Rat clone lower output

[forsakenrider]

I built a General Guitar Gadgets rat clone (etched the pcb myself), and I notice the output is a lot lower then my original RAT. The only mod I did was the Ruetz mod and that doesn't seem to change the output volume at all. I also tried a metal can opamp from Smallbear and a regular one without much difference.

Any ideas? Are there value changes I should/could make for more boost?

[Cozybuilder]

Check your value for R10- it should be 10K. If you could post a clear photo of your build, that would be a big help. I'm assuming you've looked over the board for correct component values, orientation, performed the normal checks for continuity and solder bridges, voltages etc.

For the Ruetz mod, did you simply omit R5/C6 (47R/2u2F), or did you replace R4/R5 with a 1K (or 500R) pot, connecting pin 2 (pot) to the inverting input of the op amp?

[forsakenrider]

It's not like its TOTALLY quiet, just max volume is about 3/4 or less on the real rat. I like to run my rat with a lot of volume and not very much distortion, so this pedal is a bit too quiet.

For the ruetz mod, I replaced the r5 resistor with a 1k pot, as per the beavis 4 knob rat mod....

I will get some pictures and post them.

[Keppy]

When built correctly, the diodes are the limiting factor on the volume of the Rat. The opamp can get crazy loud, but the diodes clip the signal to about +/-.6 volts max. If you used germanium or Schottky diodes it would be quieter but more distorted. LEDs would be the opposite. Once you're sure you've built it correctly, changing the diodes might help, if the ones you're using are out of spec or damaged.

The other way I can see to drop the volume, assuming the gain and filter knobs are functioning correctly, is with the biasing of the JFET. If the gate bias resistor was less than 1M you'd get a volume drop. The 10k on the source is less critical but could still be a factor, since if the JFET biases too close to saturation or cutoff then it will clip the signal, making it even smaller. JFET variance is a factor in biasing, so you might just need a different FET or different source resistor. Check that the source voltage is close to 4.5v. Speaking of the JFET, did you check the orientation?

One more thing to check: Make sure the 1uF output cap is correctly oriented (if you used an electro). An electro's properties change if it's inserted backwards. If that caused it to become resistive at guitar frequencies it would make a voltage divider on the output.

I hope you get it how you like it!

[GGBB]

GGG's rat biases the JFET gate differently than the original rat. It is via a second voltage divider that should be a little less than 3V. The source is usually at something between 1 and 2 volts, so normally the GGG rat shouldn't suffer from JFET cutoff. But you should check the voltages anyway and report back. If you see a positive G-S voltage you are probably fine. If it is less than -1V there is a possibility that cutoff is happening. In this case adjust the R13/R12 divider to raise the bias voltage (lower R13 or raise R12 - swapping them should work well). And definitely verify the orientation.

[forsakenrider]

Thanks guys.
I tried different diodes, and removing r12 and r13 and using a jumper (It says in the GGG instructions that this biases it the same as the original rat). The only thing that makes it louder is the use of red LED's instead of the stock diodes.

Thanks for all the ideas, perhaps it is something to do with the opamp, I will go over those values next.

[bool]

If the use of LEDs make it louder, that's to be expected. This will happen in virtually all of the possible diode-swap cases.

And this also makes me doubt there's some fault with your opamp.

[forsakenrider]

I know about LED's VS diodes for clipping, my point is that it is still more quiet then an actual ProCo RAT which uses diodes.

So what could be different then the stock rat? Anyway, here's a pic from before I tried the LED's and removing r12/r13 and using a jumper.
https://www.dropbox.com/s/xqwhktvw14i8q7w/image1.JPG?dl=0

[GGBB]

The only difference between the original RAT circuit (not RAT-2) and the GGG version after the clipping diodes is the bias setup of the JFET output buffer. The op-amp will not make any difference as far as volume is concerned because of the clipping diodes - so don't bother looking there. A drop in distortion level would point to a problem with the op-amp section. Since you haven't posted any voltage figures we are only able to make educated guesses. My bet is still on either the JFET biasing or possibly the 2N5458 itself since they vary a fair bit from unit to unit in terms of gain and cutoff.

[forsakenrider]

I have tried two different 2n5458's with the same results.

One thing is, when I first gave the pedal to a buddy to test, he plugged in a "centre positive" power supply for a few seconds. I thought it was ok though because of the 1n4002 for protection.

