Ultimate Octave/Foxx Tone Machine

[Hexjibber]

Hi all, hope you're keeping well!

Latest build is an Ultimate Octave on Fuzzdog board. I built one of these ages ago in a 1590B and always liked it, the foot-switchable octave layout of this kit appealed to me tho. I realise I could have just done it myself on a regular Foxx Tone Machine board but sometimes I'm just lazy!

Started out with a stock build but I'm planning on trying some of the mods from the Aion Vulcan project. With this build though I was getting below unity gain output which I don't recall getting in my original build. After checking parts and voltages etc with no issues found a suggestion to reduce the Q4 emitter resistor, auditioned a few but opted for a 270R in place of the 1k5 and now it's nice and loud. Found quite a few forum posts with the unity gain issue so seems to be common but no-one can seem to define the cause consistently, weird. Any ideas?

Anyway, I'm happy with it now, currently trialling BC108s but plan to try 2N5088s as per the Aion suggestion and see if more gain equals betterer!

Had a nightmare trying to etch this design and after 3 failures threw in the towel! Tried my first waterslide decal on this super pink box I've had knocking about, pretty happy with the end result. The powder coat itself on the box is rubbish though, chips really easily, can't recall where I got it from now but a bit disappointing.

Sorry, enough chat! Pics!

P.S. I'll actually post a PedalPCB build one of these days! Tyrian is next on the list!

 

[music6000]

One Word, Awesome !

 

[Bret608]

This looks so cool! That is funny about the unity gain issue. The one I built several years back was super-loud (unity at 9:00), and it did have the 1k5 emitter resistor.

One thing I'll say about transistors for this pedal--I personally believe it likes mid-gain ones best. Many of the originals used 2n3565. The electronics lab at the college where I work had quite a few of these still knocking about. I tested quite a few and the gains tended to be in the 200-400 hfe range, with most being 250-300. In other words, 2n3904 is the closest modern equivalent. The one time I saw someone over at the Madbean forum try out higher-gain transistors (new-production PN3565 with 500-ish hfe), he said he got oscillation at max gain. So if you do try 2n5088, be sure and try something lower gain too. I would imagine those BC108 you have in there now sound really good though.

 

[music6000]

2N4401 & 2N3904 are very similar with 2N4401 slightly less noise.

 

[benny_profane]

I just finished up an Aion Vulcan (Tone Machine) build the other day. I used 2n5088s (I think...maybe 2n5089s) with gains of 440–450. No issues with oscillation and it works great.

The updated Aion document has added to the discussion about transistor selection. It’s worth checking out:

EDIT: Unity volume (though of course dependent on the octave switch, tone shaping, and gain position) is well below noon on my build. It has a lot of available output.

https://aionelectronics.com/wp-content/uploads/2019/02/vulcan_documentation.pdf

 

[Bret608]

That is really interesting! I didn't know that the later 2n3565s were higher-gain. I'm hoping to build another one of these soon, so it seems like experimentation could be in order. This circuit does pretty much kill either way. I do hope I can find a decent pink box somewhere, though!

 

[Hexjibber]

This looks so cool! That is funny about the unity gain issue. The one I built several years back was super-loud (unity at 9:00), and it did have the 1k5 emitter resistor.

One thing I'll say about transistors for this pedal--I personally believe it likes mid-gain ones best. Many of the originals used 2n3565. The electronics lab at the college where I work had quite a few of these still knocking about. I tested quite a few and the gains tended to be in the 200-400 hfe range, with most being 250-300. In other words, 2n3904 is the closest modern equivalent. The one time I saw someone over at the Madbean forum try out higher-gain transistors (new-production PN3565 with 500-ish hfe), he said he got oscillation at max gain. So if you do try 2n5088, be sure and try something lower gain too. I would imagine those BC108 you have in there now sound really good though.

Really odd how the output on this seems to vary between people's builds, I read one suggestion that it was actually amp related but I'm running a Twin Reverb so would expect with that amount of headroom I'd get maximum volume? I'll have to try it with one of my more gainy amps just to check I suppose!

I put some 2N5088s in and I like the sound actually, didn't get any oscillation and it seemed to give it a warmer tone than the BC108s. It still doesn't kick out as much level as my other gain pedals like the Angry Charlie, Sonic Titan, Musket Muff, Fuzz War tho so am at a bit of a loss. Also not sure if I'm just remembering my previous build incorrectly but with the octave engaged its very thin and clanky unless I completely change the tone settings, then at that point it's pretty useless having the octave on a footswitch! I mainly use it for fuzz so no big deal but still a slight disappointment.

 

[Hexjibber]

I just finished up an Aion Vulcan (Tone Machine) build the other day. I used 2n5088s (I think...maybe 2n5089s) with gains of 440–450. No issues with oscillation and it works great.

The updated Aion document has added to the discussion about transistor selection. It’s worth checking out:

EDIT: Unity volume (though of course dependent on the octave switch, tone shaping, and gain position) is well below noon on my build. It has a lot of available output.

https://aionelectronics.com/wp-content/uploads/2019/02/vulcan_documentation.pdf

That's interesting, I've just had a relatively quick check between the schematics of both versions, there are quite a few parts differences, the most obvious being all the 10u caps in mine are 1u in the Aion version. I notice though that the PedalPCB one uses 10u, so perhaps that's a red herring? I've got the pedal in a state that I'm more or less happy with but still, this sort of thing annoys me!

 

[Bret608]

I think Aion's version replaced the traditional boatload of 10uf electro caps with 1uf film, similar to the Madbean version. So I think Fuzzdog and PedalPCB's are more in line with the original schematic. I know what you mean about the footswitchable octave--I did that on mine and I had maybe this one setting where you didn't have to switch things up between the two. Sorry to hear about the volume issue!

 

[benny_profane]

I think Aion's version replaced the traditional boatload of 10uf electro caps with 1uf film, similar to the Madbean version. So I think Fuzzdog and PedalPCB's are more in line with the original schematic. I know what you mean about the footswitchable octave--I did that on mine and I had maybe this one setting where you didn't have to switch things up between the two. Sorry to hear about the volume issue!

This is right. The 10uF caps are AC coupling capacitors and are not meant to affect the signal. There should be no appreciable difference between the two.

 

[bobson dugnutt]

Awesome build! I'm messing around with one I just built and the diode choice seems to make all the difference; it's not a huge fan of the D9Es I tossed in (they're the only GE ones I have kicking around). What diodes did you use for the clipping and octave bits of the circuit?

 

[Hexjibber]

Awesome build! I'm messing around with one I just built and the diode choice seems to make all the difference; it's not a huge fan of the D9Es I tossed in (they're the only GE ones I have kicking around). What diodes did you use for the clipping and octave bits of the circuit?

Thanks! I used 1N34A in this build, perhaps I should have socketed them and experimented though!

 

[bobson dugnutt]

Thanks! Gonna give those a shot

 

[Chuck D. Bones]

Looks great!
For the octave diodes, you want low Vf, low leakage and preferrably matched Vf. Some people use Schottky, such as 1N5817. I used BAT46 in my Vulcan, per the build doc. 2N5088 and 2N5089 are good choices because they are low noise and have more than enough HFE. I used D2E for my clipping diodes, but feel free to experiment. Ge diodes have low Vf and that will reduce the output volume. Schottky diodes are in the same ballpark as Ge. 1N4148 or 1N400x will provide more output, but it sounds like you have already solved that issue.