Require pedal genius help!

[monkey2410]

So I've built the following pedal from the schematic below exact other than one rather large factor, I don't have the J201 transistors that it is designed for, therefore I replaced them with 2N5457's as they are apparently very similar but unfortunately the circuit does not like these so I require the advice of some experts to maybe crunch some values within the circuit (if possible) that will allow it to operate using 2n5457's.

Currently the circuit with 2n5457's in seems to have quite a low output, a nicer sound can be produced when the gain and volume knob are balanced at a specific point although the distortion is very harsh (harsher than the circuit intends), and the rest of the attenuation sounds terrible, jumpy and has a weird phasing effect in one area.

[duck_arse]

if you've built it exact to the schem, you have an error. the D1//R4//C10 should be connected to ground, otherwise you have no DC bias on the jfet gate, I think.

also your right hand side jack and the volume pot have no explicit ground connection. a closer look shows Q3 has no DC path to ground for the source, as well.

[Fancy Lime]

Hi monkey2410,

you can almost always just replace J201s with 2N5457s. It will sound different and require some re-biasing (just as using different J201s would) but it should work. But the schematic you posted isn't going to work with either of them. As Stephen pointed out, there are several things wrong with this schematic. The missing connection from source of Q3 to ground seems to me most likely to cause the problems you speak of. I have no idea what's going on with the power filtering but I'm pretty sure there is something wrong. As for the output jack: If you did in fact connect it the way shown in the schematic there should be no sound at all coming out. The input jack is also connected the wrong way around.

Where does the schematic come from? What is it called? How do you know how it is supposed to sound?

Cheers,
Andy

[monkey2410]

your 100% right!hahah the schematic there is what i drew from the vero board below to try make it simpler for myself when breadboarding but i can see now from the diagram below that it needs grounding points exactly where you said.

After correcting the pedal, it mildly works but not much better, there is a constant very high pitch signal, probably somewhere around 15-20khz. Do you think it is possible for me to still make the pedal work with these transistors? And i realise I have the input jack the wrong way round I didn't intend for that, I wired it correctly on my breadboard.

[Fancy Lime]

Generally: yes, it should be possible to get it to work with 2N5457s. But: You need to adjust the biasing. This is different from individual transistor to transistor and even more so between different transistor types. JFETs are particularly variable with this respect. To adjust the bias, you need to adjust the drain resistors (R2, R5 and R9 in the vero layout), so that the voltage at the drains of the transistors is correct, usually around half supply voltage.

I suggest you do this one stage at a time. First build only the first stage on bread board (and make it go to the output jack) and change the drain resistor until it is loud and sounds right. Then add the second stage and repeat, then the third.

It would really be easier to help you if you could tell us what the thing is called and where your info comes from.

Andy

[EBK]

It would really be easier to help you if you could tell us what the thing is called and where your info comes from.

Looks like a Wampler Plexi Drive to me.

[monkey2410]

Yes its a wampler plexi drive  and thanks for that andy I'll have a go it tonight/tomorrow. I'm not very understanding of how JFET's work to be honest but i have a little knowledge of BJT's which will hopefully help (although i'm aware JFET's are voltage controlled rather than current). When you say to do this one stage at a time do you mean with just gate and drain resistors and no source resistors or feedback loops or anything like that (other than when moving on to the second & third stage) ?

[GibsonGM]

In your schematic, that would be input up to C3, as one stage.  You bias that to a working condition, and then move on to the next stage...most use a trimmer (or pot) to temporarily adjust the drain resistance.  You can do it by ear - you will know when it sounds right

Then remove and measure the trimmer (or pot), place a fixed value resistor there.

[monkey2410]

When i'm on stage 2 where should i be going up to? i can do stage 1 easy but i'm pretty sure im doing stage 2 wrong because im either getting a very low max output, horrible feedback from gain pot or in most cases no output at all. I think the gain pot is the main problem i might be having but ill let you guys decide.

To be clear, I have changed the R2 value to a 8.2k resistor so ignore the original 15k value in the schematic.

[duck_arse]

you will need a DC blocking cap between the Q2 drain and the outside world.

