Rat needs jump-start

[lars-musik]

Dear Forum,

I am really at a loss here.

I built a RAT, pretty standard with an adjustable RUETZ mod.
Now it seemed to have some loose connection, because when I tested it with guitar and amp it just worked intermittently. So I carried it over to my workbench, re-soldered all pads and checked it on my test-rig (as I've done before) an voilà, everything fine. Back to my amp: nothing.... Back to my workbench, signal generator and scope but couldn't troubleshoot, because it worked fine again. Spooky.

I just realised, that this thing seems to need a specific input signal strength to jump start it. Gently picked single coil guitar doesn't do it, but strummed humbucker does. So does the signal generator (at an admittedly high volume of 1Vpp).

Do you have any idea what might cause such a behaviour?

Thanks!

Here's the schematic. Maybe there's some copy-error on my part?

[antonis]

What JFET you use for output buffer..??

[lars-musik]

I just heat up my desoldering gun because I wanted to test this part. It's a 2N5457.

[GGBB]

Pleas post pin voltages of both devices.

[lars-musik]

Pleas post pin voltages of both devices.

I could try. However, probing the circuit jump-starts it more often than not. So I won't be able to acquire voltages in the "not working" state reliably...

[lars-musik]

What JFET you use for output buffer..??

same problem with known-good j201

[Marcos - Munky]

When the pedal is not working, does the bypass works?

Based on what you're saying, my best guess is something that happened to me a few times: the issue is the box lid shorting something.

When you connect the pedal to your test rig, I'm assuming you're removing the lid. So there's no shorting, and that's why it always works, but looks like it needs a jump-start.

When you connect it to your amp and guitar, the lid may be almost barely touching something inside the enclosure. The pressure required to press the switch may sometimes make the lid touch something (then the circuit won't work), while sometimes it may not be enough pressure to make the lid touch something (then the circuit works).

I had this issue on a few crowded 1590A I built. The circuit worked without the lid, then I put in the screws and started to tight them with the circuit still plugged to my guitar and amp, and then suddenly the circuit stoped working. I removed the lid and added a few layers of something to insulate the guts from the lid (duct tape, cardboard, or whatever you have on hand).

Test the pedal gain using your guitar and amp, but with the lid removed, and see if this "solves" your problem.

[duck_arse]

photos of the offending item, please?

[antonis]

photos of the offending item, please?

Just in time I started to worry..

[lars-musik]

the issue is the box lid shorting something.

I wish it would be that easy. The pots are offboard, everything's wired up, the PCB is free floating for the tests, I never screw lids on before everything's in order.... So, unfortunately that's not it. And bypass works as it should.

photos of the offending item, please?

I'll try to collect my courage for tomorrow and will deliver photos then. Of course, I only told half the story, the whole shebang is a double pedal with an order switch. So loads of wires... But I am pretty certain, that the wiring is not the culprit. It really sounds like some capacitor charging when I hit the strings hard and the tone builds up from fizzeling to as should be.
Maybe I'll replace all the electrolytic caps.

[Marcos - Munky]

Maybe the problem is how you wired it to the other circuit. Try to isolate the rat circuit from everything else, including switches, and check if the problem still happens.

[vigilante397]

I'm not sure why J201 wouldn't work, but I've never had any luck with 2N5457 as an outpur buffer, I use 2N5458.

[Govmnt_Lacky]

If this turns out to be just a bad guitar cable.. I will be supremely disappointed 

[ElectricDruid]

If this turns out to be just a bad guitar cable.. I will be supremely disappointed 

Me too, but it wouldn't be the first time. Stuff like that still catches me all the time.

[GGBB]

I'm not sure why J201 wouldn't work, but I've never had any luck with 2N5457 as an outpur buffer, I use 2N5458.

The standard RAT's implementation of the JFET output buffer is highly dependent on the JFET itself. Because the gate is biased to ground (~0V), it is close to the min spec for cutoff -1.5V in the 2N5458. So some units toward that end of the spec range can exhibit sputter/gating. Most units' spec (theoretically) would fall towards the middle of the range -1.5 to -7.0 and therefore more comfortably away from cutoff. The 2N547 range -0.5 to -6.0 is a little more susceptible to this problem. The J201 is much more susceptible: -0.3 to -1.5. As the output signal voltage off the opamp is raised, the problem is exacerbated.

The Turbo RAT - with the higher LED clipping threshold thus higher opamp output voltage - avoided the problem by adding a resistor between Drain and Gate to bias to 1/2 Vin.

[PRR]

When soft signal won't pass, but a loud signal does AND then soft sounds pass, in my experience bad-contact is the best bet. Thin film of tarnish, 0.01V won't break-through the oxide layer, 1V does, and leaves a "good" spot for a while. There is probably fascinating micro-physics down there. But when the show won't go, foo on physics, jiggle clean replace re-solder contacts until you are SURE it can't fail before the end of the night.

Poor grid/gate resistor can sometimes be the problem. Some grids/gates with NO DC path for leakage will de-bias to a no-work condition, but BIG signal may bring it back to a working condition. This is more often associated with "then it works until power-down, is dead an next power-up, until a BIG signal".

[lars-musik]

THANKS ALL!

I found it! It was indeed a bad cap! I suspected C11 (the VB cap) but that one was good. Then I started to desolder the electrolytics in the signal path and the first one was a hit! I replaced C7 with a non-polar SMD 4u7 and  voilà, everything fires up right away.
Unfortunatly my workbench is littered with parts and stuff, so I am not sure if I find the discarded one again in order to measure it. BUT as I reflowed all pads once and all electrolytic caps twice I am pretty sure it wasn't a bad solder joint.

Really the sound was quite distinctive like something charging up before working correctly, so it really might have been a bad cap. Although what Paul says  might be what happened here.

Now I have to put it all back into the enclosure... No fun.

EDIT:
Just measured the cap. Says 4.7uF on it, measures 84uF. Something's wrong there.

[lars-musik]

Finally boxed up. And although it looks pretty prototype-ish on the inside, I'll show it to you....

It's a "1981" DRV and a Rat btw. The order is switchable.

[Ben N]

How is that working with the toggle between the stompswitches? Looks scary to me.

[lars-musik]

 It's a short shaft, so there's no change to stomp on it.