Aldrin (Lunar Module) to Screw Driver transformation

[HamishR]

I have the bass and treble knobs at around 3.00, so pretty high. I forget which one sharpness is! I find it interesting how the bass knob adds so much much more gain in the Lunar Module than it does in the Screwdriver.

I guess my issue is that finding high enough gain Ge trannies is tricky. Recently I swapped the AC127 (hfe 126) I had in my Screwdriver for an MP38 (hfe 64). (I wanted to try the AC127 in a different pedal!)

I really like the sound with the lower gain Ge but it doesn't get very dirty with a Strat. I would be interested in trying a silicon tranny in that position to see what slightly higher gain might sound like - and the extra brightness is easy enough to tame. With a Les Paul or 335 the SD with the MP38 is a great low-gain OD - really, it's amazing. But I would like to try building one with that kind of tone but more gain, and seeing what silicon could do. The LM is all silicon, so it can obviously work. The only reservation I have with the LM is that it still sounds a bit like a fuzz rather than an OD, and it's the clarity of the SD that I really like. The bass strings have a fair bit more definition to them through the SD.

The Screwdriver is close to being exactly what I want. I guess that if the Screwdriver had the gain control of the Hybrid Drive rather than a Fuzz and Pre-drive knobs it would be what I want. Or a Hybrid Drive with the Bass control. So a four-knob pedal. The Hybrid Drive doesn't get fuzzy and lose low-end definition until the gain is quite high - higher than I need. Turning up the Fuzz on a SD does get, well, fuzzy. That's why I use the Pre-gain. But the Pre-gain doesn't go as far as I'd like.

And from there I would like to experiment with silicon transistors in place of the Ge - just because.

 

[Chuck D. Bones]

Sharpness is the Bass control. It seems to add more gain in the LM because the LM's 1st stage has more gain than the SD. The SD 1st stage has a 680 Ohm resistor in series with Q1's source lead. If you short that resistor out, you'll get about 10dB more gain when PREGAIN is dimed. Still won't be as much gain as the LM. See if that gets you where you want to go. I'd try that before changing Q3 from Ge to Si.

If you do end up swapping Ge for Si in the 3rd stage, the biasing of the 2nd & 3rd stages will be affected, but it should still work.

I'm pretty happy with my SD just the way it is. I have another dirt pedal in front of the SD, so I can always hit it harder if I want. But I would like to give the HFD a try. Just need to figure out the capacitor values. ?

 

[HamishR]

I think you're probably right. You have given me some ideas but you are right when you say that the Screwdriver is pretty much spot-on as it is. Thanks for taking the time to explain some things to me. I'm going to look for some more decent Ge trannies - not easy.

 

[Chuck D. Bones]

With Ge transistors, it's as much about leakage as it is about HFE. Too much or too little leakage and they don't bias correctly. I think the LM, SD and HFD are pretty forgiving in that regard, plus we can fiddle the resistors if necessary to get the desired collector currents. I put an AC127 from Small Bear in my SD, works well. I also picked up some 2N1306s from them. Decent gain and low leakage.

 

[HamishR]

I played my Screwdriver with the MP38 in it today - Les Paul and homebrew tweed - and damn it sounded good! I experimented with the gain/fuzz knob - which I usually leave on zero or close to it - and got some killer sounds. I think I have got it to sound even better than the Hybrid Drive now. It's weird but cool that swapping the AC127 out for the MP38 actually made it sound gutsier. Maybe it was just those particular transistors? Some of the MP38s I have are reasonably low leakage and while they're not super high gain they sound really nice. Low noise too.

I have some more AC127s on the way so I think another Screwdriver might be in my future.

 

[Chuck D. Bones]

As a point of interest, could you measure Q3's emitter, base & collector voltages for each transistor you install? Would be interesting to see what's going on. I built two: one on a Madbean Cosmopolitan board, and then when the Aldrin came out, build one on the Aldrin board. Used all the same parts on both. Here's what I measured:
Cosmo
E 517mV
B 618mV
C 3.51V
Vcc 9.15V
Fuzz pot 991 Ohms

Aldrin
E 483mV
B 573mV
C 4.02V
Vcc 8.95V
Fuzz pot 1070 Ohms

The Aldrin has a 100 Ohm resistor in series with the 1N5817, the Cosmo does not, which explains the 200mV difference in Vcc. I measured the FUZZ pots because sometimes pots have pretty loose tolerances.

The AC127 in the Cosmo is running 14% higher collector current, 517uA, compared to Q3 on the Aldrin board, 452uA. This could be entirely due to the difference in FUZZ pot resistances. They both sound pretty much the same to me.

 

[HamishR]

I'll check those soon. I just built another Screwdriver today (on vero) and used a 2N1308 hfe around 100, low leakage. The only changes I made to the circuit were trying a 100ohm resistor on the BS170 instead of the 680, and a 2n7 cap on the output instead of 3n3 - I had tried 1nF but it was actually too bright! I don't think the 2n7 is that much different from the 3n3 but who knows? I was surprised that the 1n was so bright - but then it could have been "Brand New Pedal" syndrome.

