Spaceman Sputnik II

[digi2t]

Managed to get my hands on this piece of vaporware. Ironically, after failing miserably to buy one when they would become available, I find one for sale about 30 minutes from my house. Go figure...

Anyway, let's get on with it. Gutshots;





As always, they're up to their usual shinanigans, like defacing components, weird a$$ values, and whatnot, but this time, they went as far as embedding the traces between the ground planes. Needless to say, lots of time poking and prodding with the multi-meter was spent, trying to follow where everything was going. This is my preliminary schematic;



I've got some unknowns here, specifically one component that reads as "N-Channel FET" on the Peak meter, but it only identifies the gate, so I'm unsure of with is the source and the drain. It might be an MPF102... maybe.

It's an interesting circuit. Really nice fuzz, with a "Drift" mode which basically send the unit into oscillation. The oscillation is adjustable, you can have it on full time, or set it to be attack sensitive. Very reminiscent of the Skyripper in a way. The filtering is lovely, really lots of usable tones in this thing. It's as if they took the Gemini III, and went a totally different direction with it.

Anyway, what do your eagle eyes see? Does this schematic make sense? Wouldn't surprise me if trace(s) are still hiding.  

[diy-tubes]

2 ge transistors are russians, no marking?

[digi2t]

3 Russian transistors. The only transistor looking part that I could lift a marking off was the VP3203, the rest were all sanded and painted over. The Peak meter gives me a reading on the first two, and diode testing the third one gives me the same pinout as the two others. I would hazard to say that they may all be the same model. The Peak gave me a gain of mid 60's for Q1, low 80's for Q2, and Q3 is unknown, but breadboarding will be the ultimate fact finder. Where the gains are concerned, and the shape of the transistors, maybe something out of the MP 15/20/etc. series. Probably in the "B" gain bracket.

EDIT:

Just noticed a problem. Low filter, in the middle position won't work, nothing going to the base of the second Ge transistor. Must be a trace going somewhere from here. I'll check it tonight.

[diy-tubes]

Do you use DCA55 or DCA75? DCA75 gives a bit higher measurements.
What about transistors housing, it can be many types: similar hFE (80) gives GT308/1T308, MP series will give a bit lower hFE (only Q1 can be used here, traditional choice of low noise MP39B, P28 and others, or something simple like MP42) Higer hFE gives P416 in the same housing. But I think any with the same pinout and hFE must do the job.

[digi2t]

DCA55. To date, I haven't really been able to justify moving up to the 75. Overkill for what I need.

Your call on the transistors is probably spot on.

That missing link between Q1 and Q2 with the switch in the middle is really bugging me. Can't wait to poke at it some more tonight.

[thermionix]

Can't wait to poke at it some more tonight.

Sig-line material right there.

(I'll delete it if you want me to.  Will anyway after a while when I get bored of it.)

[digi2t]

Can't wait to poke at it some more tonight.

Sig-line material right there.

(I'll delete it if you want me to.  Will anyway after a while when I get bored of it.)

It's all good brother. It's like a Trump tweet, once it's out there...

[pinkjimiphoton]

dino, i've found that reversed beta'd ge transistors will often read as an n channel fet.

i think at a certain point of leakage, the gain evens out either way its oriented, but it appears to the tester as a fet
and will often work as one.

if this has russkie q's in it already, i'd wonder about that. probably wrong, but i've seen some weird shit too!

[pinkjimiphoton]

Can't wait to poke at it some more tonight.

Sig-line material right there.

(I'll delete it if you want me to.  Will anyway after a while when I get bored of it.)

It's all good brother. It's like a Trump tweet, once it's out there...

its gonna be 'UGE.....

[digi2t]

dino, i've found that reversed beta'd ge transistors will often read as an n channel fet.

i think at a certain point of leakage, the gain evens out either way its oriented, but it appears to the tester as a fet
and will often work as one.

if this has russkie q's in it already, i'd wonder about that. probably wrong, but i've seen some weird shit too!

Naw, Ge is out of the question for the one with the "??". It's a plastic package, which they painted silver. If you look at the pic I posted, there are four in a row, with a fifth one further down, on the left side. The first and second from the top are for the Mill bypass. The third one is painted the exact same color as the second, which would lead me to believe that it's another BS170/2N7000 MOSFET. The fourth they painted silver, and the the Peak tells me that it's an N-Channel FET, and the gate is not the same pin position as the others. N-Channel FET with a gate on the end, that makes me think J201 or similar. The fifth one, the space boys got careless. They painted it with a darker green paint, but didn't grind it, so the writing was visible through the paint. Other than the Mill bypass, I'm not quite sure what the MOSFET/FET combo is doing here, especially when the drain of the Mill bypass MOSFET is also feeding the gate of the next MOSFET. Mill bypass of the DRIFT circuit? Acting as a cut out for the Q1 signal from hitting Q3 maybe? I'm not familiar with this arrangement at all.

Well, more "poking" revealed a missing trace. Also corrected a resistor value. Those 1/8w resistors are so tiny, the red looks like brown and vise versa.  Here's the update;



I missed a connection between the 100K/123 junction, and the Q2 base. Audio now floweth.

