no sound from mxr phase 90/diode missing

[frogman]

I ended up turning the capacitor around and the whole pedal BLEW UP



just kidding. It works now and most of the voltages are in proper range according to the GGG schematic. Except for these:

-Pin 6 of opamp's 2-5 is at 5V when it is supposed to be around 2V.
-The gate of Q2-Q4 are at 3.5V when they should be at 2.5V
-All of Q5 is down about 1V.

Plugged in, it sounds how a phaser should sound.

One additional question. I've noticed the trimpot on it still changes nothing tone, voltage or impedance wise when adjusted. Since its working,my first impulse is to box it and leave it be, but could there be an additional problem still here?

[mcknib]

The trimpot should adjust the gate voltages on your FET's if memory serves -the phase range is very small on it but if you turn it too far you should lose phasing.

Begs the question with your voltages what make of meter are you using does it have at least a 10M input impedance? mind you must be ok if the rest are ok'ish

I'd be happy if I got it phasing so I'd box it up and keep it for a while to make sure nothing else goes wrong before you give it back.

duck_arse was right on the money

[frogman]

So I just got home and was doing some testing on the trimpot affecting the gate and the voltages suddenly dropped. Now I am getting

Q1-Q4:

G - .3V
D - 0V
S - 0V

Opamp's are drawing
Pin 1 - 0V
2 - 1.9V
3 - 0V
4 - 0V
5 - 0V
6 - 1.9V
7 - 9V
8 - V

Did a continuity check on the power supply/biasing section and both sides of the D2, trimpot and C9 are now connected to ground. Would it make sense that one of the 3 components (d2 , c9, trimpot) I mentioned shorted out and made both sides of all of the 3 components grounded. The "VB" section which meets the now grounded area meets the source of the FET'S would explain their voltage being at 0V.

I pulled out the c9 and both sides are continuous.

There is still one (out of three) non- grounded pin of the trimpot and one (out of three) non grounded pin of the D2 and D1.

[duck_arse]

ahhh, you had me for a minute there, but then .....

from your this diagram:


D1 sets "A" voltage dependant on its part number, whatever it might be. that voltage is then named VB, and should appear on C9 and TR1 (where their pins connect) and both ends of R3 (figure the meter loading, obvs.), and both ends of the R5, 8, 11, 14 and one end of C10 and Q5.

TR1 reads the VB volts, and a proportion of that appears on its wiper. THAT voltage (your choice, it's adjustable) appears on both sides of R30 (loading, again), and each of the fet gates, and one end of C10.

if you measure resistance from ground to the top of the trimpot, it should be 200k or something, can't see it at the mo. it might read different due to the diodes and the meter lead directions. if you read from ground to the wiper, it should vary up to your end-to-end reading.

if either the zener or that blasted C9 is bad (or the missing protection diode), it might be short/low ohms across the VB and ground, and will mess you up. take out C9. what happens to those voltages?

can we see a new photo of that corner of the board when you have removed the cap? (boy, I hope that caps bad.)

[mcknib]

You don't say if you changed the cap - Did you put another replacement electro in there?

It does look as duck_arse says that your problem is somewhere on the VB line going by your initial voltages and your non-inverted inputs on the op amp pin 3's, FET gates etc. Hopefully it is that mo'fo' cap so again as DA says check the reading with it out.


With the diagram duck-arse is referencing you can see all your VB points which should obviously all be connected I'd be inclined to closely inspect the solder side of the board as well just in case the previous owner carried out some other work on it and bridged a trace or something.

[frogman]

ahhh, you had me for a minute there, but then .....

from your this diagram:


D1 sets "A" voltage dependant on its part number, whatever it might be. that voltage is then named VB, and should appear on C9 and TR1 (where their pins connect) and both ends of R3 (figure the meter loading, obvs.), and both ends of the R5, 8, 11, 14 and one end of C10 and Q5.

TR1 reads the VB volts, and a proportion of that appears on its wiper. THAT voltage (your choice, it's adjustable) appears on both sides of R30 (loading, again), and each of the fet gates, and one end of C10.

if you measure resistance from ground to the top of the trimpot, it should be 200k or something, can't see it at the mo. it might read different due to the diodes and the meter lead directions. if you read from ground to the wiper, it should vary up to your end-to-end reading.

if either the zener or that blasted C9 is bad (or the missing protection diode), it might be short/low ohms across the VB and ground, and will mess you up. take out C9. what happens to those voltages?

can we see a new photo of that corner of the board when you have removed the cap? (boy, I hope that caps bad.)

