MXR Phase 45 Fault - original script logo from 1975

[sohosteve]

Hi there
My dear old MXR Phase 45 appears to be dying.
When I first plug it in and switch it on the phasing is extremely subtle to the point of supposedly not sweeping at all unless I max the speed control and then I can hear a slight high speed shimmer.
After I've been playing through it for about half an hour it gradually gets deeper until I can hear it working at slower speeds - though it never seems to go as deep as it used to.
So I get the impression that one of the components is on the way out. I'm guessing a capacitor as I understand that these 'charge' up.
I say understand as I'm not an electronics guy - I can't read a circuit diagram or a resistor code though I have a rough understanding (read non existent) of what they do.
However I know how to remove and swap out components (I've a decent iron and solder sucker etc) if I'm lead by a photograph and have the right components to hand.
So I wonder, is there anyone out there who could diagnose this particular fault for me and lead me forward in repairing my trusty old phaser?
Any suggestions gratefully accepted :-)

[52 Bill]

So I wonder, is there anyone out there who could diagnose this particular fault for me and lead me forward in repairing my trusty old phaser?
Any suggestions gratefully accepted :-)

I'm sure that you will get this sorted out, but you'll have to help us by giving us more information to work with. Do you have a multimeter?

[sohosteve]

I'm sure that you will get this sorted out, but you'll have to help us by giving us more information to work with. Do you have a multimeter?

Hi Bill,
Yes I do, though I'm not too sharp with it - I've just about managed to work out how to measure the resistance of a guitar pickup though! :-)

[Mark Hammer]

Like the Phase 90, you will find a trimpot on the circuit board. This trimpot is used to "bias" the FETs in the circuit. They need to have a basic DC voltage applied to their gates (essentially, input pins) to which the LFO voltage is added.

It can happen (especially over the course of 35-40 years!) that the trimpot gets budged. You can reset the bias to where it should be, by ear. It may not be THE mosta ccurate means, but should be enough to get the job done. Feel free to move the trimpot around, since it will not harm the FETs. But I would advise you to mark off, in some manner, where it is presently set to, so you can return to it, should the problem you are experiencing not be a result of trimpot drift.

It can also happen that capacitors age, and dry out, however this is less likely to happen to the droplet-type tantalum capacitors, in comparison to the can-like electrolytic types. As near as I can tell, your script unit will likely use tantalums throughout. That said, I suppose something might have happened to it over time. So, once you are convinced that you have the trimpot set right (i.e., you are producing audible sweep that doesn't appear to seize up at any point near the top or bottom of the sweep), keep an ear out for the sorts of changes you described in your initial post.

If I look for images of the board, unfortunately, I can only see the top/component side, and no images of the copper/bottom. So the best I can tell is that one of the two 10uf cap "cornering" the trimpot might be worth a look and possible replacement, although which particular one, I can't tell.

[sohosteve]

Hi Mark,
I did think that might be the case too (trimpot fix) and it was the first thing I checked.
However the trimpot is 'glued' in place and has absolutely no movement whatsoever.
Also, the fact that after being plugged in for half an hour it slowly starts to work makes me think that it is more likely that it is a capacitor issue.
This theory is more based on an instinct developed from thirty years of chatting to techs about various component failures than any real working electronic knowledge.
I'll take some clear 'gut shots' and post them tomorrow in the hopes that this may assist in your(or other member's) assessment.
Huge thanks to you and our fellow compadres here for your suggestions and advice.
Peace
Steve

Like the Phase 90, you will find a trimpot on the circuit board. This trimpot is used to "bias" the FETs in the circuit. They need to have a basic DC voltage applied to their gates (essentially, input pins) to which the LFO voltage is added.

It can happen (especially over the course of 35-40 years!) that the trimpot gets budged. You can reset the bias to where it should be, by ear. It may not be THE mosta ccurate means, but should be enough to get the job done. Feel free to move the trimpot around, since it will not harm the FETs. But I would advise you to mark off, in some manner, where it is presently set to, so you can return to it, should the problem you are experiencing not be a result of trimpot drift.

It can also happen that capacitors age, and dry out, however this is less likely to happen to the droplet-type tantalum capacitors, in comparison to the can-like electrolytic types. As near as I can tell, your script unit will likely use tantalums throughout. That said, I suppose something might have happened to it over time. So, once you are convinced that you have the trimpot set right (i.e., you are producing audible sweep that doesn't appear to seize up at any point near the top or bottom of the sweep), keep an ear out for the sorts of changes you described in your initial post.

If I look for images of the board, unfortunately, I can only see the top/component side, and no images of the copper/bottom. So the best I can tell is that one of the two 10uf cap "cornering" the trimpot might be worth a look and possible replacement, although which particular one, I can't tell.

[52 Bill]

However the trimpot is 'glued' in place and has absolutely no movement whatsoever.

I've seen some of the early MXR trimpots that have had the wiper element soldered after they had been set.

I'll take some clear 'gut shots' and post them tomorrow in the hopes that this may assist in your(or other member's) assessment.

This will be great to help us know exactly what parts you have in your unit.

In the mean time, if you set your meter to read DC voltage, connect the black lead to the case (ground) and then read the voltages at the three terminals of the trimpot. One should be zero, one should be about half of the supply voltage and the wiper should have a fluctuating voltage that rises and falls at a rate set by the speed control.

If the effect is weak until it warms up, I would expect the wiper (control) voltage to start out at one value and then as it warms up change to a different voltage. See what you find.

[Mark Hammer]

If It's behaving that way, then the simple thing to do is replace the tantalums. Assuming you can replace them in a nondestructive way, it's a cheap and simple fix.

