No Sound coming out of my Cobalt Drive

[GregDS]

Hello there,
that's it I give up... Solid state troubleshooting is not my forte and after trying signal-tracing with my oscilloscope and getting results I can hardly interpret, I turn to you folks.
I built the Cobalt Drive a couple of months ago and no sound is coming out of it, no scratchy pots, no funny sound, just none at all.
I'm really at loss troubleshooting it. With a tube amp, it's sort of easy, you trace the signal, in voltage amplitude and when you get less than expected, you're close. But with solid state I dont really look where to look as I dont know what behavior I should expect from these components.
I check for continuity where there shouldn't be. I know the PCB looks really dirty, but it's flux, not solder all over the board.
Can you folks help me out?

Thanks a ton!

 

[TheSin]

Where did you source those J201’s? Those are pretty hard to come by.
Have you verified all component values are correct? Usually when my stuff doesn’t work its because of that. I would also check solder joints and reflow them again if they look questionable. Good luck, Im sure you’ll get it fixed.

 

[fornio]

Obligatory first question of whether it works in bypass?

 

[GregDS]

Hey,
thanks y'all for your replies!
Yes, works in bypass, I'm always using the same test box to fire up my new builds, so it can't be outboard wiring (I wish it was, because I can troubleshoot weird tone, but no sound at all....)
I sourced them on Banzai music, which is a pretty reputable German webshop... How can I test them though?
I have transistor test sockets on my DMM, but does that work for JFET?
I reflowed a whole bunch of solders but in the meantime, I'll check all the values.... ?
thanks!

 

[GregDS]

I checked the color codes on the resistor and they are the right ones. So it has to be something else...
Considering I'm not getting any sound, I'm considering the power supply, maybe?

 

[okstateblues]

Does your led work when the footswitch is activated? Also if you are familiar with following a schematic, you can trace the audio path to find where you might be faulting.

 

[okstateblues]

Also you might might look into cleaning up your board a bit with some alcohol and acid brush just to eliminate any small chance you might be getting continuity where you aren't supposed to. Don't forget an audio probe is a great friend.

 

[Chuck D. Bones]

Of the few solder joints that were in focus, some looked cold. They might be fine, it's hard to tell from the pix. Definitely clean the board, all that flux residue could be hiding something. Then perform a thorough visual inspection. Verify all the component values, we can't see most of them. Fix any problems you find. Make sure the JFETs are snug in their sockets. If it's still broken after than, then we can make some measurements. For starters, Q5-C, Q6-C and IC1-6 should all be close to 4.5V.

 

[GregDS]

Does your led work when the footswitch is activated? Also if you are familiar with following a schematic, you can trace the audio path to find where you might be faulting.

Well, I'm using a test box so I haven't hooked up the LED through the PCB.
I can follow a schematics, I'm mostly working on tube amps, I have a signal generator, an oscilloscope, DMMs, etc. but solid state technology is a big mystery to me, they are not voltage amplifiers but current amplifiers, which does not translate well visually on a scope. In other words, I dont know what I'm looking at.

Also you might might look into cleaning up your board a bit with some alcohol and acid brush just to eliminate any small chance you might be getting continuity where you aren't supposed to. Don't forget an audio probe is a great friend.

Just bought some isopropyl, will clean this bad boy as soon as I receive the bottle (tomorrow I think)

Of the few solder joints that were in focus, some looked cold. They might be fine, it's hard to tell from the pix. Definitely clean the board, all that flux residue could be hiding something. Then perform a thorough visual inspection. Verify all the component values, we can't see most of them. Fix any problems you find. Make sure the JFETs are snug in their sockets. If it's still broken after than, then we can make some measurements. For starters, Q5-C, Q6-C and IC1-6 should all be close to 4.5V.

I bought a new (better) soldering iron and I'm not getting the cleanest solder joints, I don't really know what I'm doing wrong... I'm also using SN63/PB37, so that's pretty foolproof for people like me...

So on the collector of Q5 and Q6 I'm getting a whopping 7.19VDC and 7.33VDC respectively... so that's not good... on the leg 6 of IC1 I'm at 4.84VDC, so that's a bit better...

 

[Chuck D. Bones]

The JFETs are most likely the problem, just because J201s are problematic. Lot of bogus and out-of-spec parts being sold. Unless Banzai tests them themselves, they won't know if they have good parts.

If those JFETs are good, then the source voltage will be 0.2V to 1.5V higher than the gate voltage. They're just like vacuum tubes, except they won't burn your fingers. Make sure all the surrounding resistors are the right value and the solder joints are good before you condemn the JFETs though.

 

[GregDS]

So, I got back from holiday, now it's time to look back into this stubborn Cobalt Drive.
I followed your advice, bought some isopropyl and cleaned the board, it works really well, I can't see anything wrong with the soldering though

I measured voltages on my J201 and I'm getting the opposite of what Chuck D. Bones suggests, I measured:

Q1
Drain: 7.86
Source: 2.46
Gate: 2.85

Q2
Drain: 9.75
Source: 2.47
Gate: 3.20

Q3
Drain: 7.92
Source: 2.02
Gate: 2.66

Q4
Drain: 9.75
Source: 2.02
Gate: 2.74

 

[Chuck D. Bones]

I think you have Q2 & Q3 switched in your measurements above. Take another look at page 1 of the build doc.

