Homely BSIABII vs. The New, Also Homely BSIABII

[sta63bmx]

First off, the BSIAB is the shiznit.  But will the real BSIAB please stand up?  I have two of them now.  Homely #1...



And Homely #2...



#1 is built just like the GGG layout with some 2n5457 and J201 fets.  #2 is built with ALL 2n5457 fets.  Here's the deal.  #1 has way less gain and less level than #2.  This was briefly discussed in this thread and the consensus was that I probably burned a FET or something.  Now I'm intrigued, because #1 is way more AC/DC and has a VERY different distortion character than #2.  Now I'm dying to know what's different.  I'll start by measuring the the voltages on the FETS and then I guess I can probe through the circuit until I find WHERE they start sounding different.

Here's my question: if I burned a FET soldering, what happened?  Did it just turn into a diode?  That would explain the lack of gain, I guess, since one stage would be dead. 

Here's my BURNING question: Can I replicate the heavy-handed soldering on another FET or could I just use a diode or...? 

The second one is more VH, the first is 100% AC/DC.  Man, these pedals both completely own.  But I like the first one best.  If it just had a clean MOSFET boost AFTER it, level problem solved!  I'll try to figure it out soon and keep you posted.  First pedal has a little rougher overdrive, more asymmetrical-sounding, kinda like using a distortion plus with just one diode to ground, that kind of one-way-clip sound.

[newbie builder]

first i'd try making the fets the same in both pedals

[stm]

I find it hard to believe that you fried a FET. I tend to think the particular combination of FETs and biasing voltages are responsible for your particular sound.  Maybe even one or more components with a value different than stated in the schematic--a happy accident.

In any case, it is interesting to investigate what's really going on.  As a first suggestion, you should guarantee that you can put back your original sound by measuring the drain and source voltages, and labeling each FET with a correlative number so you can put back the circuit just like i was before.

[sta63bmx]

A happy accident indeed!  I daresay that If I had these two pedals as the two sounds in a Jekyll & Hyde style pedal, my overdrive needs would be completely taken care of.  Thanks for the advice on the pedal.

1. Making The Fets The Same: I socketed the second pedal, so I can do that easy.  i was out of sockets when I built the first one, so it isn't socketed.  If I can't figure it out at all, I could desolder the FETs from #1 and use them in #2.  I'll just use the same kind in #2 and leave the originals in #1 for now.

2. Measuring Voltages: I will do this ASAP.

3. Components: One KNOWN component difference between the two pedals is that I used a 470pF ceramic in #2, and I used a silver mica of slightly lower value there in #1, I think.  Maybe like a 450pF.  It's a single in the pictures.  All my resistors are 1% metal films in each pedal, all measured individually right before I put them in.  Which means I'll go back and measure if I don't find a difference somewhere else, because my illiteracy can be stunning.

So here's my plan...

1. Make FETs the same in each pedal and measure drain and measure voltages, try them out.  If they still sound different...

2. Try to get the FETs biased the same in each pedal?  Q1 and Q4 both have their gates at 4.5V but Q1, Q3, and Q5 are handled differently.  Q5 can be adjusted with that trimpot.  Is the trimpot there on Q5 effectively changing the "voltage divider" formed by R16, Q5, and R15 to get a bigger or smaller voltage swing going to the tone stack?  I'm looking at the GGG schematic.  When I turn the trimpot too far one way, I run out of headroom and there is distortion there and I start to lose amplitude.  The other way, it stays clean, but I lose amplitude.  Q2 and Q4 seem to be doing the same thing, but I don't understand exactly what.  Are they holding the drains of Q1 and Q3 some amount above the gates or...?  I apologize for my crude understanding of electronics.  Hey, I'm a mechanical engineer, so it's a double whammy.  First of all, mechanical, second of all, engineering degrees earned in the computer age, which means zero practical knowledge of anything useful.

3. If the FETs can be biased the same and I have the same type and it still sounds different, pick through the circuit with an audio probe and listen for WHERE the two circuits start to sound different.  That might give some insight into who is responsible.  Then start measuring component values around that part of the circuit.

4. If all else fails, try the FETs from the first pedal in the second one.

Man, I've got to figure this one out.  I find it kind of funny that usually you're trying to clone something that's commercially produced, and you know exactly what sound you're going for and you usually have a schematic to try and get there.  Here, I'm trying to clone something where I don't even know what I did!  Seriously, if I could remake the first one, I'd build it with a clean boost afterwards to take care of the level drop.  Then it would be unreal.  The lower level is the only strike against it.  These two together in one pedal would completely own.

Both of these pedals sound better than any Tubescreamer. *mock horror*

[d95err]

You should not solder a FET at all. Always, always, always socket transistors!

[sta63bmx]

Yeah, it was the weekend, it was late, and I was out of sockets, so I just slammed them in there.  Believe me, I usually don't do that.   But once I get my wave soldering machine, I won't have to worry about burning parts anymore.

[sta63bmx]

Fries are up.  Angus is the first one built: low drive, low output level unless it's dimed, and 100% AC/DC sounding, unreal rhythm pedal.  Eddie is the second one, which is more like a "real" BSIAB with loads of gain and harmonics.



I used the same battery in each pedal and it was at ~ 8.4V.  Both times I had a jack inserted to get power, but the effect is bypassed, so the input is grounded.  The only marked change was the drain of Q5, and that's what you bias with the trimpot.  To bias both pedals, I just ran a 500 Hz sine wave into the pedal (not sure about P2P voltage, but low impedance and about "guitar loud") and then turned the trimpot until I got maximum volume.

I have to go fix my amps so I can play these pedals in the garage and try out any changes, but here's the voltage numbers.  Thoughts?