[sohosteve]

In reply to 52 Bill and Mark Hammer.
Thanks again for your kind help guys.

Here are the gut shots - hope you can read the values clearly.
Also, despite the fact that I know nowt about electronics I can cleanly swap out components without causing track damage (or burning the components out) with my solder sucker and iron.
If someone can visually direct me to which parts need changing and what I need to request when ordering (and a good UK supplier if possible) then I can have the parts swapped out and the whole thing back together in less than 20 minutes :-)

By the way I've just noticed there's some brown 'splurge' on the back of the pcb (flux run off? rust?) do you think this may be causing some kind of short?
Wouldn't explain why it 'warms up' after being played through for a while though. I suspect it's capacitors being charged up over time, but what do I know?! ;-)

If It's behaving that way, then the simple thing to do is replace the tantalums. Assuming you can replace them in a nondestructive way, it's a cheap and simple fix.

[DrGonz78]

Well first off take a q-tip dipped in (99-100%) alcohol to clean off that nasty flux residue. Some of the solder appear to have cracked joints. Do a light and quick re-flow of just the joints that look cracked to see if makes any difference. Perhaps wait and test voltages/troubleshoot first then solder those four main offenders to see if there is a difference. Just a thought, good luck.

[sohosteve]

Hi Bill

Well, I uploaded the 'gut shots' - hope they're clear enough to read the component values.

And I took those measurements as instructed but the results were not quite as you suggested.
1. gives the zero voltage
2. gives 4.6 (half the battery voltage I guess)
3. gives 1.9 and doesn't change wherever I set the rate control
There is no movement in any of these voltages from when I first attach the battery or after half an hour of the battery attached. And I was totally meticulous about attaching the earth and switching the meter on before connecting the battery in the hopes that I'd notice the most extreme change over the course of it 'warming up'.

Does any of that help at all?

I've seen some of the early MXR trimpots that have had the wiper element soldered after they had been set.


This will be great to help us know exactly what parts you have in your unit.

In the mean time, if you set your meter to read DC voltage, connect the black lead to the case (ground) and then read the voltages at the three terminals of the trimpot. One should be zero, one should be about half of the supply voltage and the wiper should have a fluctuating voltage that rises and falls at a rate set by the speed control.

If the effect is weak until it warms up, I would expect the wiper (control) voltage to start out at one value and then as it warms up change to a different voltage. See what you find.

[52 Bill]

And I took those measurements as instructed but the results were not quite as you suggested.
1. gives the zero voltage
2. gives 4.6 (half the battery voltage I guess)
3. gives 1.9 and doesn't change wherever I set the rate control

Good photos, these will help. I see that the wiper on your trimpot has been soldered down. The top terminal of the trimmer that you refer to as #3 is the wiper. The wiper is directly connected to the 470K (yellow/violet/yellow) resistor that is directly to the left of the trimmer.

Using the same setup as before read the voltage at the bottom side of the 470K resistor. Hopefully you will get the fluctuating voltage here. Also take readings on both ends of the 3M9 (orange/white/green) resistor that is horizontal at the very top of the board, where the wires connect to the board.

The 10uF tantalum capacitors that Mark has suggested replacing are the three blue epoxy blob caps that have the black and green paint markings. The one that is directly to the right of the trimmer is the timing cap for the LFO. If it is a bad cap, I would think that this one would be the one most likely to cause your problem.

[sohosteve]

Hi Bill,
I'll check that later.
Just to clarify (in case I wasn't too clear initially) the problem has never been about the speed of the sweep - it's the depth that is the problem.
When first plugged into and switched on it seems not to phase at all except when the rate is turned up full - then a very shallow fast, light shimmer is heard. As it warms up (5-20 mins later) it gradually becomes deeper until it is very noticeable at fast rates and very gently there at slow rates.
If the pedal had a Depth control as well as a Speed it starts with the pedal sounding like Depth is set to zero when first used then gradually gets to the point where it sounds like it has increased to about 4 (out of a possible 10).
Compared to a vintage block logo 45 a pal has (or any other Phase 45 pedals I have checked out on Youtube) my pedal sounds like it doesn't sweep as DEEPLY as it should - Speed adjusts just fine.
I only mention this again as you just referred to the 'timing cap for the LFO' and I wondered if you mean timing as in time - ie speed/rate - apologies if I misunderstood!

[52 Bill]

I only mention this again as you just referred to the 'timing cap for the LFO' and I wondered if you mean timing as in time - ie speed/rate - apologies if I misunderstood!

The cap does affect the speed of the LFO, but if it fails, it is possible that it can change how strong the waveform is, which will change the depth of the phasing.

The real test will be the voltage readings that you get from the oscillator output. And I would still want to check the setting of the bias control.

[sohosteve]

The first requested measurement fluctuates from 1.34 to 1.44

Measurement across horizontal resistor is fluctuating quickly between 7.6 to 8.7 at fastest speed and slowly at the slowest between 0.00 - 0.11

The cap does affect the speed of the LFO, but if it fails, it is possible that it can change how strong the waveform is, which will change the depth of the phasing.

The real test will be the voltage readings that you get from the oscillator output. And I would still want to check the setting of the bias control.

[soundguruman]

Yeah, the tantalum caps die with age. That's pretty common.

FOAM- used to wrap the circuit board... it ROTS.
After it starts to rot, it eats into the circuit tracks and components.
LOOK for tiny spots of green corrosion on the circuit tracks and components.

The foam will actually turn into a sticky corrosive liquid goo. (with age)
All of this residue must be carefully removed.
We suspect that this foam is formaldehyde based. It can eventually destroy the unit.