Little Jim - a Marshall inspired MOSFET distortion pedal design

[jonny.reckless]

This is the Little Jim distortion pedal. It's inspired by and named after Jim Marshall, and attempts to capture that bright, crunchy, raucous sound of a cranked Marshall amp, which is also deceptively complex and rich with harmonics and timbral movement.

Most of my guitar preamp designs for the last 30 years have used JFETs for the gain stages. I experimented with using MOSFETs for this one. They sound quite similar, but are a bit more aggressive and crunchy, although they do produce more hiss than their JFET counterparts. They are not subtle, but work quite well in this high gain application. The pedal is a bit noisy (hiss) as a result.

The circuit is pretty simple. It's 3 cascaded BS170 MOSFET common source gain stages. The input stage uses a J112 as a buffer, because the input capacitance of the BS170 is quite high and will suck the tone of a guitar if connected directly without some form of buffering. The tone control is basically a variable frequency treble roll off but boosts the bass which compensates for the bass cut in the first gain stage to avoid flubbiness and farting out.

I found that the bag of BS170s I bought from Tayda were all very closely matched for Vth so I didn't have to do any special MOSFET threshold matching, you may need to check before soldering your MOSFETs that they have a similar spec, as the tolerance on the datasheet is quite wide, but in practice, mine were all within about 150mV of each other.










Listen to it here:

[Steben]

killer sound.

(But I can't help to note once more the use of a jFET correctly followed by quite a lower impedance pot than usually)

[Mark Hammer]

I have to compliment you on not showing something until it is really complete.  Not just a drawing of an idea, but a finished product, including pics of a build.  Nicely done.  Like the name, too.  Hommage, but not outrageous.  Kudos.

[11-90-an]

I have to compliment you on not showing something until it is really complete.  Not just a drawing of an idea, but a finished product, including pics of a build.  Nicely done.  Like the name, too.  Hommage, but not outrageous.  Kudos.

And each build comes with a pcb, layout, and a  great video!

[Steben]

Jonny, what software do you use?

[cab42]

Sounds great.

A couple of questions:

In  the schematic C4 and C7 are labeled as NF with no values. In the photo C4 is 4.7n and C7 seems to be missing?

C1, C3, C5, C13 (10u) looks like tantalums in the photo. Is that correct?

I had some time tonight and did a quick vero layout. Do you mind me posting it?

[jonny.reckless]

Sounds great.
A couple of questions:
In  the schematic C4 and C7 are labeled as NF with no values. In the photo C4 is 4.7n and C7 seems to be missing?
C1, C3, C5, C13 (10u) looks like tantalums in the photo. Is that correct?
I had some time tonight and did a quick vero layout. Do you mind me posting it?

NF means "not fitted" - I removed C4 during testing as I preferred the more aggressive sound without it. You could put it back and it softens the high frequency a bit and reduces the hiss very slightly. The 10uF caps are 5mm pitch multilayer ceramic. I generally don't use tantalum caps for anything any more. You could probably use tantalums or even radial electrolytics in there if you like, I doubt it will affect the tone very much.

Happy for you to share your vero layout, thanks for doing that. My vero prototypes are always a total mess (and tend to oscillate at high gain settings) so I don't generally share layout until I have a working PCB

[jonny.reckless]

Jonny, what software do you use?

Lots
For EDA, I've used Protel (now Altium) since the mid 90s when it was on Windows NT. It's too expensive to buy for hobbyist use but I have an old license tied to a (now defunct) startup I used to work for a few years ago. I use OBS Studio to capture the videos, and Vegas Movie Studio to edit them. I use LT spice for simulation, Daqarta and ARTA for various audio testing. I don't work as an EE these days - my day job is software engineering - so I can't afford expensive test equipment or CAD software any more

[deadastronaut]

love it..great sound man right up my street...nice work. 

[Phend]

Not knowing about the ins and outs of circuit design, curiosity question, where does the hiss originate in this mosfet design?