I only had the other SD I had built with the hfe 50 Russian Ge tranny to compare it with. Overall the SD I built today has a little more gain - probably mainly to do with the 100ohm resistor I guess - and a hair brighter. But that may be down to the 2N1308, which I have found to be quite bright sounding.

Both of the pedals sound great, but I think I like the newer one more. With the Les Paul it has a little more cut and definition and a bit more of a vocal sound - some would say "honk". The new one also sounds amazing with a Strat. Both sound huge with a Strat but the newer one has a little more twang, which I like. I would have to guess that most of the tonal differences are purely down to component variation, and probably has a lot to do with the 2N1308.

This is such a great design.

 

[Chuck D. Bones]

Another thing to try, if you want to reduce the low freq content at high PREGAIN settings, is to lower C2 from 100uF to 10uF or 22uF. Pretty sure that's what Skreddy does with the HFD.

 

[HamishR]

That makes sense. I might experiment.

Tried them again today with different guitars... The newer one seems to have a little more mids I guess - I wonder where they came from?? I think in future I might split the difference with the resistor on Pregain 3 and call it 220 or 330ohms. Apart from that i really can't think of anything I would change. It sounds amazing.

I've run out of BC109Cs. I assume I can sub BC549C?

 

[Chuck D. Bones]

That resistor in series with the PREGAIN pot sets the maximum gain available in the first stage. I'd only change it if the top end of the PREGAIN range is too high.

Q2 and Q3 form a modified Fuzz Face circuit. Q2 is operated at the edge of saturation. Different parts numbers could sound different and there is nothing in the data sheets to help us predict which transistor will sound good. The BC109 is a popular transistor to use in silicon Fuzz Face type circuits. I'm assuming that the pedal builders have a reason for that. The BC549 will definitely "work." You will have to determine empirically whether they sound good. You could also try other NPN high gain Si transistors: MPSA18, 2N5088, 2N5089, 2N5210...

 

[HamishR]

That's very helpful - thanks!

 

[HamishR]

One more thing - Just built another, this time with a BS170, BC549C and an AC127 (hfe around 126, low leakage) and it sounds almost identical to the one I built with a BS170, BC109C and 2N1308 (hfe around 100, low leakage). So the BC549C works exactly as I had hoped - I really can't hear a difference which was perhaps to be expected!

FWIW I did try a BC108C and it sounded terrible - thin and weak with hardly any dirt. And that's maybe a bit weird because the hfe of the BC108C and the BC549C was practically identical.

Anyway i have another incredible sounding Screwdriver now.

 

[Chuck D. Bones]

Strange that a BC108C would sound different from a BC109C. According to the data sheet, the only real difference between the two is the noise figure; BC109's are quieter. Maybe you got a bad one.

Where are you getting 2N1308s?

 

[HamishR]

I forget where I got the 2N1308s. Probably off Ebay - I bought a bag of them and they weren't too expensive but they're mostly quite good. I did test one with ridiculously high leakage but also one with an hfe of around 100 and zero leakage - is that even possible??

I was puzzled by the BC108C thing - but it definitely changed completely when I put the BC549C in there. The BC549C sounds great. So my builds indicate to me that yes the hfe of the Ge transistor does matter because the only one I have built which sounds different is the one with the 50hfe MP38. It still sounds wonderful but has less dirt. I should probably swap out the 680ohm resistor on the pregain for something lower to help offset it I guess.

Thanks for that tip! Lowering the value of the pregain tranny sounds great! It's so interesting listening to what happens when you use the two different gain knobs to get the dirt you want - the pregain knob gives you a tighter, more aggressive dirt while the gain knob gives you a more compressed, gentler dirt. So by varying which one you use you can get similar amounts of dirt but quite different levels of compression. And lowering the 680ohm transistor to 100ohm lets some more of that aggressive, honky sound out.

 

[Chuck D. Bones]

I have some 2N1307's & 2N1309's (PNP), they have very low leakage. So yeah, I believe it's possible. I'll keep looking on eBay for 2N1308s. The ones I've found there are almost as expensive as tested ones from Small Bear.

With the 680 Ohm resistor, the 1st stage gain only goes up to like 2.5 or 3x. That 100 Ohm resistor lets it go up to 7 or 8x. With that much gain, strong pickups can drive the MOSFET into saturation. Also, with the SHARPNESS turned all the way down, the MOSFET can drive the shit outta the 2nd stage. That's something that Marc Ahlfs changed on the HFD. He put a bigger resistor (10K vs 100 Ohms) after the SHARPNESS control to limit how hard the 1st stage can hit the 2nd stage. I do like the Screw Driver.

