Mojito deluxe gain/crunch

[comradehoser]

This may or may not be a troubleshooting thread.

How much crunchiness should I be getting out of this? I am only really getting very tame breakup with boost engaged.

Unity without boost engaged is almost full CW, so it could maybe an issue with Q2 if Q1 is responsible for boost.

Seems pretty quiet and mild overall, so I'm thinking something went awry, as I don't always get on with SMD, and I thought I could rock it with some mmbf4392. Well. Separated a leg off one, so gives you an idea of my proficiency. How fryable are smd transistors?

On the other hand, I am also prone to "is it supposed to sound that way-itis" so I need some independent confirmation, and the only demo I unearthed was a frickin slide player, damn. I am not a slide player.

Ran an audio probe, and there's nothing that jumps out.




Gain pot maxed:

Q1 (bottom) mmbf4393
1: 3.03
2: 3.04
3: 2.48

Q2 (top right). Mmbf 4393
1: 1.95 (variable)
2: 3.10 (variable)
3: 0.5 (variable)

IC 1 ca3130ez
1: .088
2: 4.51
3: 4.23
4: 0
5: 0.88
6: 4.42
7: 9.02
8: 4.75

Ca3130 from Mouser, Mmbf4393 from Ppcb or mouser.

Pins 1 and 3 of Q2 look a bit low compared to Q1, and the fluctuation/instability with gain up make me think that could be it? Q2 is stable with gain down. But what do I know?

I miiiight have one last little smd buddy to try, but I didn't think so. I have bad luck with these guys in particular.

[Edit: tried reflowing Q2 following with what has worked for me in the past: liberal flux on the legs and a retouch with a solder loaded tip. Now I have no output at all ]

 

[comradehoser]

I used BA482 because they have the same I-V curve as the JFETs BJ used in the pedals that the Mojito is based upon. He wired D & S together and used the gate-channel junction as a diode. Other diodes will work and the differences are subtle. LEDs will not work well because Vf is too high.

What drain voltages do you have now on Q1 & Q2?

I promised somebody in another thread that I would report back on the difference.

At the moment, with in jack ground clipped to ground and drive pot at zero, boost off, we have

Q1
D 2.97
S 3.15
G 1.68

Q2
D 0.040
S 0.337
G 0.027

And, noticing that this looks an awful like the old voltages, plugged in and we're in the same old place. Almost a clean signal, just getting to unity with everything dimed. IC voltages remain the same.

Really really nonplussed. I did nothing between last night and this morning other than unplug the pedal and move it.

I guess circuit board sacrifice and waving my wand +5 of eutecticity didn't actually solve anything.

 

[comradehoser]

Maybe it was all a dream...

 

[Chuck D. Bones]

Something definitely wrong with your build.

The gates should be at 0V.
The drain voltages should be higher than the source voltages, by at least 3V.

Blindly reflowing solder joints is not a good path to success. Visual Inspection under high magnification is needed.

SMD soldering requires good light, a magnifier, a sharp soldering tip, a steady hand and very little flux or solder.

If you apply too much solder, you run the risk of either getting a short between pads or the part can float up on the pool of solder leaving one or more leads up in the air.

 

[comradehoser]

Something definitely wrong with your build.

The gates should be at 0V.
The drain voltages should be higher than the source voltages, by at least 3V.

Blindly reflowing solder joints is not a good path to success. Visual Inspection under high magnification is needed.

SMD soldering requires good light, a magnifier, a sharp soldering tip, a steady hand and very little flux or solder.

If you apply too much solder, you run the risk of either getting a short between pads or the part can float up on the pool of solder leaving one or more leads up in the air.

Well, I took those SMDs out before you posted, and I'm using through-hole 2N5457 with sockets to test to same effect. It's very possible I messed up some traces in the SMD solder pads (actually I know I did), and I DEFINITELY take your advice for future SMD soldering, but would that affect the board if I've defaulted to the through-hole ports?

Do you mean reflowing without ensuring that the reflow has been successful? I wasn't erratic in my reflow in the sense that I literally reflowed every single solder point or anything with solder on it (minus the SMD pads).

Why that should yield success for a half hour and then failure after 12 hours is really bizarre to me.

 

[Chuck D. Bones]

Let's see a pic of the current state of your build.

The sockets that most people (mis)use for transistors are the machined pin sockets intended for IC leads. Transistor leads are almost always too small for the sockets and don't make a secure connection. That alone can explain why sometimes your board works and sometimes it doesn't.

