MID BOOST RC Filters

[looselectron]

Hi. I'm new here, finally jumping into DIY pedals as a hobby with a mid-booster kit which I bought in the Akihabara electronics district of Tokyo. It's basically a tube screamer without a gain control, FET switching and input/output buffers. However, I'm confused by an extra RC filter in parallel from the feedback (gain) resistor to ground that is not in the typical tube screamer circuit.

The filter is  comprised of a large electrolytic cap 22uF (C3) and a 3.3k (R2) resistor to ground, parallel to a 47nf cap and 1.5k resistor to ground (similar to TS). The corner frequency of this second RC circuit is VERY low (abt. 3Hz) just for filtering noise or does  it have tone shaping properties? Isn't 22uf a rather large value here? (Link to schematic below).

http://www31.ocn.ne.jp/~takagi/Diy/fx/image/02boost.gif

[Digital Larry]

The heart of the Tube Screamer is a non inverting op-amp configuration.  At VERY low frequencies the gain is just 1.

In general the gain is 1 + Rf/Rl where Rf is the feedback resistor from output to (-) input and Rl is from (-) input to ground.

So this other RC network makes the gain to go up to 47k/3.3k around the frequency you calculated, and the other smaller cap makes the gain go up further yet to 47k/(3.3k || 1.5k) around the frequency dictated by the smaller cap with whatever load it sees (which is starting to get a little complicated to figure out after Thanksgiving dinner).

So anyway, it looks like two shelves.  From "0" up to f1, it has a gain of 1.  At the first frequency it goes up a bit more (but flattens out) and then at the next frequency it goes up a bit more (then flattens out again).  Why have the first corner frequency so low?  Hard to say.

[Digital Larry]

Why have the first corner frequency so low?  Hard to say.

Sorry to reply to myself!  Since it's a kit, you have the opportunity to play with two separate low-rolloff points by adjustment of component values.  Whoever put the kit that you have together decided to have a really low cutoff.  But if you wanted to you could reduce the cap value to reduce the bass going through.

Also, I completely ignored the high frequency rolloff contributed by the cap in the feedback network.  But I figured you already knew about that.

[looselectron]

Thanks for the reply, Larry.

I put the kit away for a while, but decided to take it out again after buying a VHT Special 12/20RT that could use a good mid-boost.

Comparing this to a Tubescreamer again, I noticed another major difference. The high-pass filter in the TS's clipping-stage is an RC pair connected to GROUND on the schematic.

In the Mid Boost schematic (which did not come with the kit, BTW), both RC pairs are connected to the 4.5V source voltage tap. R1 + C2 are very similar to TS specs, and my guess is they form a 2.2 kHz low-pass filter -- perfect for a bright upper midrange boost.

I recalculated the other pairing of R2 + C3 to be 2.1 Hz high-pass filter.  Because of the polarity of the cap, looks to me like it's just keeping DC and ripple respectively out of the signal path.

What I don't understand is what happens when the RC networks are connected to 4.5V instead of ground? An active boost?

[aron]

> when the RC networks are connected to 4.5V instead of ground

Apparently no difference.

From a post here:

The trick is now that we "tell" the opamp that 4.5V is actually ground, so the opamp thinks it has a bipolar supply of + and - 4.5V. It operates around it's own local "ground" which is really located at 4.5V.

[looselectron]

Thanks, Aron. So then that means the clipping circuit is referencing 4.5V instead of ZERO volts?

More obviously, this is really the only return path from the clipping circuit to ground or supply voltage, right?  I guess without that, the clipping circuit won't work at all. Also, would my understanding be correct that the clipping circuit is also acting as the bias for the op-amp's (-) input?

 Actually, I just saw the same arrangement in the schematic for the Barber Silver LTD. The values are different but the topology of that part of the circuit is the same.