## News:

SMF for DIYStompboxes.com!

## All transistor Rat

Started by soggybag, July 09, 2022, 01:02:00 PM

#### antonis

Quote from: Vivek on July 15, 2022, 09:23:47 AM
Can we put big cap on wiper so no DC on wiper ?

Of course but IMHO it shouldn't be essential..
(after all, 1M pot isn't placed inside signal axis..)
"I'm getting older while being taught all the time" Solon the Athenian..
"I don't mind  being taught all the time but I do mind a lot getting old" Antonis the Thessalonian..

#### soggybag

Quote from: antonis on July 15, 2022, 08:48:51 AM
JUst another mode for restless Mitchell..

A band-pass zone variable from 340Hz - 15kHz..(you can make it as narrow as you like..)
A 22k stopper resistor in series with 1M pot should make its full rotation more usefull..

(a couple of μV DC across 1M pot shouldn't make it crackle..)

What's going on here? Is that a low pass filter R12+C9 followed by a high pass filter C8+1M pot?

R12 and C9 form a low pass of 16kHz = 1 / (3000 * 0.0000000033 * 6.28)

C8 and 1M pot form a high pass. 338Hz = 1 / (1000000 * 0.000000000470 * 6.28)

The 22k resistor limits the high pass filter to 16kHz = 1 / (22000 * 0.000000000470 * 6.28)

I'm starting to get the math! I need more practice problems to practice!

So what's happening here? Looks like two filters, one cuts lows below the threshold while the the rolls offf highs above the threshold?

#### antonis

Quote from: soggybag on July 15, 2022, 03:52:43 PM
R12 and C9 form a low pass of 16kHz = 1 / (3000 * 0.0000000033 * 6.28)
Yes, buy only for CUT pot set FCW..

Actually, they form a LPF of 468Hz (pot FCCW) to 16kHz (pot FCW)

Quote from: soggybag on July 15, 2022, 03:52:43 PM
C8 and 1M pot form a high pass. 338Hz = 1 / (1000000 * 0.000000000470 * 6.28)
The 22k resistor limits the high pass filter to 16kHz = 1 / (22000 * 0.000000000470 * 6.28)
15.5kHz to be more precice but yes..
(you've got the point..)

So you have a band-pass filter of variable Q (abusive use here but study link below..)
https://en.wikipedia.org/wiki/Q_factor
"I'm getting older while being taught all the time" Solon the Athenian..
"I don't mind  being taught all the time but I do mind a lot getting old" Antonis the Thessalonian..

#### antonis

Next challenge will be to place one of hard clipping diodes into Q1 NFB loop to mix soft and hard clipping configuration..
(I let you decide for proper diodes orientation for maintaining symmetrical clipping..)
"I'm getting older while being taught all the time" Solon the Athenian..
"I don't mind  being taught all the time but I do mind a lot getting old" Antonis the Thessalonian..

#### Vivek

If that 1M pot is set with wiper to pin 3,

Instead of being a filter

It's just a short to ground

Ie FET gets no signal

Maybe a resistor in series with the pot to set minimum resistance of that leg would be a good idea.

#### antonis

#45
Quote from: Vivek on July 15, 2022, 06:52:58 PM
If that 1M pot is set with wiper to pin 3,
Instead of being a filter
It's just a short to ground
Ie FET gets no signal

As long as it's wired as variable resistor, it makes absolutely NO difference for "grounding" Q3 Gate..
(it just sets resistance change in relation with rotation direction..)

Quote from: Vivek on July 15, 2022, 06:52:58 PM
Maybe a resistor in series with the pot to set minimum resistance of that leg would be a good idea.

Already suggested a 22k "stopper" resistor above..

P.S.
What I'd expect is an objection concerning one diode placed inside Q1 NFB loop in relation with input impedance..
"I'm getting older while being taught all the time" Solon the Athenian..
"I don't mind  being taught all the time but I do mind a lot getting old" Antonis the Thessalonian..

#### D_Ex_Patria

What's the impact of dropping to the resistor from Q3 gate to ground to 22k on Q3? Is it solely a highpass with C8 or will it have some other effect on the operation of Q3?

I ask, because it seems like it would make a fantastic variable highpass in earlier stages of other circuits. Like ye olde FAC controls on older orange amps.

#### antonis

Quote from: D_Ex_Patria on July 16, 2022, 11:23:08 AM
What's the impact of dropping to the resistor from Q3 gate to ground to 22k on Q3? Is it solely a highpass with C8 or will it have some other effect on the operation of Q3?

Just a HPF..

Secondary (and negligible) impact is about 45 times lower Gate to GND voltage difference, compared to the one for 1M..
"I'm getting older while being taught all the time" Solon the Athenian..
"I don't mind  being taught all the time but I do mind a lot getting old" Antonis the Thessalonian..