The wiper of the trimpot reads between 62 and 63k and no amount of adjustment seems to change it. The wiper is at about 20mV.

With the cap out. The op amp and fet voltages are the same as they were with the capacitor soldered in, after it malfunctioned..


Q1-Q4:

G - .3V
D - 0V
S - 0V

Opamp's are drawing
Pin 1 - 0V
2 - 1.9V
3 - 0V
4 - 0V
5 - 0V
6 - 1.9V
7 - 9V
8 - 0V





I know my soldering is sloppy, however, it functioned for a while after I replaced the tantalum caps (with new aluminum electrolytic caps). And both polarities of the C9 were separate until the most recent malfunction. So I dont think I bridged a trace, although I could have. Also the stuff to the right in the picture is flux, not a scratch.Also if the picture is too small, dont be afraid to ask for a bigger one.

I have a slightly off topic question just for the sake of learning. You said that D1 sets voltage, wouldn't it be the R29 in the schematic setting the voltage for VB (voltage drop) since it connects straight from 9V?

[duck_arse]

search time for the frogman. RG and PRR have written of the dark art of zener diode usage hereabouts. read theirs so I don't give you the raw prawn.

basically, the zener is going to have 5V1, if so marked, across it, so you need to help lose the other 3V9 somewhere. a resistor will do this for you, quite nicely. the trimpot wiper then reads a proportion of that voltage, except in your case.

the picture is penty large, I just wanted to make sure you'd taken the cap out. your total loss of supply volts seems to be the major problem to fixxe, just at the moment. where did they all go?

[frogman]

So we've narrowed it down to the zener shorting, now I need to figure out why it died in the first place correct?



Would it be best to replace the zener, so that the correct voltages are present, then I could figure out what overloaded it? There is now only about 4mV across the zener.

I'm not sure that the schematic is 100% which is making it hard to envision exactly how I can trace the problem to the source.

[mcknib]

Could just have been a power spike perhaps the previous owner used the wrong power supply or an AC supply or whatever I'd just replace it and see how it goes from there. That would be the best way to see if you get the correct voltages etc and see if it functions correctly.

The GGG schematic is probably the nearest if not the same as yours it being the script phase 90 so I'd use that for reference.

[duck_arse]

measure the value of the resistor in series w/ the zener. if it went funny (ho, I don't know, short?), and the zener got too much current, it mighta failed. if it is the root cause, that is.

(you could pull the zener out of circuit, and if, only IF, there is now no findable supply shorting funnies still going on, put a temporary resistor, same value as the zener series R, to produce a tempy bias voltage. this would show if the rest of the circuit is working, but it relies on the rest of the circuit working.)

[mcknib]

FYI here's the values of the components in the audio path schematic aside from the op amps, labelling and use of nearest modern value electro's + supply filtering electro's the values should be the same SW1 is just to switch the 22K script feedback resistor in or out but as I said I'd use the GGG schem for ease of reference

R1 1M       R17 150k      C1 10n      IC1    TL071 Mono Op Amp
R2 10k      R18 150k      C2 47n      IC2 TL074 Quad Op Amp
R3 470k    R19 150k      C3 47n      IC3 TL061 Low Current Mono Op Amp
R4 10k      R20 56k        C4 47n
R5 22k      R21 150k      C5 47n      Q1 -Q4 2N5952 Matched
R6 10k      R22 150k      C6 47n      Q5 2N4125 General Purpose Silicon PNP
R7 10k      R23 3M9       C7 22μ
R8 22k      R24 4k7        C8 10n      TR1 250k Trim Pot
R9 10k      R25 470k      C9 10μ      SPEED 500k Reverse Log Pot
R10 10k    R26 150k      C10 47n     SW1 SPST
R11 22k    R27 22k        C11 100μ
R12 10k    R28 22k        C12 100μ   
R13 10k    R29 10k        C13 10n     
R14 22k    R30 1M
R15 10k    R31 100R     D1 5v1 Zener
R16 22k    R32 3k3       D2 LED

tempy bias voltage

....and whatever you do don't get frustrated and lose your tempy

Great word that duck_arse worthy of much more use I think