Based on these measurements, it's a fair bet that your JFETs are bogus. We can do a quick test on them out of the circuit.

Remove one of the JFETs, then measure resistance from D to S. You should get <200Ω and it should read the same in both directions. If that reads correctly (I'm betting it doesn't), then use the DIODE measurement function on your DMM to measure the forward voltage from G (red lead) to S (black lead). You should get around 0.6V.
I think you already know the pinout, but here it is just in case.

 

[GregDS]

Okay, so first off yes I had Q2 and Q3 switched in haha
here are the measurements I'm getting:
Q1:
resistance between S and D: around 15k (but it keeps rising and sometimes I'm not getting any reading)
forward voltage: 1.14V

Q2:
10k (but it keeps rising and sometimes I'm not getting any reading)
1.14V

Q3:
no resistance reading (except if I touch the transistor with my fingers)
1.14V

Q4:
No resistance reading (except if I touch the transistor with my fingers)
1.17V

 

[Chuck D. Bones]

Just to confirm, this is measuring the individual transistors out of the circuit. You cannot touch any of the leads with you fingers because that will corrupt the reading.

 

[GregDS]

Yes, transistor out of the circuit! The resistance readings changes when I touch the plastic case of the J201, Im not touching the leads. Otherwise I'm just letting it lay on the table with mini grabbing probes on but I'm not always getting a reading at all (no matter the range on the DMM)

 

[Chuck D. Bones]

Got it. Whatever those things are, they are definitely NOT J201s, or any other JFET with that pinout. You can get surface-mount MMBFJ201s mounted on adapter boards from PedalPCB. You want the Vp to be matched as close as possible for the Q1-Q3 pair and for the Q2-Q4 pair. Best way to accomplish that is to buy 2x as many as you need and cherry-pick them. The rest can be used in pedals that don't require matched pairs.

These JFETs do not have to be J201s. PF5102 is a good sub.

I'll bet you thought this would be an easy build.

 

[GregDS]

I'll bet you thought this would be an easy build.

yesss I thought so...
Besides going the SMD routes and the PF5102, I'm curious what data should I pay attention to when searching for other subs? What do you call Vp?

Also, let me back up a little, because I'd like to understand what I've been doing.
I'm looking at the Cobalt Drive schematics and it looks like, as you said earlier, a JFET is like a tube, so is it correct to see those as the Gate between the tube's grid, the Source being the cathode (where you bias your amplifying component) and the Drain is the plate (where you feed (high) voltage and take the output signal from)?

You asked me to check the resistance between the Drain and the Source and it should have been really low (less than 200 ohms), is that the property of j201 in particular, or transistors in general, or is it a way to check if the transistor is an open circuit?

And lastly, the forward voltage is how much voltage needs to go through the transistor to start conducting, is that right?

 

[Chuck D. Bones]

Quick JFET lesson.
Like vacuum tubes, JFETs are depletion-mode devices, which means that with zero gate-to-source voltage (Vgs), they are fully ON. In the case of N-channel JFETs, we throttle them back with negative Vgs. When Vgs reaches Vp (pinch-off voltage), the JFET is fully OFF. Vp is also called VgsOFF. When we use JFETs as amplifiers, we run them somewhere between fully ON and fully OFF. When you measured the resistance between source and drain with no voltage applied to the gate, you were measuring RdsON, the drain-source ON resistance. The spec for J201 is RdsON < 150Ω. When you measured Vf from G to S, it was to see if there was a diode between those two pins with the cathode at the source. Normally, we do not forward bias the gate, so that diode never turns on. We did it for this test to see if there might be an N-channel JFET inside that hunk of black plastic masquerading as a J201. Your parts failed both tests.

The two JFET specs we care about in pedals are Vp and Idss. Vp I already explained. Idss is the drain current when the gate is shorted to the source. This is the most current the JFET will carry. Those two specs determine the operating range of the JFET. Vp and Idss vary a LOT from one part number to the next and even within one part number, the variation is considerable. The specs for J201 are Vp is between -0.3V and -1.5V; Idss is between 0.2mA and 1mA. Some circuits require the JFET to be at or near one end of the range. Pedal builders have to cherry-pick JFETs in a lot of cases.

 

[GregDS]

Wow, thanks a lot for taking the time, I am glad I understand tubes because what you're saying makes a lot of sense when I translate it in a tube context.
However, in the meantime, I changed the fake J201's for 2N5457's and I'm still getting nothing, I've read that in some applications they can be swapped for one another, is that right?

 

[Chuck D. Bones]

Sometimes. The Vp spec for 2N5457 is -0.5V to -6.0V. If the Vp for the ones you have is close to the low end, and they're matched, then they'll work in this pedal.
The other thing to keep in mind is we know the JFETs were bad, but we don't know if there are any other problems lurking. Assuming that nothing else is wrong, a quick test for this pedal is to go back and repeat the Q5 & Q6 Vc measurements. If the JFETs are good and fairly well matched, then Vc will be close to Vref.