Buffered Big Muff Input

[D_Ex_Patria]

I've been reading through older posts about the first stage in the Big Muff, and wanted to try putting in a low noise JFET buffer in the first stage to do two things: have more gain on hand and reduce the thermal noise caused by the ~39k series resistor (Russian values) at the input.

Here's the circuit I'm thinking of putting to board:



Decoupling cap values are a bit arbitrary here and it's missing power filtering caps, but the important bits for gain and bias are where I want them.

To get the roughly 10dB gain that happens in a stock Russian Muff with the new reduced source resistance, R2 is raised to 1.5k, and R5 is raised from the stock 100k to get the circuit to bias at about 4.5V depending on the transistor used. The switched cap on the emitter in Q1 changes the gain from about 10 dB to 30dB, which I think would approximate putting a boost in front of a stock Russian Muff. Like a Blunderbuss fuzz, but hopefully with less noise.

So questions for the more learned of you:

1) Are the R8 and R9 reasonable values for series resistors in this circuit? It seems to be a standard value in a lot of pedals, is it appropriate here?

2) Is the raised R5 going to hurt this in any way? It seems it would improve input impedance to that gain stage and the bias works out, but I continually see 100k values everywhere in Muffs.

thanks to Fancy Lime for humoring me and my messages. Also, my bias circuit and JFET buffer is borrowed from his Goat distortion.

[antonis]

Read: 2. Big Muff Pi Input Booster Stage.

https://www.electrosmash.com/big-muff-pi-analysis

[D_Ex_Patria]

Read: 2. Big Muff Pi Input Booster Stage.
https://www.electrosmash.com/big-muff-pi-analysis

I have read the ElectroSmash article. The series input resistor appears to be there increase input impedance and decrease the impact that different pickups might have on that first input stage. It also attenuates the signal, which in most cases will keep that first stage from clipping.

If I put in the JFET buffer in the schematic I posted, the input impedance of the whole circuit will increase to about the value of R3, and without the 39k series resistance dropped to 1k, there's less attenuation when it hits the gain stage.

Is there something specific you believe I'm missing here?

[antonis]

Is there something specific you believe I'm missing here?

Other than it will not be a Big Muff anymore..?? 

Yes, there is another (dual) thing..

Lowering R8 calue (39k) doesn't offer any bennefit 'cause:
1. Signal source output impedance is just JFETs 1/g_m (about 400R or so) and not guitar or previous effect output impedance..
2. Q1 stage Gain isn't just R4/R8 (ideal shunt-feedback configuration) due to open loop gain waaayyy lower than "infinite"..

What I mean is:
1. Voltage dividing effect difference should be considered negligible
2. Gain increase should also be considered negligible

P.S.
No theoretical argument about JFET buffer input impedance increase..

[D_Ex_Patria]

I'm very aware of that this is not a pure Muff.

I built my own Russian Muff not long ago, and I loved the sound I got when running an LPB1 in front of it... But it was a bit too hissy.

From what I understand, that 39k resistor is the source of a lot of that hiss. So my goal is reducing hiss and increasing gain. With C2 engaged, I may be close.

[POTL]

Maxon released Big Muff with input and output buffers. if I remember correctly, their new mini fuzz is made according to this scheme.

[iainpunk]

Maxon released Big Muff with input and output buffers. if I remember correctly, their new mini fuzz is made according to this scheme.

that schem. seems wrong, look at the first transistor, its sending steady 9v in to the next stage...

adding buffer to big muff is a bit redundant, since the input impedance is Hfe x Remitter so 300x1500=450kohm. that's reasonable to begin with.

cheers

[D_Ex_Patria]

Maxon released Big Muff with input and output buffers. if I remember correctly, their new mini fuzz is made according to this scheme.

that schem. seems wrong, look at the first transistor, its sending steady 9v in to the next stage...
cheers

You're right. That should be on the emitter. That's a good catch. That schematic's probably been up for a decade.

I think the bias arrangement of the first stage factors in for  input impedance. That 47k to ground in the original muff schematic is going to drop impedance significantly, isn't it?

[antonis]

I think the bias arrangement of the first stage factors in for  input impedance. That 47k to ground in the original muff schematic is going to drop impedance significantly, isn't it?

Let's make some things clear..

Input impedance (for particular stage) is whatever is seen when you sit on transistor Base and look around you searching for AC ground..

A remarkable deviation is feedback resistor (470k), the apparent value of which is divided with particular stage gain plus unity..

P.S.1
I'm not aware of any "original" big muff with 47k resistor to ground..
Could you plz elaborate..??

P.S.2
You said you've read Electrosmash article but I'm not sure you've studied it..