 

[HamishR]

So that explains why the HFD sounds so much gentler to me... I do like the HFD but like the Screwdriver more. FWIW I used a 220ohm resistor on the first gain stage in my latest Screwdriver. Sure it's slightly less gainy I guess but it's not super noticeable.

My other favourite drive for a Gibson is the Rockett 45 Caliber. Side by side with a Screwdriver they sound amazingly similar (at least the way I set them up) but the Screwdriver has a better feel, plus a little more fullness. It' surprising how similar they can sound given that they are so different circuit-wise. The Screwdriver is perhaps creamier and more touch responsive - more fun - but the tones are very similar.

Chuck you are a font of knowledge and I am very grateful for your input. I will never really understand what I am doing but I love building pedals and you are making it more fun.

Oh, and I'm pretty sure my 2N1308s came from a place in the UK. I bought them to build some BJF pedal. Skreddys are better!

 

[Chuck D. Bones]

I'm starting a new thread for the Hybrid Fuzz Driver. Bought one today and finished tracing it. And played thru it a bit too!

 

[MattG]

Only problem with building a Hybrid Drive is we don't know all of the capacitor values.

I got a Hybrid Drive pretty cheap, for the express purpose of pulling the SMD caps and measuring them!

First of all, on the one I have, Q1 is 2n5210, and Q2/Q3 are 2n5172. So no germanium transistors here!

I'm using the component references from the board itself:

C9 - 4962 pF
C7 - 61 pF or 1447 pF
C11 - 467 pF
C10 - 508 pF
C13 - 267 or 292 nF
C12 - 6070 or 6013 pF
C6 - 181 pF or 137 pF

I don't know how helpful that is... I'm using an el-cheapo "Multie-function Tester - TC1" for measuring capacitor values. Where I only listed one value above, multiple readings were very close.

C7 is tricky - that's when I was still trying to figure out how to actually measure these tiny SMD caps. I don't know why I got two radically different readings. And then the cap itself went flying (could be in the dog's food, could have gone down a vent, could have simply not been able to see it on the floor, etc). But all the others I saved and taped to a piece of paper. So if you have any tips on getting better measurements, I'm all ears.

Since my transistors are different, I'll list my resistor values - these I just inferred from reading the number codes on the resistors themselves:

R1 - 2k (Edit: accidentally left this out in original post)
R2 - unknown/not labeled (Edit: accidentally left this out in original post)
R3 - 100k (on footswitch board)
R4 - 100R (on footswitch board)
R5 - 15k
R6 - unknown (not labeled, would have to de-solder and measure)
R7 - 100k
R8 - 10k
R9 - 56k
R10 - 5.1k
R11 - 5.1k
R12 - 75k
R13 - 100k (Edit: mistakenly typed 10k in original post)
R14 - 56k
R15 - 10k
R16 - 4.7k

FYI, here's another thread with pics of the Hybrid Overdrive PCB.

Let me know if this is enough of a hint to complete the Hybrid Drive schematic, or if I can do anything else!

 

[Chuck D. Bones]

Matt,
Thanks for doing this! Can you take some better board photos of the main board so I don't have to guess how the parts are hooked up? Top & bottom so I can see the traces and read the silkscreen.

For now I'll assume they're the same as the photos I've collected.

You don't need to remove R6. Just measure the voltage across R6 with the power on and no signal.

Please recheck R13. It's 100K on the pix I have.

It's very odd that Q2 is not AC127. I wonder if the one you have was modd'ed. Do the solder joints on Q2 look any different from the others?

 

[MattG]

Ask and ye shall receive!

You're right about R13, it's "1003", 100k, I mis-typed in my earlier post (now corrected).

To me, the soldering on all three transistors looks "virgin". However, the resistors at the top of the PCB (R12, R14, R15) look like they may have been reworked - you can see all the flux residue in the pics, that's not from me, I didn't touch any of the resistors. The rest of the resistors (and the transistors), to me, look like clean, original soldering (though I still suspect they were hand-soldered, or at least done in DIY fashion, the soldering doesn't look as "pristine" as robot-soldered SMD, IMO).

On Skreddy's own Screwdriver page, it says, "Note: I use a low-gain silicon in place of the former germanium transistor; after a thorough review and revision of the circuit to dial it in perfectly it sounds identical to the former version and will be immune to temperature variations and will see lower failure rates." So it's possible he switched to silicon for all the pedals in the Screwdriver "family".

I'll need to resolder the caps before I can power this back up and measure the voltage across R6. I was kind of hoping to wait until we're all pretty certain on all component values. I might get new caps anyway, I don't know how many heat cycles they can take. (It seems strange to me to use such small SMD components, I'd at least use 1206-size, they are fairly easy to hand solder, there's plenty of room on the board.)

Anyway, on to the pics! Please let me know if you want me to take more or better photos of anything. (I just got a new phone, seems to take pretty good pics!)