Gemini III dual fuzz dissected

[slacker]

Q7- (Mystery transistor, not sure of pinout. I'm going from left to right, looking at the flat side of the tranny. Need verification here.)
E=1.430v
B=0.694v
C=161.8mV

I simmed the Q7 circuit, using a variety of common NPNs, and get the following voltages

C = varies with device between approx 1.3 and 2 volts, higher gains give lower voltage.
B = approx 0.7 - 0.8 volts
E = approx 150 -160 mV

So basically what you've measured if you reverse E and C.  I'd just go with Keen's law with this tranny.

[digi2t]

Q7- (Mystery transistor, not sure of pinout. I'm going from left to right, looking at the flat side of the tranny. Need verification here.)
E=1.430v
B=0.694v
C=161.8mV

I simmed the Q7 circuit, using a variety of common NPNs, and get the following voltages

C = varies with device between approx 1.3 and 2 volts, higher gains give lower voltage.
B = approx 0.7 - 0.8 volts
E = approx 150 -160 mV

So basically what you've measured if you reverse E and C.  I'd just go with Keen's law with this tranny.

Ian.... I love you man!!!!

So, that means that I just have to flip the Q7 symbol on my schematic, right? It`s just a bit bizzarre for me, I`ve crossed paths with a tranny yet that has the E on the right, and C on the left, when you`re looking at the flat side with the legs down. To me, it`s always been the other way around. Maybe that`s what was throwing me off. Also probably why they sanded the numbers off this sucker, they didn`t want to give the baby away.

Ian, you`re a true scholar and gentleman. Thanks dude! Now I have to do some research and find out which tranny fits this bill, backwards pinout and such.

On another note, I scored a bag of 75 AC128`s and 75 AC176`s last week. So far the Atlas tester is giving me pretty good numbers on a majority of them. So, it`s going to be a toss up between them, or Russian trannies.

Cheers,
Dino

[dorrisant]

Yet another reason I didn't label those suckers before... just kidding!

Do feel free to label them and repost the trace if you can...

Tony

[slacker]

No problem Dino, happy to help figuring this beast out.
If your voltages are like this



then your schematic is fine, no need to flip anything. If the 1.4V and 160mV are switched then it's back to the drawing board on what the transistor is.
A lot of European transistors have the pinout you described, have a look at the datasheet for a BC547 or BC550 for example, they're high gain general purpose trannys, basically equivalent to your 2N5088/2N5089.

[Earthscum]

then your schematic is fine, no need to flip anything. If the 1.4V and 160mV are switched then it's back to the drawing board on what the transistor is.
A lot of European transistors have the pinout you described, have a look at the datasheet for a BC547 or BC550 for example, they're high gain general purpose trannys, basically equivalent to your 2N5088/2N5089.

Drive them hard, and BC's have a nice tone (to my ears). I used a BC108 (or 109?) in my Muff because of the way it spiced up the sound (into SS, anyways... can't vouch for tube). I've really only noticed the most tonal difference when they are getting pounded pretty good, though. I've kind of made a habit of putting MPSA13's in the front of my circuits and BC's at the end (or where they get driven the hardest with the least post-filtering). Just an idea, might be "flavoring" this circuit.

[digi2t]

No problem Dino, happy to help figuring this beast out.
If your voltages are like this



then your schematic is fine, no need to flip anything. If the 1.4V and 160mV are switched then it's back to the drawing board on what the transistor is.
A lot of European transistors have the pinout you described, have a look at the datasheet for a BC547 or BC550 for example, they're high gain general purpose trannys, basically equivalent to your 2N5088/2N5089.

Yup, that's it. It's a normal EBC configuration, like a 5088, etc. I think the fact that I measured from the backside before must have screwed me up (mirror image deal ). I think we've got that problem licked. This time I snuck a probe under the top board to the legs on the component side to take the voltages. That way I'm sure of the orientation. It's not a BCxxx or 2N2222A can package, just a normal silicon package. Does your sim allow to see what gain range gets us into the voltage range we have here?

