Low Gain Silicon Transistors

[MobyOctopad]

Maybe look into something like the 2N3705s, gain between 50-150 hfe according to this Mouser datasheet

 

[MobyOctopad]

I forget which suffix is the lowest gain of the Soviet KT315 (I believe КТ315И, but I'll double check when I'm home) but it's an awesome and cheap low hfe transistor. Weird pins but easy enough to finagle. I have a big ol' section of my drawer bins for low Hfe SI and I'll check to see if any I have haven't already been mentioned here.

I picked up a small batch of these from one of the forum's esteemed Ukrainian eBay sellers but stashed them because I thought they'd be difficult to use on a PCB. Was it too much of an issue dealing with their leads?

 

[finebyfine]

I forget which suffix is the lowest gain of the Soviet KT315 (I believe КТ315И, but I'll double check when I'm home) but it's an awesome and cheap low hfe transistor. Weird pins but easy enough to finagle. I have a big ol' section of my drawer bins for low Hfe SI and I'll check to see if any I have haven't already been mentioned here.

I checked and it is the yellow КТ315И (sometimes has the suffix “D” in latin). Mine measure from 50-80. The J suffix (orange) measure around 140.

Also in my bins are KSP06TAs from Tayda that measure around 95. Most others I have on hand have been mentioned

I picked up a small batch of these from one of the forum's esteemed Ukrainian eBay sellers but stashed them because I thought they'd be difficult to use on a PCB. Was it too much of an issue dealing with their leads?

I’ve had the best luck with them by using rectangular female header pins and socketing them. I’ve tried to slim the leads down enough to fit in round sip sockets with flush cutters but it didn’t work and wasn’t easy to boot. Might work to just slam them with pliers too now that I think of it.

 

[Chuck D. Bones]

Look for high-voltage transistors. They tend to have lower HFE. I picked up some 2N6517's at Electronic Goldmine. Mouser has 'em for about the same price.

 

[finebyfine]

Here's some more part numbers that are binned by Hfe, ranges pulled from datasheets. last letter here is the bin
- 2SC1815-O 70-140
- S9013-D 64-91
- S8050-B 64-91
- 2SA1020-O 70-140
- 2SC1384-Q 85-170
- KTC3202-O 70-140
- A44-A 80-100
- A42-A 80-100
- 2SC945R 80-100

 

[Chuck D. Bones]

The datasheet values indicate the extreme bounds for every part that rolls off of the manufacturing line. Even the "typical" values may not reflect the HFE in a pedal circuit unless the datasheet HFE is measured at the same Ic as the transistor will see in-circuit.

The datasheets can provide some rough guidance on which parts might work well in a given pedal circuit. Ultimately, the builder will need to audition parts.

You can also try the power transistors in TO-39 packages. 2N5154 is a goodie, but apparently everyone selling them on eBay has figured that out.

 

[finebyfine]

The datasheets can provide some rough guidance on which parts might work well in a given pedal circuit. Ultimately, the builder will need to audition parts.

I'd say I'm always surprised at the Hfe ranges that sound good to my ears when I'm socketing transistors, even in circuits that are more dependent on variations. Honestly, it's such a fun thing about pedal building to me.

 

[t1redhands]

Great timing - I got the itch to try some lower (than typical) gain transistors for my Big Muff builds.
My main one is a GGG-tuned BMP (VRH variant), which was running 2N5089s. Just a ridiculous 'thick wall of lightning' fuzz.

I binge-bought some BC109Bs (Smallbear) and BC550Bs (Tayda); the former measured between 200 and 350-ish, while the latter were almost ALL 400-450. I tried the 200 hfe in my BMP and I'm hooked. It can still get that big fuzz sound with the sustain maxed, but everything lower is more articulate.
I have some BC549As and BC550As en route, and will be sorting those out ASAP.

I LOVE using low-HFE transistors in a Big Muff. It gets the thick wall of doom sound I'm a sucker for as well as actually usable low gain tones. Try giving it only 5V.

The KT*** silicon transistors make great partners with the plentiful low-gain Soviet germanium PNPs for Sziklai pairs.

 

[CR0SSBL4DE]

Take into consideration that the Bipolar transistor hFE (current gain) it's not the actual gain of the amplification stage in the circuit. That will be determined by the topology and the feedback used. Is that feedback the one that makes the huge variations in hFE usable in real world products.

For example, the first stage of a big muff pi is a common emitter and has 2 FB resistors. One is 470K between base and collector, and the other is 100ohm, between emitter and ground.
In that stage, if you place a BJT with hFE=100, the voltage gain of the stage will be about -4.66 (13,37dB). But if you place a BJT with hFE=400, the voltage gain of the stage is now -6 (15,5dB).
So a transistor with 300% more AC current gain, will only result in a 29% AC voltage gain increase on the circuit.
(you can play with values here, if want to check http://guitarscience.net/calcs/cecbemf.htm)

So that's why low gain bipolar transistors are not common. When adding feedback to get consistent results (and also to make it's response a bit more lineal), you actually get a lot less gain in the circuit.