[KarenColumbo]

Why not put an LPB-1 before your muff? Those seem to like each other.

[D_Ex_Patria]

From section 2.3 of the ElectroSmash article:

The input resistance of the Input Booster due to the emitter resistance and the feedback network effect is much smaller than R2. So, the value of the input resistor R2=39K accounts for almost all of the signal loading at the input.  The 39K it is indeed a low input impedance...

The ElectroSmash example circuit uses a 47k at R14, and that value also shows up in several versions of the Ram's Head and V3 Muffs:



ElectroSmash also discusses the effects of INCREASING the value of that 39k, and does state that doing so will decrease available gain:

Increasing the 39k R2 input resistor, the input impedance is increased but it also forms a voltage divider at the input, reducing the available voltage gain.

That's at least one of the reasons I wanted to try to decrease that value in my modified version of that stage.

[iainpunk]

Why not put an LPB-1 before your muff? Those seem to like each other.

i would recommend a range master if you like a bit more definition and less ''sponge-like'' attack, if you like the big muff as-is, character wise, an LPB-1 is great for that.
i think having multiple boost pedals in front of the big muff is a good option as well, so you can change the feel based on which one you engage.

cheers

[D_Ex_Patria]

Why not put an LPB-1 before your muff? Those seem to like each other.

Good thinking. And I tried that. It's how I got into this whole mess. 

You're right; they DO like each other, and they sound great when I'm playing, BUT the hiss when I'm not playing is just a little too high. Either I learn to live with it, or I get a noise gate, or I try to find an elegant way around it. From previous posts by Fancy Lime and a few others, I've gleaned that the noise in your first stage will have a huge affect on the the noise in the whole circuit, so I'm trying to put my focus on that.

[antonis]

That's at least one of the reasons I wanted to try to decrease that value in my modified version of that stage.

Let's take it once more..

39k resistor should be considered in series with signal source output impedance (voltage source internal resistance)..

For a guitar pickup of 10k (say), input voltage dividing effect is:  47k//(470k/Gain +1)//120R*(h_FE+1) / [47k//(470k/Gain +1)//120R*(h_FE+1)] + (10k+39k)]

You can realize that the main dominator of input impedance is 470k apparent value (about 5k) in parallel with 120R apparent value (about 10k) resulting into about 3k3..

P.S.
You don't have to bother with 39k resistor value too much..

[PRR]

Maxon released Big Muff with input and output buffers. ...

Aside from the mis-draw noted; this *still* has the 39k series resistor at guitar level. Oh, it is now 33k and moved (we assume) behind a buffer, which amounts to the same as before. (If buffer gain is 0.91 then the hiss is unchanged.)

Since actual discrete design is hard, and this stage(s) are not the distortion, I would replace the two transistors and the gain control with a good opamp with gain control in the NFB loop. Millions of places to steal that from.

[POTL]

Maxon released Big Muff with input and output buffers. if I remember correctly, their new mini fuzz is made according to this scheme.

that schem. seems wrong, look at the first transistor, its sending steady 9v in to the next stage...

adding buffer to big muff is a bit redundant, since the input impedance is Hfe x Remitter so 300x1500=450kohm. that's reasonable to begin with.

cheers

there is definitely a mistake in the diagram, but it is so obvious that I decided not to talk about it.

[POTL]

Maxon released Big Muff with input and output buffers. ...

Aside from the mis-draw noted; this *still* has the 39k series resistor at guitar level. Oh, it is now 33k and moved (we assume) behind a buffer, which amounts to the same as before. (If buffer gain is 0.91 then the hiss is unchanged.)

Since actual discrete design is hard, and this stage(s) are not the distortion, I would replace the two transistors and the gain control with a good opamp with gain control in the NFB loop. Millions of places to steal that from.

do you suggest replacing the input and output buffers with one two-channel amplifier? I thought about it in other circuits, but it always seemed to me that the input and output of a high gain circuit within a single microcircuit was a bad idea, it seemed to me that there could be a lot of noise, hum and interference, from the close location of the input and output.

[ElectricDruid]

do you suggest replacing the input and output buffers with one two-channel amplifier? I thought about it in other circuits, but it always seemed to me that the input and output of a high gain circuit within a single microcircuit was a bad idea, it seemed to me that there could be a lot of noise, hum and interference, from the close location of the input and output.

I think you're right and putting both the input and output buffers on one chip in a high gain circuit like this could lead to problems. Not for certain, but definitely worth considering. But Paul wasn't suggesting that. He only suggested replacing that mess of transistors buffers followed by a passive 'sustain' control with one decent clean op-amp stage with an active gain control. We can leave the output alone.