This Week on the Breadboard: OKKO Diablo v3

[comradehoser]

It's done. Here's the final schematic. The only changes from the schematic above are the Pres. Trim pot, voltages noted on the dwg, the notes and the revision #.

View attachment 91445


A few words about selecting Q1-Q4. The J201 Idss spec range is 0.2mA to 1.0mA. On my breadboard, Q1 & Q3 are in the middle of the range (0.54 and 0.60). Q2 and Q4 are at the top of the range (0.92 and 0.95). These are not magic numbers, but they work well for me. If you have an already built circuit, an easy check is to connect a DMM + lead to Vref and the - lead to Q2-S. Turn DYNAMICS all the way up. The reading should be a negative number, in the ballpark of -200mV. Now clip a jumper across R4. The voltage should still be negative. If it is, you're good. Do the same thing for Q4-S & R13.

Pres. Trim affects the range of the TONE knob and MIDS switch. Pres. Trim has the most influence when TONE is dimed.

All of these mods are easily kludged onto a Deofol board. If you don't do anything else, at least get the biasing right. Check the biasing on Q1-Q4. Delete R20. Move R18 from GND to Vref. Use a JFET with a higher Idss for Q5, like 2N5457.

Have started implementing these changes, but I have some sharing issues with my Figment Preamp build around my limited stock of JFETs in specific ranges, so I'm going to have to test on both boards and make my picks. Luckily I socketed everything and have a handful of through-hole interfet J201s and NOS 2N5457s, but all of my interfet repro j201s are surprisingly consistent in the .38-.48 IDSS range, so I can't get as wide a spread of IDSS as indicated between the odd and even Qs if I stick with j201s.

I can say that putting the higher-IDSS 5457 (tried 1.3 and 2.0) in Q5 does definitely make it sound much more "open" for lack of a better word. Way more "overdrivey" and less "distortiony/fuzzy" especially at maxed gain and feed. I actually haven't decided what I prefer most yet, though. The presence trim I'll add. It looks very doable, even though I'll have a lot of floating stuff.

I lifted R18 and connected to VREF at the closest point (bottom of R6). I had hoped that between this and the changes around the odd and even Qs and the presumably higher headroom at Q5, the changes would mitigate the very odd ducking/swelling behavior of my pedal when dynamic is below 10:00 and gain/feed are maxed. But no.

Just curious--- what does deleting R20 do or don't do?

 

[Chuck D. Bones]

R20 pulls Q5's bias towards GND, which is what we don't want. You may or may not hear the difference when you remove R20 since R18 has a much stronger effect on Q5's bias. R20 is kind of a belt-and-suspenders guarantee that Q5 will have a bias source in the unlikely event that the wiper of the TONE pot loses contact with the resistance element. Since we'll have other problems if that happens, I don't see the advantage of R20. Crappy pots should be replaced, not band-aided.

If you can't find the thru-hole J201s you need for Q2 & Q4, you can try other part numbers. Look for 2N5457 with Idss around 1mA.

You can temp install a PRESENCE trimpot and once you find the desired tone, remove the trimpot and replace R19 with that value.

I seriously doubt that OKKO intended for Q5 to saturate. They probably hand-picked J201s with high Vp & Idss.

 

[comradehoser]

R20 pulls Q5's bias towards GND, which is what we don't want. You may or may not hear the difference when you remove R20 since R18 has a much stronger effect on Q5's bias. R20 is kind of a belt-and-suspenders guarantee that Q5 will have a bias source in the unlikely event that the wiper of the TONE pot loses contact with the resistance element. Since we'll have other problems if that happens, I don't see the advantage of R20. Crappy pots should be replaced, not band-aided.

If you can't find the thru-hole J201s you need for Q2 & Q4, you can try other part numbers. Look for 2N5457 with Idss around 1mA.

Ah, that's very informative on Q20, thank you.

Yes, that's what I've been doing with the 5457s-- thus the conflict with the Figment preamp, which wants a higher IDSS in Q1, and quite low in Q2 and Q3 according to PPCB spec (I think--the PPCB JFET spec categories listed in the Figment build doc don't bear relation to what my JFET tester reports on). Lowest 5457 I have are 1.1 IDSS that I'd want to also use in the Deofol. I've yet to test if it makes a difference in the Figment, though.

And, I have to decide if I prefer more the more overdrivey incarnation of the Deofol or the more distortion incarnation of the Deofol.

 

[Chuck D. Bones]

What are you using to test your JFETs?

I have a hard time making sense of the JFET specs that Robert puts in the Build Docs. I don't have a DCA75, so I don't know how to interpret the displayed results. I can tell you that interpreted literally, they don't make sense. Example: Q1 indicates that Idss is 2.23mA. Idss is the highest drain current possible in a given JFET. Semiconductor physics says so. How is it possible that the same instrument reports Rds(on) at Id = 5mA? Is this a measurement, or is the DCA75 reporting the datasheet pass/fail limits?

What we really need to know are the drain voltages and drain currents in the unit that was traced. When trying to match the JFET specs in the Build Docs, one has to ask: Which JFET specs are important in this circuit? How closely do we need to match them? If we can't get an exact match, what do we do then?

 

[comradehoser]

You, me, and owlexifry. I asked him about jfet selection and the published figures for the figment and he couldn't really make sense of them either.

I have ye olde made in China Amazon special tc-1 tester

 

[Chuck D. Bones]

The CCTT* I have does not measure Vp or Idss. It simply measures drain current at some arbitrary gate voltage between 0 and Vp.

* Cheap Chinese Transistor Tester

 

[comradehoser]

[Edit to remove irrelevant block quotation]
So if I'm reading what's going on in your schematic and the PPCB schematic, we're replacing R19 with a 50K trimpot, right?

 

[Chuck D. Bones]

Is there a reason you quote my previous posts verbatim?

we're replacing R19 with a 50K trimpot, right?

Affirmative.

 

[comradehoser]

Is there a reason you quote my previous posts verbatim?

Hahaha... Kind of-- just so that you'll receive notice that I replied to a comment you made, and I guess for the context of what I'm replying to. In this case, I think I confused R19 and R20, which is why my intent could be puzzling as the context is incorrect. Hold on, I'll delete it.

[Edit: and I usually don't edit my quotations to be precise as I'm lazy, and also the default setting when you press reply is to quote the entirety of the post you are replying to]

 

[comradehoser]

Hey Chuck,

I finished the rest of the mods and my pedal gated hard when I turned down "Dynamic," and stopped working permanently. I did some measurements and found out my VRef was close to 0v--everything else was correct otherwise in VCC land (16v to 3v) and continuity checked out in the power section around everything related to VRef (R100/R101/C104). Continuity and correctness of mods all checked out.

Since R18 to VRef seemed the most directly related to VRef, I reversed the R18 to VRef mod to ground again, and I have my pedal back. Originally I had tied R18 into VRef at the VRef side of R6. I can't figure out why that would be verboten in the circuit. Is there an optimal place to tie into VRef, or should I just stick with R18 to ground?

 

[Chuck D. Bones]

Mine doesn't do that.
On my breadboard, Vcc varies from +5.1V to +17.2V. Vref is always within a few % of 1/2 Vcc.

With TONE at zero, what voltages do you get at Q4-D and Q5-D? I'm accounting for the fact that they are installed "upside-down" in the Diablo/Deofol circuit.