 

[CR0SSBL4DE]

Take into consideration that the Bipolar transistor hFE (current gain) it's not the actual gain of the amplification stage in the circuit. That will be determined by the topology and the feedback used. Is that feedback the one that makes the huge variations in hFE usable in real world products.

For example, the first stage of a big muff pi is a common emitter and has 2 FB resistors. One is 470K between base and collector, and the other is 100ohm, between emitter and ground.
In that stage, if you place a BJT with hFE=100, the voltage gain of the stage will be about -4.66 (13,37dB). But if you place a BJT with hFE=400, the voltage gain of the stage is now -6 (15,5dB).
So a transistor with 400% more AC current gain, will only result in a 29% AC voltage gain increase on the circuit.
(you can play with values here, if want to check http://guitarscience.net/calcs/cecbemf.htm)

So that's why low gain bipolar transistors are not common. When adding feedback to get consistent results (and also to make it's response a bit more lineal), you actually get a lot less gain in the circuit.

That said, if I were to play with a big muff to modify it's sound, instead of using low gain transistors to modify the gain staging of the circuit, I would modify the resistors values around the transistors to reduce each stage gain, but keeping the DC bias the same.

 

[Chuck D. Bones]

Increasing the emitter resistors is a good way to mellow-out a BMP.
Changing HFE or any of the feedback resistors changes the input impedance of a given stage, which affects freq response, linearity and how it loads the previous stage. Bias point affects all of that as well. It gets complicated.

 

[Ginsly]

Very, very glad I posted this - lots of great info, thanks everybody! As a VERY new builder I sometimes feel pretty dumb posting these questions, but once in a while it's nice to see that it's something others have been wondering about.

I can't afford much at the moment, but I'll eventually have to check out some of the places I'm less familiar with for these low-gainers...

So Electronics Goldmine is generally worthwhile? They been mentioned a couple times here. Do people still go through Futurlec at all?

Antique Electronic Supply seems to be in the same realm as Small Bear and Stomp Box Parts, but mostly for tested/curated transistors. I'm sure it's worth it, but I kind of need to find some low-gain Si's on the cheap. Lots of leads in this thread, but I'll have to see about sourcing...

 

[jessenator]

In my BJT addiction collecting almost specifically for the Super-Fuzz, these are ones that measure around 100 and below on my T7 and have been used in my Super-Fuzz builds:

part num	type	mat	min	mostly	max
MPS6548	NPN	Si	30	50	60
A5T2243	NPN	Si	30	50	70
2N2369A	NPN	Si	30	90	90
MPS3394	NPN	Si	60	70	90
MPS4274	NPN	Si	70		90
MPS6512	NPN	Si	30	50-80	90
2N2369A	NPN	Si	60	90	100
MPS5131	NPN	Si	20		100

Occasionally I'll get lucky with NOS 2N3903s, but the ones I have still hover over 100.

If you want some unsolicited combinations I've tried and a one-word comment

Q1—Q3 should have the highest current gain (anywhere from 140-200 is approximately OE); some builds suggest these be in descending order of gain.
Q4 & Q5 are sort of the """secret""" to this pedal's tone, and are typically matched as closely as possible (anywhere from 60-140). However, the bias trimmer is meant for you to tune them to get the desired effect (more on that later).
Q6 should be as low as possible, around 50 being typical for original units. Some older silicon can get down this low (e.g. MPS3394, 5AT2243) but you can piggyback if necessary/desired.

Your project schematic might be drawn differently, so your mileage may vary.

I've got to retract one recommendation here:

The MPS6548 is a VHF/UHF specialty application type and has a weird pinout of B E C—yes, the emitter pin is in the middle. So, don't buy it unless you want to bend pins for funsies.

 

[Ginsly]

The MPS6548 is a VHF/UHF specialty application type and has a weird pinout of B E C—yes, the emitter pin is in the middle. So, don't buy it unless you want to bend pins for funsies.

Thanks for the heads up! Ha, I also checked out the KT315 @finebyfine and @MobyOctopad mentioned... yeesh! Thing looks like a thumb piano!

 

[Chuck D. Bones]

EG has 2N3903s on sale. Get 'em before the sale expires or they run out.

Antique Electronic Supply is very pricey IMO. Same goes for SB. Gotta shop around. There are deals to be found on eBay, but the usual caveats apply.

 

[Ginsly]

EG has 2N3903s on sale. Get 'em before the sale expires or they run out.

Antique Electronic Supply is very pricey IMO. Same goes for SB. Gotta shop around. There are deals to be found on eBay, but the usual caveats apply.

Right on! Yeah... I was thinking about AES because they have a couple pots that Tayda was recently out of, and a couple other specialty items. The basic transistors didn't seem too pricey, other than NOS stuff... I'll take another look around, though - thanks for the advice. Gonna be a minute till I can make any moves anyways.

 

[Zanshin]

I picked up some 2N3903s from EG (thanks @Chuck D. Bones !) and some MPS6548s - I don’t mind a little twisting…

The KT315 is intriguing. Is it possible to twist them to fit TO-92 shape?

 

[Guilherme Collateral]

2N2369 is a pretty low gain silicon transistor, and so is 2SC828