Setback / FX90 volume drop

[Aentons]

Im finishing up a Setback build and yes the thru volume drop is a bit annoying. Is it something that was solved for in later revisions?

I used trimpots for R4 (200k) and R48 (100k) and even maxed out they don't seem to help. Any Ideas?

[madbean]

Aside from increasing that 18k resistor to 47k as suggested, anything further and you will probably need to modify the output section a bit. One thing you could try - jumper R9 and use a 100pF cap for C4. Then use a 68k resistor for R18. If that's too much output reduce R18 until you get the volume you want.

[Aentons]

Thanks for the tips!

I tried your suggestions and used a 100k trimpot for R18 and it does not seem to do much if anything at any position.

The R9 and C4 changes made everything really bright and noisey(hiss). Turning R48 back down to around 18k clears the noise but anything above that is pushing it and the volume drop persists. The brightness stays too.

I'll fiddle around with it some more and see if I can maybe find some values that balance it out.

[madbean]

Thanks for the tips!

I tried your suggestions and used a 100k trimpot for R18 and it does not seem to do much if anything at any position.

The R9 and C4 changes made everything really bright and noisey(hiss). Turning R48 back down to around 18k clears the noise but anything above that is pushing it and the volume drop persists. The brightness stays too.

I'll fiddle around with it some more and see if I can maybe find some values that balance it out.

Right, well you have two resistances in parallel there so just adding lots of resistance to just one of them only gets you so far. Changing R18 to 47k while leaving the 3k3/33n in parallel was what I found to be the best compromise without really getting into altering the circuit.

The trick is to first set the output volume where you want then choose the appropriate value cap that reduces the hiss to a minimum without rolling off too much high end. There may be some perfect combination of values that does this, it's just a matter of finding them. Or, maybe leave that as is and reduce R2 to 390k. A little more at the front end without lowering the input impedance of IC1A too much.

[Aentons]

I got a chance to open up my buddy's FX90 to see if it is a later version and the component placement doesn't seem to match up with the revA schematic & layout diagram. He modded it for true bypass but that was all. Anybody know what revision this might be?

My goal in lookin at it is to see if there changes for how they may have solved the volume drop because this one doesnt have it. It actually has a slight volume boost.

I will try to start tracing down around the TL062.



Also, I did notice a cap on the bottom that is on pin2 of the 1M trim pot.


And here is the front if it helps to identify:


Here is the revA diagram im referencing:

[tcpoint]

Here's a link to the Rev H schematic.  There's a lot of differences.  https://www.madbeanpedals.com/forum/index.php?topic=29170.msg308109#msg308109

[Aentons]

Here's a link to the Rev H schematic.  There's a lot of differences.  https://www.madbeanpedals.com/forum/index.php?topic=29170.msg308109#msg308109

Awesome! Thanks for that.

This is much closer to the one I have.

It looks like the layout diagram is rev H, but the schematic is rev I. (If you look at the IN jack label, the uppercase I matches what is over in the bottom right rev box)

[Aentons]

Here is the modern art masterpiece I managed to scrawl out. It's not perfect but it confirms the the pedal I have does indeed match the input and output sections on the rev I schematic.

The differences from revA are almost all in the feedback loop for the IO opamp

[TeleCrunch]

The volume drop on this project had been a real disappointment, since the stock circuit sounds very good, but way too low in level to be usable. This weekend I simulated the circuit in LTSpice and found a couple mods worth experimenting with to achieve unity or even a boost in volume level:

If you add a much smaller value resistor (470 ohm-2.2k) in parallel to R7/10k the overall output level increases. I used a 1.5k (in parallel to the 10k = 1.3k), which doesn't hit the second op amp with too hot of a signal. Additionally, if you change the Mix pot to a higher value it will also increase the output level for all delay mix levels without changing the width of adjustment too much. I used a B25k, which in combination with the R7 change made the pedal completely gig-able with a few dBs of signal boost overall. With these two changes, the build doc's R48 18k to 47k swap isn't necessary. I thought the R48 swap changed the character of the stock pedal too much anyway by adding a lot more low frequencies with only getting a modest gain in output level.

The one trade-off I've found with the above mods is that the sweep of the Mix control changes, so the lower delay mix levels are a little harder to dial in, but I normally use delay levels set pretty high, same or louder than the clean signal, so it's not an issue for me. With the component values I chose, I don't hear any increase in distortion by hitting that last opamp stage with more signal, but I use a Tele with passive vintage output pickups, so YMMV.