[GibsonGM]

When i'm on stage 2 where should i be going up to? i can do stage 1 easy but i'm pretty sure im doing stage 2 wrong because im either getting a very low max output, horrible feedback from gain pot or in most cases no output at all. I think the gain pot is the main problem i might be having but ill let you guys decide.

To be clear, I have changed the R2 value to a 8.2k resistor so ignore the original 15k value in the schematic.

How about you get the first stage done, then add the entire rest of it...use the same value drain resistor as 1st stage, and if it's wonky use a trimmer to dial that one in, too...

Duck is spot on, without an output cap you'll have DC going out, which will sound like crap.  Try 1u or something, adjust to taste after it's working.

Forgive me if I'm not seeing something, but why is the gain control done that way??   Why not set the pot up as a divider, one lug to ground with wiper feeding the gate of the next stage, as is the usual?   The way it's drawn seems like a feedback control sort of job...maybe I'm missing how great it is that way (?)  C10 will allow (negative) AC feedback.

The other way to do it is make the source resistor a pot to adjust gain...but maybe you are working from something that is done this way, I dunno, just seems an odd way to adjust gain (outside of an actual amp etc).  Anyway, if you keep that, be sure of your connections - you could get all bungled up in that area, which could lead to what you describe 

[monkey2410]

Hi, I've just tried that out with a few different values from 22nf up to 22uf but i'm still having the same problem. It provides terrible feedback somewhere roughly between 30k-90k and the rest is virtually silent or clips.

[duck_arse]

pull C5 out of circuit, see if that tames it some.

[monkey2410]

sadly not 

[KarenColumbo]

Is it just me, or are there no coupling caps at all in the circuit?

[GibsonGM]

Is it just me, or are there no coupling caps at all in the circuit?

Well, C2, C3, C4 are acting as such....no output cap (yet)...don't need an input cap, but it would be good practice.  All that stuff up by the diode seems to me to be a lot of busy-ness, but perhaps I'm missing something?

[Fancy Lime]

Well, if you need an input cap depends on what you feed it with. If there is only a guitar in front of the input, you don't need it because there is no DC component. But if other electronics come first, there might be some DC. There shouldn't, but there might.

The whole power supply filtering business by the diode looks super weired in the schematic but I think it is correct. Not the bes way to do it, imho, but not wrong.

If the thing still oscillates and does not bias even with input and output caps added, I'm going to hazard the guess that there is something wrong with the build, not the schematic. I've had these cases where I just could not find the error, schematic was correct, build was according to schematic, did not work. So I took it all apart and reassembled it and: TADAA, it worked. Both on breadboards and on perf- or veroboards, even on PCBs, you can get strange things like parasitic capacitors between to solder blobs that are too close to each other (even without touching). Or creep currents through or over some part of the board that should not be conducting but is conducting a little bit because there is some dirt on it. The possibilities are endless. Sadly.

Don't be disheartened and keep trying. It's part of the learning experience and all (or at least most) of us went through the "what the hell is it now?" phase.
Andy

[Cozybuilder]

I traced the vero layout, this is the resulting schematic. This looks like it should work, so I'd suspect wiring error, or perhaps a cut was missed?

[GibsonGM]

That should work.  That's the problem with vero....IMHO there is no substitute for doing 'each connection by hand'.   But for some, that's what they like.

If you can follow the schematic...have you considered perf board?  Doing it based on your schematic WILL let you isolate sections so you can be sure they work.  R2, R5, R9 could be trimmers/pots until you have dialed each section in.

Fancy Lime: I know an input cap isn't required (amps don't have them....), it's just good practice.  If another effect leaks DC, it gets into your circuit.     If I had my way, amps would have a 100V rated cap up front, lol.  I've sent more than 20V AC into them, usually by mistake. 

[monkey2410]

Guys i've finally got it working! Big thanks to you all!! The schematic you drew was correct Cozybuilder and a lot simpler to follow than the the one I did. If anybody wants to try this pedal out for themselves I used an 8.2k resistor for R2 and the rest is all the same, must say I'm impressed with the sound and operation! I've not put it down on a PCB yet though so lets not jinx it! There is a bit of DC hum but this may just be because its breadboarded and hopefully that cleans up a little when its stuck on a PCB??