So, ever since I learned that these inverters we use for some of our pedals can deliver 5 to 25mA output, I wondered if it'd be possible to make a headphone amp out of them.

And today, after not being able to sleep, I decided to see it for myself.



After tinkering a bit with the circuit, I ended up with the following schematic:



The trimpot for adjusting the bias for the output stage is actually an idea from Tim Escobedo's Circuit Snippets, specifically from the Calavera circuit.

I think it could do with a filter at the output to kinda "simulate" a speaker cabinet, but I'm terrible with filter stages. Wouldn't even know where to start tbh.

[Edit]

Sound sample:

https://youtu.be/8VQJ2Yj2IJ8?si=pFdULSaG7aHjAICO

v2



I figured I could use a 10k trimpot with a high value resistor from the taper to do the bias adjustment.
I also figured that I have like a half dozen of those 4M7 trimpots with the red knob, and that I'd like to use them on something, and they worked when I put them on the circuit, so I'm gonna use them in the final build.

Also V+ resistor reduced from 220R to 68R for even better driving output capabilities.

Sounds miles better than a 386!

Love seeing weird stuff like this - just saw somebody making a sketchy audio amplifier from a 7805 and shocked that it worked too 

Quote from: drdn0 on March 18, 2025, 06:30:21 PMSounds miles better than a 386!

Love seeing weird stuff like this - just saw somebody making a sketchy audio amplifier from a 7805 and shocked that it worked too 

Another fun one is this one with a TL431:



I remember at one point, on some other post (I think on this very forum), there was a project for a booster using the same IC, but seems to have been lost to time.

https://www.diystompboxes.com/smfforum/index.php?topic=97109.0

user tca did some designs using the TL431. I don't know if the images will work in that thread, but I've got copies if needed.

Quote from: duck_arse on March 19, 2025, 10:02:06 AMuser tca did some designs using the TL431. I don't know if the images will work in that thread, but I've got copies if needed.

YES That's the one!
If you could PLEASE add those images to the original topic it'd be tremendous (to not derail this topic and to make it easier to find in the future).

Thank you!

Since there's volume to spare on headphones and earphones, I've decided to take an idea from the earliest Runoffgroove designs of a "cab sim" at the output of their many JFET amp emulator circuits, using a pair of RC low-pass filters. There's an alternate output (amp) bypassing the RC filters for use as a pedal.

I've also drawn a layout that I plan on fitting on a Mentos CleanBreath enclosure (the thin plastic ones).

I am curious about the four paralleled inverters in the output.  Would they not need a feedback resistor to keep them in the linear range?  Otherwise you would just get high / low levels at the output.

Quote from: amptramp on March 23, 2025, 10:06:00 AMI am curious about the four paralleled inverters in the output.  Would they not need a feedback resistor to keep them in the linear range?  Otherwise you would just get high / low levels at the output.

I might be wrong, but from what I understand, the trimpot is used to set a bias voltage such that the inverters are kept at a linear operation point between open and closed state.

Quote from: blackcorvo on March 24, 2025, 12:00:04 AMI might be wrong, but from what I understand, the trimpot is used to set a bias voltage such that the inverters are kept at a linear operation point between open and closed state.

It does but it's far less reliable than a feedback resistor, or possible a feedback resistor and that trimpot.   If the output amplifier has a gain of 20 and you want to adjust the output within 0.1V then you need to adjust the trimpot to within 0.1/20 = 5mV.    With a 9V rail on the trimpot that's not going to happen.

The problem with the circuit as is the the very high output impedance from the tone control (4M7 pots).   If we tried to add a feedback resistor then the input impedance to the output amplifier would be much lower than the trimpot design.   It's kind of painted itself into a corner.

If you went back to the original 100k pots, perhaps with some part tweaks to behave the same as the 4M7 pot version, then it would give you some chance to add a feedback resistor (perhaps a 10M feedback resistor).

That's just the thing, I don't want to add a feedback resistor. It works as is and I like how it sounds.

Feedback resistors increase the gain too much to my taste, and make the sound go into a flubby mess when you go too high (which is why I limited the first stage with 1M and lowered the second to 680k).

And I'm using 4M7 pots because I have a handful to spare, which came from a bulk kilogram of components purchase, and I'll pretty much never use them in anything else. Plus they're perfect for something that could go in your pocket, cus they're essentially 14mm trimpots with a knob.
Lastly, the bias trimpot is multi turn, I have a space of like 3-4 full turns where within the output works, so it's pretty easy to find the sweet spot once you get there.

Anyone else who wants to follow your suggestions, go for it. I'm just sharing a project as I'm working on it, with the things I have available.

Feedback resistors reduce the gain, they don't increase it.  Adding one would make the trimpot adjustment less critical.

Quote from: amptramp on March 24, 2025, 09:05:24 AMFeedback resistors reduce the gain, they don't increase it.  Adding one would make the trimpot adjustment less critical.

Alright, I'll try it on the protoboard, but AFTER I'm finished building the current perfboard (I'm halfway through). If it does help with the bias, I'll bodge the resistor onto the board.

Quote from: amptramp on March 24, 2025, 09:05:24 AMFeedback resistors reduce the gain, they don't increase it.  Adding one would make the trimpot adjustment less critical.

Although with the trimpot bias you could set the output to bias at less than maximum gain.

Finished build:



Used a tiny 12v bulb instead of a resistor (because it looks cool when the brightness changes with harder picking).

Tried the 10M resistor on the last stage, barely noticed any changes.

Sample:

Nice build and sound for something so compact! Has the final schematic changed or are you gonna experiment with feedback resistors?

Quote from: pacificplazacircuits on March 25, 2025, 06:28:41 AMNice build and sound for something so compact! Has the final schematic changed or are you gonna experiment with feedback resistors?

Only difference is the added 10M feedback resistor on the last stage.
And that I'm currently using a 4069 (adapted to the 4049 pinout with a couple 8 pin sockets) because I fried my only 4049. I'll get a replacement sometime this week.

Oh, and I put it inside a Barkley's tin.