Mu Doubler - Geofex

[lowell]

Anyone build this pedal? I just finished my first one. I built the jfet matcher and matched 2 jfets for the differential section. I am also incorporating the millenium bypass too... It is doing something, sounds like even order harmonics, however the effect is much thinner sounding than the straight buffer signal and a kind of ugly distortion too... i.e. when I turn the mix pot to ALL effect it sounds thin and when turned all down to just straight buffered signal it is a fatter sound w/ more lows. I know it's not wired backwards because the effect is only really happening when it's up all the way and thin sounding. The thing has a fairly unfavorable distortion going on too.

What might cause this?
Is something wrong or is this what others who have built this sound like too?
If it sounds like something's wrong what might I check? What should the transistor voltages look like? I am getting what RG says in the article on the sources of the differential - 1.5v.
I have gone over it a lot and the connections seem correct, maybe I fried a component or jfet when I first fired it up?

[lowell]

I just realized I missed something in the article.... Q2 must be matched to Q3. I thought only Q3 and Q4 must be matched... oops! This is probably why. Will report back w/ results.

[52 Bill]

It is doing something, sounds like even order harmonics, however the effect is much thinner sounding than the straight buffer signal and a kind of ugly distortion too...

I haven't built this circuit, but I think that it is meant to be an octave doubler fuzz, which will not give you a smooth distortion.

It will give you the sound of playing one octave above the input signal, when you feed it a sine wave. So play single notes above the 10th fret with your neck pickup and see how it sounds.

Also try using your bridge pickup, and play single notes below the 7th fret, while picking close to the bridge and see if you can get a raspy sitar-like sound.

[Mark Hammer]

Actually, the intent was to provide a cleaner octave than typically is found. Exactly how clean, I couldn't say.

[lowell]

Ok guys I matched Q2 to Q3 and Q4 and the pedal still sounds.... gatey I guess you might say.

What might lead to the tone being so thin when the effect is on?

[R.G.]

It's been over six years since I did that one. In retrospect, it may have a bug in the drawing. I've tried to remember the particulars and haven't dredged them up yet.

If I were doing that today, I'd put a 1uF cap between the source of Q2 and the 100K blend pot, and move move the 0.1uF cap that attaches to the gate of Q2 so it goes between Q2 gate and Q3/4 drains.

Give that a try and let me know how it works.

[lowell]

Hey thanks for replying! Nice... ok i'll try that and report back. cheers.

[lowell]

RG..

Ok I added/moved those components and it definitely fixed the "thin" problem of earlier and the distortion is more subtle. However, the remaining problem I'm having is that the "effected" signal is very quiet comparatively speaking. When the pedal is bypasses it's almost twice as loud. I double checked the sources of Q3 and Q4 and they are still the 1.5v. Any ideas as to volume loss or is this a side effect of the circuit? Also, how might I go about increasing the effect?

[R.G.]

It's a side effect of the circuit. It probably needs another gain stage before the 100K blend pot.

Here's how that works. Every active device has some inherent distortion. The kind of distortion depends on exactly how the device works internally. JFETs happen to have a distortion that depends on the square of the input signal. Q3 and Q4 form a special differential amplifier. The signal is split into two equal-but-opposite halves and fed one to Q3 and one to Q4. Then the resulting signal is added at their drains by connecting the drains together. What that does is to cancel out almost exactly the actual signal as amplified by Q3 and Q4. But the even-order distortion is not cancelled - it adds at the drains. Q2 then buffers this signal.

Unfortunately, the amount of distortion in your JFETs isn't all that big, so the majority of the signal that's left, the distortion, is very small. That's what's giving you such low signal.

The simplest thing to do is to put a JFET gain stage after Q2 source and before the 100K mixing pot. Here's the paper-dolly way to do it:
1. Print another copy of the schematic. Change the device number on Q2 to Q5 and Q3 to Q6.
2. Right between the "M" and the "u" in the title, cut the sheet of paper straight down just to the right of the two stacked 10K's.
3. Just to the right of the last "r" in "Doubler" in the title, cut the sheet of paper straight down, leaving the new Q5 intact.
4. Now cut away the 47uF and Q4, leaving the new Q6 alone.

This circuit fragment is a high gain amplifier, with signal going in at the gate of the new Q6 and coming out of the source of the new Q5. To insert it into the old circuit, open the connection between the new 1uf cap from Q2's source and the 100K mix pot. Connect the 1uF to the gate of the new Q6, and connect the source of the new Q5 to the 100K mix pot. This should give you quite a bit more signal, maybe too much. But we can cut it back if it's too much.

Did I muddy that up too much?

[lowell]

RG,
That's more help than I hoped for. Do I ditch the 1M resistor on gate of new Q5 or does that get grounded? Thanks very much!

[R.G.]

Oops, sorry. That 1M goes to the same 10K/10K/capacitor bias voltage as the one for Q2 does.

You're replicating Q2/Q3 as a gain stage, but without Q4 to cancel the main signal voltage. It should have a lot of gain. This is an SRPP stage, a so-called mu-amp, which has gain to burn. You may have to increase the source resistor on the new Q6 to lower the gain.

[lowell]

Oh ok cool thanks!

Aside: What is an SSRP circuit?

On another note any chance you can help me repair an EH Blackfinger?.. the version w/ the 748 op-amp in it? I brought the thing back to life but the sustain knob does not work. Let me know and I can start a thread for this. thanks again!

[R.G.]

Aside: What is an SSRP circuit?

Series Regulated Push Pull. It's a tube circuit going back at least to the 1940's that has been converted over to JFETs. A fellow discovered it in a National Semiconductor application note and renamed it the "Mini-Booster" for the effects circle until the National app note was found.

On another note any chance you can help me repair an EH Blackfinger?.. the version w/ the 748 op-amp in it? I brought the thing back to life but the sustain knob does not work. Let me know and I can start a thread for this. thanks again!

Actually it would be better if you posted over at the diystompboxes forum. The interchange would reach a lot more people there and be more instructional.

[lowell]

RG,
Ok I added the SRPP circuit after the "effect" signal per your description. I think it helped, however I'm not sure how much... should it be plenty loud?... comparitively to the bypass signal?... cause it's still quite a bit softer than when the pedal is off/bypassed. Maybe I wired it wrong. Let me know if you think it should be sufficiently loud now.