Here's some voltages

BAT:
8.86v

2n5458:
G:0 (ground, as per original with no divider)
S:0.27v
D:8.49v

LM308
1: 7.77
2: 4.24
3: 2.11
4: 0
5: 0
6: 4.24
7: 8.48
8: 4.30

[GGBB]

Here are the voltages from my Vintage Rat reissue:

BAT: 8.96V

2n5458
G: 0V
S: 1.69V
D: 8.93V

LM308
1: 8.29V
2: 4.45V
3: 4.02V
4: 0
5: 0
6: 4.45V
7: 8.92V
8: 4.52V

Your 2N5458 source voltage seems low to me, but even more strange is the low number on LM308 pin 3. Not sure why. It also seems strange that you have comparatively such a large drop from Vin to LM308 pin 7 and 2N5458 drain - that could suggest a problem with the 47ohm resistor which could have occurred during reverse polarity, or some kind of current draw issue. Perhaps check that resistor and the diode. But I just don't know how that could explain a volume drop. I don't have answers unfortunately, but your numbers do seem a little off.

[Keppy]

Pin 3 is connected to Vb through a 1M resistor. If you have a meter with a 1M impedance, the reading might be a case of meter loading. This seems likely, since the other pins have the expected voltages. Check your Vb at the 100k-100k divider just to be safe, though.

Your gate voltage should be higher from what GGBB said about the GGG version of the Rat. Check voltages on R12/R13.

Check the voltage on the diodes and the orientation of C7 as well. If any DC is leaking through that cap it will throw off the action of the diodes.

[forsakenrider]

Keppy, as per this from GGG:
"The original had that 1 meg resistor
set to ground voltage and if you want your Rodent to be as original,  leave R13 blank
and use the included 0 ohm resistor (jumper) for R12.  This will set the voltage into
R9 at ground voltage as original."

That is how I have set up my "rat"

I will check the other voltages and try a different mutli meter

[forsakenrider]

With a better quality meter Pin3 is at 3.82v, so thats a bit closer. I also changed the 9v in wire to be sure it wasn't half broken or bad solder joint. The 47ohm resistor still reads correctly, and I changed the 1n4002.

Voltage at the diodes: -0.334v, But I also measured it once at 1v or so. It seems to settle in the negative region though.

vb at the divider is 4.2v now.

S (centre pin) is now 0.277v, so it hasn't changed.

[Keppy]

Sorry, I misunderstood your previous posts about the gate biasing. I should have read more closely.

It sounds like pin 3 is fine. If C7 was leaky there would be a positive voltage on the diodes, so it seems like that's fine as well, though I'm not sure why it would show negative voltage there.

The fact that the pedal got louder with LEDs proves that the opamp is putting out enough signal before the diode clipping (like bool and GGBB said). The problem has to be in the clipping stage or output buffer. It may be that you built it correctly but have FETs with a Vgsoff that's off the mark for this circuit. The clipping diodes allow the signal to swing approximately .6v up or down. Right now, your FET only allows for swings of .27v in the downward direction, so it's definitely re-clipping the signal that comes off the diodes. The source voltage needs to be higher. Anything over 1v should be just fine.

Have you had the gate biased to ground the whole time, or did you start with the voltage divider in place (R12/R13)? The voltage divider was probably added to make the buffer work with a wider range of FETs. ProCo might use something different that a 5458, or might select FETs from within the part number for this purpose. This seems like the solution, except that this quote makes it sound like you had the divider in when the problem started.

I tried ... removing r12 and r13 and using a jumper...

[forsakenrider]

I started with the divider, and its always been quieter. I have been A/Bing the pedal with my vintage rat (83-86 model I believe) and its noticeably quieter at low gain settings. I also did try two different 5458's.

Should I put a socket there and try all my fets until I find one that works?

[Keppy]

Before doing any more desoldering to the board, I'd try an audio probe to make sure the volume drop is really at the FET. Your voltages seem fine except for the FET, but the source voltage should have been higher when you used a positive gate bias voltage and yet you had the same problem. Check both sides of R7, R8, C9, Q1. You should hear full volume where R7 meets the diodes, reduced treble as you move past R7 & R8, and then identical sounds as you test both sides of C9 and Q1.

[forsakenrider]

I've never tried an "audio probe" before, seems like a good way to make lots of cracks and pops on an amp! So I tried feeding a sign wave and then reading the wave at those points on my oscilloscope. At full volume the wave is almost double until Q1 and at that point the signal goes in the gate and comes out the source about half.

Do I have a bunch of bunk 2n5458s? I'm not really sure where I got them.

Is there something else I can try? maybe a different part I can try?

Is my test even valuable? I just got an oscilloscope and am still learning how to use it.

[Keppy]

You can try any N-channel JFET, like J201, 2N5457, 2N5485, etc. If you use the voltage divider, preferably with equal values to bias at 4.5v, then you can use a bipolar NPN transistor such as 2N3904 or 2N5088.

I'm glad you found away to trace the problem to the FET, so we could rule out the other causes for good.

[forsakenrider]

I put in a j201 and it is now loader then my vintage rat.

Does that mean all my 2n5458's are bad/fakes? I think I have 10 of them.