Issue with adding an extra gain boost to Jackson 1484 Twin Twelve pedal

[kenjib]

Hello everyone,

I am building a modded clone of the Jackson 1484 Twin Twelve pedal using veroboard and based on this schematic:

https://docs.pedalpcb.com/project/Argentum-PedalPCB.pdf

I wanted to add an extra gain stage that would essentially allow me to switch between this version and something similar to the higher gain JHS Twin Twelve version, which is an almost identical circuit, based on this schematic:

http://pedalparts.co.uk/docs/TwoDozen.pdf

So what I did was add a switch to insert the third gain stage of the JHS pedal as an extra gain boost option. This is what I did -- it is the Jackson/Argentum schematic with the modded addition highlighted in red:



Here is the vero layout:



The lower gain circuit seems to be working correctly. However, when I switch to the higher gain circuit the tone gets really thin, almost like a high pass filter is being applied somewhere. Here is a sample -- first with low gain, and then switching the high gain on:

https://soundcloud.com/kenjib-364738420/1484-issue-m4a?si=c32cedf4a66f4daf9f28f4e0bfa81389&utm_source=clipboard&utm_medium=text&utm_campaign=social_sharing

I did proper biasing to @9V (due to 18V charge pump) with the trim pots -- and even played around with the biasing a bit extra to see if it helped. I tried adding a 10n cap in parallel with the 2.2n C2 to increase the value there in case that was causing filtering. I also tried shorting out R11 since that is one of the only other differences between the two circuits that I can find. I also built a complete second copy of the pedal to make totally sure that it wasn't a build error and get the same result. None of these options fixed the issue. Is anyone able to see what the problem is causing the tone problem when I engage the gain boost switch?

Thank you!

-Kenji

[Matthew Sanford]

I'll chime in but please take any advice from me with a gallon of salt

I didn't really hear much different in your 14 second recording, but according to the extra gain page you should only use the 7660S (well, they said 7660SEPA, typo? maybe 7660SCPA?) or you will add high pitched whine to it. The other thing they mention is to bias the first two to 9v then remove another 0.25v from them, and the third should be biased after by ear. Now on it the 10n/100k should be a HPF with cut off below 20Hz - I first thought maybe a bad solder on a resistor that follows could change  that, but you said you've built two and both behave the same. Other thought is maybe bad biasing there is causing it to reflect resistance to ground in parallel with that 100k resistor reducing it's value and raising the cut off frequency.

Now again, I'm spitballing a bit and do not take my thoughts as correct here, but maybe some things to look at.

[kenjib]

Hi Matthew. Thank you!

I think the recording quality just isn't capturing it well. Here is a better quality sample so you can hear it better:

https://soundcloud.com/kenjib-364738420/01-240229-1603a?si=a008daec6a2241b281510138ec1ed3df&utm_source=clipboard&utm_medium=text&utm_campaign=social_sharing

All the bottom end disappears and it gets a lot more fizzy.

I am using a 7660SCPAZ with the freq. boost pin connected to pin 1 so luckily there is no whine. Thanks for pointing that out. I have messed with all kinds of bias settings and can't get the problem to go away. I did start with it using the 9v minus 0.25v at first but I think it is actually a little fizzy that way so it seems to sound better biased a tad higher. Either way it has the problem.

After your suggestion I tried swapping out the 100K at R5 with both a 470K and 2M (and adjusted bias) to test your theory about the cut off and it doesn't seem to make much of a difference unfortunately. I also used a multimeter and get 100k with the boost switch both on and off.

[ElectricDruid]

Assuming that an inadvertent highpass filter *is* what we're looking for, it would help to know which cap is causing it.

You said you tried sticking 10n in parallel with C2, which is a good start. Grab a 100n and try sticking that in parallel with C2 and then C3 and see if the bass comes back.

If "C3" is the answer, then Matthew's theory might be close - if any current flows gate to source on Q4, then (to the extent that current flows) that R4/47R is in parallel with the R5/100K and the effective resistance goes through the floor, while the cutoff goes through the ceiling!

[duck_arse]

your circuit notations don't match your layout notations around that boost switch. circuit shows "3" and "6" shorted, but layout notes say "5" and "6". layout boost "2" seems to be connected to a 2n2 cap from boost output, whereas schem shows boost "2" connects to the clean output 10nF.

this is why we always want to see photos of what you have built. this is why we should never have to ask to see the photos of what you have built.

[amptramp]

At the output of the added amplifier, you have C2 = 2.2 nF and the series connection of R6, R7 and the volume control gives you 124 K.  This gives a highpass turnover of 583.7 Hz that isn't there when the extra amplifier stage is bypassed.  This is definitely going to be noticeable as the bass being suppressed and the remaining sound being thin.  You need a larger C2.

C3 and R5 give a turnover of 159.2 Hz but the 100 K gate resistor has the drain resistor of the previous stage in series with it, so this will be pulled down as the resistance in the drain gets higher.  It would be a good idea to increase R5 from 100 K to a higher value.  In the circuit, 1 megohm would have no effect on DC bias, so I would do that.  You would then get 15.92 Hz which is well below the lowest frequency a guitar would produce (82 Hz).

[kenjib]

Okay so I tried the 1M resistor at R5. I don't think it fixed the filter problem but I think it seemed to boost the max gain quite a bit, which was really nice (maybe I am mistaken?). So next I added 100n to C2 and now it sounds great! I guess adding 10n just wasn't enough.

Thanks to all of you. Your recommendations were all spot on right away! I'll post some pics and audio later when I get it boxed up and double check the sound on my big amp in the garage.

your circuit notations don't match your layout notations around that boost switch. circuit shows "3" and "6" shorted, but layout notes say "5" and "6". layout boost "2" seems to be connected to a 2n2 cap from boost output, whereas schem shows boost "2" connects to the clean output 10nF.

this is why we always want to see photos of what you have built. this is why we should never have to ask to see the photos of what you have built.

Sorry about that. It's actually just because the pinouts are different. I should have changed the pin numbers on the schematic. The DPDT in KiCad by defaul does this for the schematic:

1  4
2  5
3  6

...but for the vero I used the DPDT layout from the datasheets:

1  2
3  4
5  6

My bad! They are actually exactly the same in the finished result. I will make sure to post photos of the board next time. Sorry about that. I can sometimes be very long winded so I was trying to not overload you guys with everything.    Lesson learned!

[kenjib]

So I ended up going back to the 100k resistor for R5. I felt like I was getting more sputtery/velcro issues with the 1M resistor but it's hard to tell with all of the bias variables. Either way, I am getting a good result with the original 100k so the only change I made in the end was increasing C2 from 2.2n to 100n.

Here is what it sounds like. The first four bars are with the pedal off. The second four bars are with the pedal on but the extra boost bypassed. The rest of the audio clip is with the boost engaged:

https://soundcloud.com/kenjib-364738420/twin-twelve-test

A pic of the finished pedal is attached. Thanks again for the help everyone!!!

[amptramp]

Sounds good!  I don't think you need to make any more mods on this one.