[PRR]

> where does the hiss originate in this mosfet design?

http://lamp.tu-graz.ac.at/~hadley/psd/problems/fetnoise/s.pdf
http://web.mit.edu/klund/www/papers/UNP_noise.pdf
https://www.nikhef.nl/~jds/vlsi/noise/sansen.pdf

[Steben]

> where does the hiss originate in this mosfet design?

http://lamp.tu-graz.ac.at/~hadley/psd/problems/fetnoise/s.pdf
http://web.mit.edu/klund/www/papers/UNP_noise.pdf
https://www.nikhef.nl/~jds/vlsi/noise/sansen.pdf

I'm going to be noise silent or some will call me OCD

[rutgerv]

Great work, and what an awesome demo as well. A rockin' sound (though always difficult to judge in the context of other factors determining the sounds (e.g. microphone type and position, cabinet type, power amp)).

I'd like to try Little Jim on a breadboard, but was wondering about other types of Mosfets. How critical is it to use a BS170? I remember having BSS138 and 2N2007 mosfet in my parts collection, but would have to order BS170.

Rutger

[rutgerv]

Looking at the datasheet, 2n7002 seems to have a comparable Rds On, but a much lower drain current rating (which I assume you're not needing here).

[jonny.reckless]

Looking at the datasheet, 2n7002 seems to have a comparable Rds On, but a much lower drain current rating (which I assume you're not needing here).

I only had BS170 and 2N7000 in my parts bin. I found they sound pretty similar, although the 2N7000 was slightly worse for hiss. Any small signal N channel MOSFET with a Vth less than 3V should work OK, you just might need to adjust the VBIAS to get the drains to sit around 2/3 to 3/4 of VDD. If you look at the drain signal on a scope, you should see asymmetrical clipping with the top half cycle rounded and fat, and the bottom half cycle clipped pretty hard.

You do need 3 MOSFETs with similar Vth, ideally around 2V or so. I found the bag I bought from Tayda were all closely matched. I guess you could just test them in a component tester to select for similar threshold to keep the bias points consistent. Maybe even different MOSFETs in each position might sound cool.

[jonny.reckless]

A rockin' sound (though always difficult to judge in the context of other factors determining the sounds (e.g. microphone type and position, cabinet type, power amp)).
Rutger

Signal chain for the demo video was Guitar -> Little Jim -> Hall of fame 2 reverb -> Marshall JCM800 50W head set to completely clean -> Marshall 1960A 4x12 -> pair of large diaphragm condensers mics in XY capturing the room sound. It's a fairly accurate representation of how it sounded in my shed

[rutgerv]

Hi Jonny,

thanks for the elaborate answers! Will let you know how my experiments work out with other MOSFET types.

One other thing I was wondering about: the 100 ohm resistor (R15) in the PSU input stage, does it serve any purpose for the dynamic response (sagging) of the rest of the circuit? Or is it purely for keeping out the PSU noise?

Rutger

[tonyharker]

BS170 transistors frequently have a suffix eg -D27Z. What is its significance and does it matter in your circuit?

[PRR]

> BS170 and 2N7000 .......... slightly worse for hiss.

I *suspect* (you may know better) that between MOSFETs of similar scale (die size, active area), the difference in hiss may be more about when and where they were made, not what they are marked as.

Hiss in most MOS applications is not a "problem" except it indirectly indicates contamination ("dirty Silicon") and leakage. The industrial buyers won't pay the price for "perfectly clean". (You and I and all of us don't buy enough to matter.) So, like fish-fry oil, one day it is fresh and clean, next week not-so-clean, and maybe clean again the day after the manager says "Ugh..."

[jonny.reckless]

BS170 transistors frequently have a suffix eg -D27Z. What is its significance and does it matter in your circuit?

I believe it is just to describe the packing style and pin spacing - ammo, reel, 0.08 vs 0.1 pin spacing?

Threshold voltage is the most important property in getting the circuit to bias correctly. I have one of those cheap component testers you see on eBay and measure threshold voltage of MOSFETs and JFETs before I install them into a circuit. They are amazing value for money, it's worth having one:
https://www.ebay.com/itm/12864-Mega328-ESR-Transistor-Resistor-Diode-Capacitor-Mosfet-Inductance-Tester/322883932945?_trkparms=aid%3D1110006%26algo%3DHOMESPLICE.SIM%26ao%3D1%26asc%3D20131231084308%26meid%3Dd7f8392b48ee408797cf2f6f659dedca%26pid%3D100010%26rk%3D1%26rkt%3D12%26sd%3D262474886090%26itm%3D322883932945%26pmt%3D1%26noa%3D1%26pg%3D2047675%26algv%3DDefaultOrganic&_trksid=p2047675.c100010.m2109