to a higher capacitance from the get go to, in the case of a BB112, around 200pf (it's ~25pF at >1V and ~450 at 8V). But since you'd have to have a lot of them in parallel that might not be the best choice. I've also found a high capacitance diode, the BA123, (1.6-2.4 nF) but that thing is just really expensive and you'd still need a few of them.

So the multiplier using a few diodes in parallel should be the way to go for now. I ordered a few BB112 to play around with and I'll see, where this takes me. Using this technique in an oscillator circuit might still be something worth looking into, since I have a hunch that the low capacitance might be less of a problem there.

But I'm glad that possibly poorer performance than other VCAs is the reason I've not seen this before and not that I was just heading down the wrong direction and this doesn't work at all. I was just curios how a simple circuit like this might behave

Best
Jannis

The capacitance varies with voltage. The audio voltage is directly on the varactor. Distortion will happen.

This is true of other devices, yes.

Rather than paralleling a thousand diodes to get a useful impedance, wrap a TL072 around the bridge so the varactor works into many-Megs and the TL072 can drive useful load.

I don't see ANY advantage over an ordinary diode bridge. But if it keeps you off the streets and out of trouble, why not?

I think we've discussed this before?

about that shunting a (pf-range) capacitor across signal works as an attenuator in condenser microphone front ends..

/Jakob E.

Hey everyone,
thank you so much for your replies!

PRR said:

The capacitance varies with voltage. The audio voltage is directly on the varactor. Distortion will happen.
This is true of other devices, yes.
[...]
I don't see ANY advantage over an ordinary diode bridge. But if it keeps you off the streets and out of trouble, why not?

Click to expand...

Haha, it definitely does! ;D
I thought it might not be particularly good performance wise, but since I'm really new to electronics, projects like these are just there to help me wrap my head around things. Creating kind of a "new" problem for myself motivates me to dig a little deeper and it keeps me interested.

PRR said:

Rather than paralleling a thousand diodes to get a useful impedance, wrap a TL072 around the bridge so the varactor works into many-Megs and the TL072 can drive useful load.

Click to expand...

That question outs me as a total beginner yet again, but still, I'll have to ask: by "wrapping around" do you mean putting it in the negative feedback path of the opamp?

gyraf said:

I think we've discussed this before?

about that shunting a (pf-range) capacitor across signal works as an attenuator in condenser microphone front ends..

/Jakob E.

Click to expand...

Yes, we have discussed this before, at least the part about me wanting to research into RF-circuits and varactors after you posted about varactors as possible VCAs in the aforementioned thread, when I came to your booth at Studioszene in September in Cologne. I remember that talk very fondly and it encouraged me even more to keep going with that idea, so thank you! 

That particular thought actually came to me when I was looking at the U87 pad circuit. That's why I thought this should work that way. But in microphones the problem of a rather high highpass filter is circumvented by using 500m-1g resistors after those small value capacitors (the capsule and the pad), correct?

Best
Jannis