If the JFET biasing is wrong, it's not gonna work. Getting that right is crucial. JFETs are not plug-n-play. If you have a breadboard, you can check the biasing on the JFETs before they're installed. All it takes are a few resistors, 9V power and a DMM. If you're interested, I can provide how-to instructions.

when I said "blindly reflowing solder joints" what I meant was reflowing solder joints without first determining if they need to be reflowed. Visual inspection is the recommended method of determining whether a solder joint is good or not. Every time you touch the soldering tip to the board, you risk collateral damage. Parts & pads can take only so much heat before they fail.

 

[comradehoser]

Let's see a pic of the current state of your build.

The sockets that most people (mis)use for transistors are the machined pin sockets intended for IC leads. Transistor leads are almost always too small for the sockets and don't make a secure connection. That alone can explain why sometimes your board works and sometimes it doesn't.

If the JFET biasing is wrong, it's not gonna work. Getting that right is crucial. JFETs are not plug-n-play. If you have a breadboard, you can check the biasing on the JFETs before they're installed. All it takes are a few resistors, 9V power and a DMM. If you're interested, I can provide how-to instructions.

when I said "blindly reflowing solder joints" what I meant was reflowing solder joints without first determining if they need to be reflowed. Visual inspection is the recommended method of determining whether a solder joint is good or not. Every time you touch the soldering tip to the board, you risk collateral damage. Parts & pads can take only so much heat before they fail.

Haven't bothered to reclean, sorry.

Got you on the reflow feedback.

Both jfets are holding in the sockets with some tooth. Moving them about does nothing to the readings, so not a case of intermittent contacts. Also, I had soldered the 4393 SMD directly into the through hole ports without sockets and it was behaving the same.

This makes me think it's something external, so maybe my clod-handed soldering ways shorted or burned up something early that is also affecting the through hole space. My continuity probe didn't turn up anything, nor did audio other than quiet signal after r13 going into the gate.

Source pad on Q2 is lifted.

JFEt biasing: sure! I do have a breadboard that I am loath to use but I just used to build the Moonn Sojus 31/bohm- klingenhoffer trickverzerrer fuzz circuit successfully.

But, I don't mean to ask a whole lot of your time either. If the instructions on breadboard JFEt biasing are in your boneyard articles on jfets, I can just go look at those for sure or Google it, but please feel free to instruct me if it's more arcane.

Also, if the board is just fried, then I am eventually going to be able to swallow that this little buddy might have to join the median comp and the twin face in the bin of over soldered sadness, and I'll just get another board and transfer (?) the parts. Out of more than 100 builds, 3 isn't too bad though.

At this point, I'm just trying to solve the puzzle to understand why. But I get that not everything has an answer and you DFU'd again can be sufficient.

 

[Chuck D. Bones]

The simple instructions are to build the circuit around Q1 consisting of R3, R4 & R19 & the BOOST pot. You can use a 47K resistor to simulate the BOOST pot. Measure source & drain voltages. Try different JFETs until you get close to 6V on the drain. Then tweak R19 if necessary to dial it in.

Repeat for Q2 using R15-R17. Select a JFET that gets the drain close to 6V, then tweak R17 if necessary to dial it in.

If you have to change the resistors by more than 1/3 to dial-in the drain voltage, then you have the wrong JFET. If you have J113, try those.

That's all there is to it.

Sometimes the best move is to punt. I would not bother trying to reuse anything that isn't in a socket. Parts are cheap and it's too easy to damage them. Reused parts are always suspect.

 

[comradehoser]

The simple instructions are to build the circuit around Q1 consisting of R3, R4 & R19 & the BOOST pot. You can use a 47K resistor to simulate the BOOST pot. Measure source & drain voltages. Try different JFETs until you get close to 6V on the drain. Then tweak R19 if necessary to dial it in.

Repeat for Q2 using R15-R17. Select a JFET that gets the drain close to 6V, then tweak R17 if necessary to dial it in.

If you have to change the resistors by more than 1/3 to dial-in the drain voltage, then you have the wrong JFET. If you have J113, try those.

That's all there is to it.

Sometimes the best move is to punt. I would not bother trying to reuse anything that isn't in a socket. Parts are cheap and it's too easy to damage them. Reused parts are always suspect.

Rock and Roll. Thank you very sincerely for your help. Will suck it up and try again, folding in the protocols I learned today from you.

If anything, I learned a lot of new information, and now it's here for anyone else to learn from as well. Aces.