#### soggybag

I put the MOSFet version on the breadboard. It works pretty well. Here's a few thoughts.

- over all output is not too far above unity at max volume.
- the range of gain is as great as the Rat. Though I don't have a real Rat to compare with at the moment. Just my impression.
- the voice is very suited to rock guitar.

I tried the  470p and 1M + 22k bass roll off. This worked but the volume drops significantly.

Seems like I could  rearrange Q2 to get a volume boost at the output. Looks like the current configuration is a source follower which is just under unity gain.

I'm going to breadboard the transistor op-amp version also and see what that sounds like.

#### antonis

Quote from: soggybag on July 20, 2022, 12:51:20 AM
- over all output is not too far above unity at max volume.

You're free to put the blame on D1, D2 pair..
(and secondarily on LPF..)

Quote from: soggybag on July 20, 2022, 12:51:20 AM
I tried the  470p and 1M + 22k bass roll off. This worked but the volume drops significantly.

But of course..
(that's the way ALL passive filters work..)
Just consider CUT pot FCCW and 1M pot FCW (only 22k stopper resistor between Gate and GND)..
You have a voltage division of 22/125 = 0.176..
By "adding" Q3 voltage gain of about 0.9 we result into 0.158 X D1/D2 forward voltage drop, which hardly is 100mV on Q3 out..
(it actually is even lower due to C8 & C9 voltage dividing effect..)

Quote from: soggybag on July 20, 2022, 12:51:20 AM
Seems like I could  rearrange Q2 to get a volume boost at the output.

You can try Jack's boosters http://www.muzique.com/lab/boost.htm or some kind of BJT booster (like Big Muff Π last stage..)
"I'm getting older while being taught all the time" Solon the Athenian..
"I don't mind  being taught all the time but I do mind a lot getting old" Antonis the Thessalonian..

#### soggybag

I got back to this project after a break. It was sounding alright but the volume was a little low. The original J201 follower provided about unity gain on the end, if I understand correctly. I replaced this with a second MOSFET with a gain of ~x2.

I can adjust the volume of the output with R14. It also inverts the signal to get it back in phase, so that's a small improvement.

I used BS170 for Q1 but ran out of these. I had some 2N7000 types and used one of these for Q2. The internet says these are interchangeable but the pin out is reversed.

I added D5 to protect the MOSFET.

Q: what exactly is going on there? I saw this on other schematics. People were using 1n4001 9.1v zeners and LEDs? Should I add a diode like this on Q2?

#### antonis

Quote from: soggybag on September 18, 2022, 01:35:05 PM
I can adjust the volume of the output with R14.

You can adjust the GAIN, to be more precise..
From signal amplitude point of view, Gain & Volume are interchangeable only for undistorted signals..
(which is the case here for D1/D2 Si, Ge or Schottky diodes..)

Quote from: soggybag on September 18, 2022, 01:35:05 PM
Q: what exactly is going on there? I saw this on other schematics. People were using 1n4001 9.1v zeners and LEDs? Should I add a diode like this on Q2?

https://electronics.stackexchange.com/questions/573662/is-a-diode-on-a-mosfets-gate-source-really-necessary
"I'm getting older while being taught all the time" Solon the Athenian..
"I don't mind  being taught all the time but I do mind a lot getting old" Antonis the Thessalonian..

#### PRR

A simple diode protects one way not the other.

5V Zener is fine for small MOSFETS.

The Base-Emitter junction of any small BJT silicon transistor will work (~~7Vz) but it is awkward to figure the polarities.
• SUPPORTER

#### Steben

Never ever experimented with Mosfets... but is the gain of q1 stage at its max? If not, have higher gain and use double the amount of diodes. There you have volume. Classic ds1 mod....
• SUPPORTER
Rules apply only for those who are not allowed to break them

#### soggybag

I guess my strategy was something like what is happening in the Big Muff. The signal is clipped pretty heavily at the 0.7v diode threshold. The tone section brings the signal down more but the last stage has a gain of roughly x5.

This set up gives you some room to play with the volume control. There's a spot about 75% that is unity and some room to boost or reduce the output.

I'm guessing the clipping setup in the big muff clips similarly to diodes to ground giving each clipping section an output range of roughly +0.7v to -0.7v?

In my schematic Q2 is about x2 gain. Which gives a little more range for the volume control.

I'm wondering if I'm correct about this thinking about the "Haunting Mids Fuzz." Haunting Mids is basically a Big Muff without the tone and output transistor. It's really loud.

Looking at the schematic it seems Haunting Mids uses 3mm red LEDs for the first stage which the internet says have threshold of 1.58v and the second set of clipping diodes are 1n4742 types which look like 12v zener diodes. These diodes could explain the higher output.