[pinkjimiphoton]

you are a very sick young man. i respect that and salute thee!! 

[digi2t]

More corrections. Unfortunately, the new glasses haven't helped.  The trimmers make more sense now.



The unknown FET (?) is still, well... unknown, but I'm thinking that maybe it's being used as a sort of threshold gate for the oscillation. By adjusting the SCAN voltage, you can adjust where the oscillation will kick in, according to the signal strength coming from Q3 collector. You crank up the SCAN, and the oscillation is on full time, turn it down and you can have it come on at a desired signal output level.

As for the BS170 below it, I'm guessing it's switching off the unknown FET when in bypass, maybe to prevent any oscillation audio from bleeding into the clean signal?

Does that make sense, or am I whistling Dixie here?

[bluebunny]

DCA55.

As I understand, the 55 doesn't have the chops to differentiate between D and S, so all it does is identify the gate on a JFET.

[digi2t]

DCA55.

As I understand, the 55 doesn't have the chops to differentiate between D and S, so all it does is identify the gate on a JFET.

Not my experience. I think the problem stems from the fact that I'm testing the components in situ, so the surrounding components are messing with the readings. If I take a FET or MOSFET and test it directly, the 55 has always been able to tell me what's what. The same applies to testing transistors. In some cases, depending on the surrounding cast of characters, the 55 might give you a proper declaration, other times, it comes back with "Common diode network", which is understandable considering that other components in the circuit are influencing the reading. Another thing when test transistors while they're in the circuit, the gain/leakage readings  may or not be accurate. I would tend to take them with a larger margin of error in this case.

All in all though, it's a pretty fine tool.

[bluebunny]

Ah, OK. I was misinformed. Thanks Dino. (I should really have tried it myself!)

[digi2t]

Ah, OK. I was misinformed. Thanks Dino. (I should really have tried it myself!)

Marc, my sincerest apologies, you are correct where JFET's are concerned. Only the gate is identified. It will however identify all the pins on MOSFET's.

I spoke too soon. Sorry.

On the other hand, your astute observation on the 55 helps reinforce that the unknown component is indeed a JFET. Thanks!!

I think I have enough info now to start breadboarding this sucker.

[diy-tubes]

Measuring JFET with DCA75 or chinese transistor testers like gm328 can measure Vgs(off), so we can suppose the type. Even fake ones like modern J201 with Vgs(off) about 1.2-1.5.
Common diode network DCA will show if hFE is too low, or if it fail to determine the type.
BTW do you measure in curcuit? DCA manual warn from it.
Waiting for breadboarding results! At least I will try to offer russian ge transistors types with proper hFE.

[bluebunny]

No problem, Dino!  Sometimes I just regurgitate stuff that other far cleverer folks say in the hope that I appear intelligent.  Doesn't always work!      Either way (and in the best scientific tradition), we end up learning something.  And either way, I'm intrigued by what happens next.  Over to you.    <reaches for popcorn>

[digi2t]

Common diode network DCA will show if hFE is too low, or if it fail to determine the type.
BTW do you measure in curcuit? DCA manual warn from it.

In cases like this, where I'm really not interested in removing components, I'm not looking for accurate numbers. Just a simple pinout to point me in the right direction. I understand that the readings can/will be erroneous measuring in circuit, so I don't put that much faith in them. It gets me into the ballpark, and breadboarding gets me the more accurate numbers.

If the 55 does give me numbers in circuit for Ge transistors, I've noticed that the gains always tend to be lower, and the leakage higher, than what the transistor would actually measure if it was off the board. As such, I've developed a sort of "guesstimation" of what it is probably close to. For instance, if I see "hFe 62" and "leakage = 0.72mA", I find that choosing a transistor in the mid 70's, with a leakage around 0.20 or better, usually gets me pretty close to the voltages in the original circuit. Not an exact science by any stretch, but it helps noobs like me to keep plodding along.

In the case where the 55 returns "Common diode network", I refer to the DMM diode test for pinouts. Breadboarding then becomes the process of elimination where gains and leakages are concerned.

Finally, where tone is concerned, well... everything goes out the window. You can plug in an AC128, and then an MP20, with exact same gain and leakage specs, and get two totally different feels from a circuit. Why, I don't know, and frankly, I could care less. It's that unknown that keeps me interested. A fool in blissful ignorance. 

No problem, Dino!  Sometimes I just regurgitate stuff that other far cleverer folks say in the hope that I appear intelligent.  Doesn't always work!      Either way (and in the best scientific tradition), we end up learning something.  And either way, I'm intrigued by what happens next.  Over to you.    <reaches for popcorn>

Thanks just the same Marc. You're absolutely correct, we always have more to learn.
Would that happen to be buttered popcorn by any chance? Yum!

[diy-tubes]

Just remark for DCA55 users, that DCA55/75 not intended for testing in circuit. It can be damaged by DC stored in capacitors, or if AC/DC is connected.
digi2t, thank you for sharing you experience!