Another thing I was thinking about, is ditching the hi/med/lo gain switches for pots. I'm going to have to sit down and figure out the parallel resistor values, and see if I can fake them with a pot. That would be great, allowing us to more precisely adjust the gain of either side.

@David - Right on the BC108/9 deal. I run everything into SS, and they do sound pretty good in my rig as well. Could be worth while trying these in here.

Geeeez.... I can't wait to get my other projects cleared off the bench, so I can start working on this this.

@dorrisant - I looked over the PCB layout, but I found that there is a .22uF cap missing before the transformer. Might want to have a look at that. Also, I think it's going to be a tough board to work with, I'm not sure that the transformer will fit onto the board. Maybe have a look at the dimension specs of the transformer. There is no connection point on the board for the P1 trimmer to connect to (going to ground). That's all I had time to look at today.

Thanks for your work guys. I'm sure we're going to lick this sucker.

[digi2t]

Just had an update from someone on the other forum who's breadboarding this right now.

There is an error on one of the resistor values. The resistor between B and C on Q3 is NOT 15K. It is 150K.

His wife actually confirmed the color code from the pics I posted. He said that it went from an overdrive, to a great fuzz when he swapped out the resistor. I went back into it, and this time I lifted the resistor to test, and sure enough, it reads 150K. The 220K on Q2 is good. I had to test these two, because the colors are just so hard to read on these 1/8 watters.

I've updated the schematic, and the correct value should appear shortly.

Sorry guys.  Guess an appointment with the opthamologist is in order.

[dorrisant]

Sorry guys.   Guess an appointment with the opthamologist is in order.

No problem, man... have another homebrew... uh... you probably know what I mean.

Good call for the mystery contributer for the good eye! Please pass along the gratitude!

I love the video you posted... I heard some mean sounds there that I want to try out...

My layout remains unaffected by the resistor change... until you guys tell me what I f@#$ed up.

...And look at how many people have viewed/are watching this thread! This is so worth finishing. We are almost there!

Tony

[slacker]

Does your sim allow to see what gain range gets us into the voltage range we have here?

Sim said a 2N5088 was in the right ballpark. I whacked the circuit snippet on the breadboard and tried it with a few 2N5088 in the 400-500 Hfe range and the voltages were spot on to what you measured and what the sim predicted. Looks like we have a winner

[digi2t]

Does your sim allow to see what gain range gets us into the voltage range we have here?

Sim said a 2N5088 was in the right ballpark. I whacked the circuit snippet on the breadboard and tried it with a few 2N5088 in the 400-500 Hfe range and the voltages were spot on to what you measured and what the sim predicted. Looks like we have a winner

Excellent news Ian. Thanks a bunch... again!!

I`ll add the info to the schematic. I`ll advise buddy who`s breadboarding it as to your findings as well. He`s waiting an a 2N3700 right now, but with a close sub, he says the circuit sounds really good.

Cheers,
Dino

[digi2t]

More news on the Q7 front.

One of our fellow DIY`ers who is testing has reported that the following transistors will work here (with a description of the tone).;

BC239C - agressive, but smooth
BC550C - cleaner tone
2N5088 - more agressive than the BC550, but clean

All three, in the 400 to 500 hfe range, give close to, or exactly the same, voltage readings. Testing any one of these after the circuit is build should seal the deal as to the tone prefered.

Also, it was found that the 2N2222A`s tend to match the voltages best, are at, or close to, 100 hfe.

Another note; The 18.5K and 308K resistors may be sub`d with 18K and 308K. Much easier to find, and shouldn`t make any noticable difference tone wise. If you`re real picky, buy a batch and sift through them, I`m pretty sure that you`ll get the occassional one that comes close.

