[OLD PEDAL REPAIR] Guyatone PS-004 Flanger not working

[rileyismycopilot]

Hello team,

I got this early 80' Japanese Guyatone PS-004 Flanger that is in untouched but not working condition.

Symptoms: sounds ok in bypass, LED lights up BUT bypassed sound only, no modulation, no noise no nothing, just the clean guitar signal coming out.

Chips inside are RC4558P (x2, 1 for LFO, 1 for ), MN3001 clock and MN3006 BBD.

How can I check them? (I know nothing about ICs really).

Thanks!

[PRR]

Welcome.

https://www.diystompboxes.com/smfforum/index.php?topic=29816.0

[ElectricDruid]

Chips inside are RC4558P (x2, 1 for LFO, 1 for ), MN3001 clock and MN3006 BBD.

Sounds like all the important bits are there at least!

[rileyismycopilot]

Hello!

Thanks for the replies folks. I'm out of town and will get back to you next week with a proper schemo and detailed measurements.
Good to know you guys are around.

PS: assuming you are the PRR from the Hoffman Amp Forum, I'm very glad to see you here. I've been posting / reading the HAF for years now and have always appreciated your incredible wealth of knowledge and ability to share it.

[Mark Hammer]

First commercial pedal I've heard of that used a 128-stage BBD. 
Won't have any trouble hitting those sub-1msec delay times.  That's for sure.

[ElectricDruid]

MN3006 seems like a pretty rare beast all around.

The short end of the delays is fine, but the long end is going to involve an audible clock or some heavy filtering. You could probably go stupid-high without even needed buffering, since the very short delay line hardly has any clock input capacitance (100pF according to the datasheet I've got). Even just 500KHz would be 0.128 msec!

I shall be interested to hear a report on how this sounds once it's working.

[Mark Hammer]

That's why the "sweet spot" for flanging tends to be either 512 or 1024-stage devices.  The dramatic stuff requires a very wide sweep.  128-stage and 256-stage devices don't have a problem nailing the short-delay end of the sweep, but the long delay at the other end of the sweep would require the sort of lowpass filtering - dictated by the lower clock rate required - that ultimately detracts from the magnificence of what the shortest delay times are able to do.  Low clock rates and wide bandwidth do not live comfortably under the same roof in BBD-based devices.

[Steben]

512-1024 Or double channels ....

[ElectricDruid]

That's why the "sweet spot" for flanging tends to be either 512 or 1024-stage devices.  The dramatic stuff requires a very wide sweep.  128-stage and 256-stage devices don't have a problem nailing the short-delay end of the sweep, but the long delay at the other end of the sweep would require the sort of lowpass filtering - dictated by the lower clock rate required - that ultimately detracts from the magnificence of what the shortest delay times are able to do.  Low clock rates and wide bandwidth do not live comfortably under the same roof in BBD-based devices.

+1 agree. The really wide sweep is hard to do. If you go from 50KHz to 500KHz, you've only got a 10:1 sweep. Not bad, but not really good enough. Pushing the top end up to 1MHz (which is when things start to get trickier) only gains you another octave, so you're to 20:1. Getting wider than that generally means dropping the long end lower. 25KHz to 1MHz is a really good 40:1, but you won't find that kind of range on many units (a few, but not many). Only the Paia Hyperflange goes wider, as far as I know.

[rileyismycopilot]

Hello gang,

I'm back with a schematic and measurements of the ICs.
(measurements taken with 9.0V battery, Rate control at lowest and Color / Width control at max).

Please let me know if something looks fishy.

Thanks a lot for your help.


RC4558P numero uno
1 = can't take a measurement, it starts and about 4V then slowly goes down
2 = same thing
3 = 4.9
4 = 0
5 = 4.9
6 = 8.8
7 = 8.7
8 = 9.0

RC4558P numero dos (LFO)
1 = 1.4
2 = 4.7
3 = oscillates from 2 to 7V
4 = 0
5 = 4.7
6 = 4.9
7 = oscillates from 2 to 7V
8 = 8.6

MN3101
1 = 8.9
2 = 8.6
3 = 0
4 = 8.6
5 = 8.5
6 = 8.4
7 = 8.4
8 = 8.9

MN3006
1 = 8.9
2 = 8.6
3 = 4.1
4 = 8.9
5 = 9.0
6 = 8.6
7 = 8.9
8 = 8.9

[Fender3D]

Replace ic1

[ElectricDruid]

RC4558P numero uno
1 = can't take a measurement, it starts and about 4V then slowly goes down
2 = same thing
3 = 4.9
4 = 0
5 = 4.9
6 = 8.8
7 = 8.7
8 = 9.0

Could that happen if the input buffer transistor was dead. Then the cap C3 would be connected to +V via R3 and down to virtual ground by R5. It'd start to charge up, and force the input voltage lower as it did?

