Debugging help - Torchy Phase 45 JC Maillet mods

[Neko Neko]

Hello

I've been lurking on these forums for years but I finally posted needing some help. Previous experience is limited but have successfully built a few pedals & a full modular analogue synth.

I've built the Phase 45 with JC Maillet Mods four univibe sounds & have some problems with this tricky build. I have used the search function & gained some insight but not enough to fully debug the pedal.

1.What does it do, not do, and sound like?

I can hear a distorted signal with tiny amount of phasing, very low gain. Rate LED works Mix pot changes sound between univibe & Phase 45, Bias pot does work as intended & helps tune in the phasing LFO sounds. I can hear the LFO sometimes.

2.Name of the circuit = Torchy Phase 45 JC Maillet mods
3.Source of the circuit (URL of schematic or project) = http://www.neko-neko.co.uk/MXR_Phase_45.gif
4.Any modifications to the circuit? No
5.Any parts substitutions?

I used MPF102's as the phasing trannies & 10k pot for Mix pot.

6.Positive ground to negative ground conversion? Negitive ground

Powered by power supply, all power voltages correct.

IC1
1 - 6.84
2 - 6.07
3 - 5.53
4 - 0
5 - 4.95
6 - 5.62
7 - 8.43
8 - 9.12

IC1
1 - 6.84
2 - 6.07
3 - 5.53
4 - 0
5 - 4.95
6 - 5.62
7 - 8.43
8 - 9.12

IC2
1 - cycles 0 - 8ish
2 - cycles 4.5 - 5.5
3 - cycles 3.3 - 5.5
4 - 0
5 - 4.79
6 - 4.86
7 - 1.3
8 - 9.12

Q1
D - 4.95
S - 4.79
G - 2.10

Q2
D - 4.79
S - 4.79
G - 4.80

D1 = 4.80

Obviously there is a problem surrounding the zener Diode as its supposed to be 5.1v, i have changed it but didn't change anything. I have checked for shorts but no change. This is where i'm stuck I'm not sure which parts of the circuit is in need of attention. Any help would be appreciated. Thanks in advance.

[Seljer]

The 4.8V from the diode is fine. A zener diode needs more current flowing through it before it approaches its rated voltage. In a stompbox you usually want rational current use so you don't want to waste excess amounts through the zener diode to get up to the full rated voltage.

The odd thing is the different gate voltages between q1 and q2 (the both should probably be in the range of Q1), so something around q2 may be iffy (R16 maybe?)
That and the voltages of audio sections IC1 and IC2 are all over the place!
what opamps did you use?

edit: something is making pin 7 of IC2 go down. It may be related to something around q2.

[Neko Neko]

Thanks so much for the help Seljer

I used NE5532 as in schem.

ok i understand about the zener now so thats ok. I debugged the Q2 area & found that the MPF102 was dead. The gate reading for Q2 is inline with Q1. Bias pot working better now. The circuit was in the state before but must ahve killed Q2 during resoldering checking for shorts.

Now for the op amps. IC2 pin 7 is the same as before. Going through the audio lines logically everything looks ok unto the op amp in IC2, I fear it may be dead & it isn't socketed (doh). Going to try a TL072. Will report back.

[R.G.]

I used MPF102's as the phasing trannies & 10k pot for Mix pot.

Did you do any matching on the JFETs? If you used my JFET matching circuit, what were the voltage readings on them for Vgs?

IC1
1 - 6.84
2 - 6.07
3 - 5.53
4 - 0
5 - 4.95
6 - 5.62
7 - 8.43
8 - 9.12

Odd stuff here. Is R2 **really** supposed to be 22K? The reason I ask is that this forces a DC gain of 1+ 10K/22K = 1.45 on this section. The input voltage at pin 3 is then amplified and the opamp tries to make it 8Vdc on the output. Just offhand, I'm not sure how it gets to 6.84V, but I believe that must be all it can do, given the power supply.

I would pull up R2 and insert a cap in series with it. This would at least force this stage to have DC conditions that would let it amplify. 0.1uF is enough for a guitar's 80Hz, but 0.22uF or more would make that a bit wider. Use either one for testing. Just lift one end of R2, solder one lead of the cap into the hole R2 just vacated, then solder the free leads of R2 and the cap.

This should correct the DC conditions a lot, and may make sound come through.

Are you using an NE5532? Those things need a lot of bias current and may act odd with 470K bias resistors like R5. Not that they can't work, but R5 may need to be a lower value.

I also find it odd that pin 3 is at 5.53V, not the same voltage as the top of zener Z1. But the DC amplification stuff needs fixed first, I believe. IC1A having the wrong DC on its output messes up the DC conditions on the rest of the DC coupled stages, IC1B and IC2B. That in turn may mess up the biasing on the JFETs that lets them act like phasing resistors.

[Neko Neko]

Ok I replaced ICB with a TL072. No difference so ICB is ok.

The pedal now acts a a tremolo type sound. I can hear subtle phasing but more tremolo type sounds. To me this says the original signal is not amplified so does not mix with phased side to create full phasing sound.

I did not match the trannies as read on the forum that random MPF102's worked well & didn't need to be matched.

[PRR]

> Is R2 **really** supposed to be 22K? ....that this forces a DC gain ... voltage at pin 3 is then amplified and the opamp tries to make it 8Vdc on the output.

Way I see it: R2 network returns to the same 5.1V reference as the "+" input. Differential input is zero. Output should be steady 5.1V.

Agree that the reported numbers say otherwise.

Might just remove (lift one leg of) R2. Any DC gain is gone. AC/audio gain drops a small amount; not sure why it has gain at all (surely just two all-pass stages can't lose 3dB?).

[R.G.]

I did not match the trannies as read on the forum that random MPF102's worked well & didn't need to be matched.

Anything's possible. However, the specified range on Vgs for the MPF102 is from 0.5 to -8V. What are the chances that any two will match?

Well, that depends on whether they came from the same batch, or even from the same wafer, which is possible. It's also possible that halfway through making up 10,000 of them to ship to a distributor, they used several wafers, and any two could be wildly different from each other.  So it's possible that the poster was telling you the exact truth - at least within their experience. Remember the blind men and the elephant?

> Is R2 **really** supposed to be 22K? ....that this forces a DC gain ... voltage at pin 3 is then amplified and the opamp tries to make it 8Vdc on the output.
Way I see it: R2 network returns to the same 5.1V reference as the "+" input. Differential input is zero. Output should be steady 5.1V.
Agree that the reported numbers say otherwise.

That's all I could come up with, excepting a hidden soldering/component flaw. Sure, it should make no difference. But I bet sticking a decent cap in series with that 22K makes the voltage right itself to the bias voltage. Could even be an artifact of the bias currents through the dramatically different resistances, although I would have expected subbing in a TL072 to fix that.

Might just remove (lift one leg of) R2. Any DC gain is gone. AC/audio gain drops a small amount; not sure why it has gain at all (surely just two all-pass stages can't lose 3dB?).

That's why I wanted him to put in a cap - he gets the gain and the no-offset at the same time.