Cheers,
Dino

[dorrisant]

Just to clarify... the pcb is not made to mount to the transformer. I was gonna remote mount it and wire it from there... just two wires...

I will add the cap and repost... good eyes! Anything else?

Tony

[puretube]

More news on the Q7 front.
...
BC239C - agressive, but smooth
B550C - cleaner tone
2N5088 - more agressive than the BC550, but clean
...

Cheers,
Dino

?
why?

[digi2t]

More news on the Q7 front.
...
BC239C - agressive, but smooth
BC550C - cleaner tone
2N5088 - more agressive than the BC550, but clean
...

Cheers,
Dino

?
why?

Why?... Because it is there  .

I must confess, I`m not aware of what he`s using for guitar and amp combo. I`m just repeating his findings for the sake of anyone else that might be working on this. Personally, I`m swamped in other projects right now, but in my opinion, the more info, the better.

@ dorrisant - I kind of figured as much on the the xformer. I`ll have another look at the layout over the weekend, and see if any other gremlins pop out at me. Thanks dude.

[digi2t]

Well... first and foremost.... I wish to thank, from the bottom of my soul, chaospere for all his hard work. His wife too, long story, but she deserves it as well. He's build the clone, and reports that it sounds amazing. A real keeper. Not only that, but he's put together a wonder build .pdf, including a PCB transfer, parts list, wiring guide, and hell... why not, tranny voltages.

I'm so happy, I could almost cry.... OK, get a grip Dino...

The build paper is here; http://www.aronnelson.com/gallery/main.php?g2_view=core.DownloadItem&g2_itemId=47579&g2_GALLERYSID=ace1fcc31b7d711f6135f011d3090860

To cut down on parts, and make the PCB a bit smaller, the Millenium bypass has been omitted, and the wiring is done in true bypass fashion. He also added a 1N400x protection diode, since we both agreed that original's use of a 1N4148 diode for protection was a bit weak.

There are only 215 of these pedals ever made, and word has it that the last batch of 50 made, were the last ever. I was freakin' lucky to get my hands on one, and now everyone can enjoy this beauty.

I'll eventually get down to doing a vero of this as well, for those of us that don't have the capacity to etch (your's truly included  ).

Cheers,
Dino

[Skruffyhound]

Nice project. Complete after just three pages. Way to go.
Props to Chaosphere (and his wife).
I'm making no promises about when, but next time I'm etching I'll let you know, for this or something else.
-Aston

[digi2t]

Nice project. Complete after just three pages. Way to go.
Props to Chaosphere (and his wife).
I'm making no promises about when, but next time I'm etching I'll let you know, for this or something else.
-Aston

Thanks Aston. I'd be happy to pay you to etch me a board, when you get the time. No rush. PM me if that's kosher with you.

Cheers,
Dino

[dorrisant]

Wow! Great job guys!... Gonna go etch...

[Earthscum]

Well, something new to add here... this just came up in another thread:

Not with a modern silicon transistor, the reverse beta is far too low. For most silicon transistors, the reverse beta is below 10, and it's often even below 1. You can read about the manufacturing differences here:
http://www.muzique.com/lab/reverse.htm

Howz about that?!?  

I love a good mystery.  

eta: Best highlight: "However, the most interesting effect is when this reverse booster circuit is driven into clipping - it clips in a matter totally unlike that of the basic common emitter circuit! The overdriven reverse booster generates a mess of harmonics different from other clipping circuits and configurations."

If that's the way this was designed, bravo to them! That's a really neat trick I'm gonna have to explore... I have a couple of old metal can Si's, too.

[CynicalMan]

I don't think there are any reverse-bias transistors here, but there is a Harmonic Percolator-style stage, with the AC176 and one of the AC128s. It looks sort of like this:



I'd like to take a closer look into how that works. It's probably one of those old high-gain cookbook transistor circuits like the Fuzz Face or the Mini-Booster.

I should also point out that the voltages you have on Q7 do indicate an hFE of around 500.