There's definitely *something* suspicious going on around that first op-amp though. I'm just not completely convinced that it's the op-amp itself that's at fault. If it's socketed it's an easy test to make and I'd try it, but if it isn't, I'd want to be more sure before mucking about trying to get it out.

[Mark Hammer]

...and in the interest of low ticking, IC3 could benefit from being replaced by a low-current type, such as LM358, TL062, or TL022.

[Fender3D]

Could that happen if the input buffer transistor was dead. Then the cap C3 would be connected to +V via R3 and down to virtual ground by R5. It'd start to charge up, and force the input voltage lower as it did?

There's definitely *something* suspicious going on around that first op-amp though. I'm just not completely convinced that it's the op-amp itself that's at fault. If it's socketed it's an easy test to make and I'd try it, but if it isn't, I'd want to be more sure before mucking about trying to get it out.

Symptoms: sounds ok in bypass, LED lights up BUT bypassed sound only, no modulation, no noise no nothing, just the clean guitar signal coming out.

Pedal sounds ok in bypass... Q1's alive and kickin'

[ElectricDruid]

Pedal sounds ok in bypass... Q1's alive and kickin'

Yep, you're quite right. Sorry, I missed that. Ok, so not the buffer transistor.

So is it the op-amp, or just "the stuff around the op-amp"?

T.

[rileyismycopilot]

Thanks a lot for the detective work!

"RC4558P numero uno
1 = can't take a measurement, it starts and about 4V then slowly goes down
2 = same thing
3 = 4.9
4 = 0
5 = 4.9
6 = 8.8
7 = 8.7
8 = 9.0"

Are the 2 sides of the opamp looking weird? Or only one? If so, which one?

[Fender3D]

I don't know the pedal layout...
it looks like you're missing a jumper or contact somewhere since there's no GND on BBD and it's output voltage might affect IC1b voltages...

Anyway:
half IC1 does not work ok, you can safely replace it

[rileyismycopilot]

Hi gang!

I've replaced the IC1 with a socketed TL062 (as per Mark Hammer's input) and the problem is still exactly the same: all I have is clean guitar signal, whether in bypass or fx mode, zero flanger sound.

I've re-measured the IC's voltage and here's what I have now:

IC1 (TL062 - replaces dead RC4558)
1. 4.8
2. 4.8
3. 4.8
4. 0
5. 4.8
6. 4.8
7. 0.7
8. 8.9

IC2 (MN3101)
1. 8.9
2. 8.5
3. 0
4. 8.5
5. 8.4
6. 8.3
7. 8.2
8. 8.8

IC3 (MN3006)
1. 8.9
2. 8.5
3. 4.0
4. 8.8
5. 0
6. 8.5
7. 7.3
8. 7.3

(didn't measure the LFO IC as this one seems to be allright).

Anything obviously weird?

[Mark Hammer]

Have we verified that the switching FET is okay?  The simple test is to bridge the drain and source pins of the FET to mimic it being "on".  Keep in mind that is the FET is dead, all you will hear in effect or bypass mode is the clean signal.  Pedals of this type do not bypass; they merely cancel/engage the effect by allowing/blocking the wet signal to mix with the dry.

[rileyismycopilot]

Have we verified that the switching FET is okay?  The simple test is to bridge the drain and source pins of the FET to mimic it being "on".  Keep in mind that is the FET is dead, all you will hear in effect or bypass mode is the clean signal.  Pedals of this type do not bypass; they merely cancel/engage the effect by allowing/blocking the wet signal to mix with the dry.

Hi Mark,
Thanks for the tip.
Are you talking about Q2 (2SK30GR), at the output of the circuit?
If it happens to be dead, what could I replace it with?