Treble loss with a Boss Ph-1 Phaser

[Rand]

Hi guys, this is my first post here and I am sorry if this has been asked before. I did try to do an extensive search but couldn't seem to find any clues:)

I have inherited an original Boss Ph-1 which didn't pass any signal in either bypass or on mode. I reviewed the schematic and decided it was almost certainly an issue with the first IC chip. I replaced this and the pedal works, however I notice there is a huge tone suck even in bypass mode. Its a big loss of treble and some signal even when bypassed or with the effect on.

Now I know that these pedal are not true bypass and a bit of capacitance can be an issue with these buffered pedals, but this is a really noticeable degrading of input signal even if its the only pedal between the guitar and amp. I dont recall other boss pedals having such a treble "suck".

I've attached the schematic. Can someone help me with where I would start looking at why the problem exists? I'm guessing since its the same with bypass or the effect on, its somewhere right at the input or output?

Thank you

[iainpunk]

yes, this pedal is just a notorious tone sucker, the engineers have taken quite a few ''''clever'''' shortcuts to simplify the circuit. its also kind of hard to mod it for more treble. a slight improvement is to gain with a 4.7nF (or smaller to taste) cap parallel to the 22k resistor called R20, but its a small subtle improvement, not a full solution.
what happens if you have another buffered pedal in front of this one?

cheers, Iain

[antonis]

No Welcome, Iain..??

[iainpunk]

No Welcome, Iain..??

oops

Rand, welcome to the forum
(are you the reincarnation of Ayn Rand?)

cheers, Iain

[Mark Hammer]

Hi and welcome to the dark side.

The fact that the issue is common to both the effect and bypass mode suggests that the problem lies either in the output mixer stage, OR the switching.

Q5 and Q6 are both turned on or off simultaneously by the flip-flop circuit.  When they are turned off (bypass mode) Q5 blocks phase-shift signal from being mixed in with clean.  The gain of that op-amp output stage is set by C6 (1uf) and R19 (1M), such that the output op-amp is juuuuuuusssst a hair above unity gain. When Q5/Q6 are turned ON, the phase-shifted signal is now allowed to pass to the mixer/output, but equally important, R18 is essentially placed in parallel with R19.  This reduces the effective resistance of the ground leg of that op-amp, and the gain of the stage is increased to just under 3x.  This compensates for the drop in output level when you take out some of the signal with notches.  And since the PH-1 has no feedback/emphasis, the additional signal normally produced by use of feedback isn't there to increase overall level in effect mode.

Normally, in a pedal of this age, one looks for electrolytic capacitors in critical positions, since they can dry out over the course of 40 years.  Here, all signs point to C6 as being a potential source of signal drop.  If it is not behaving like it did in its younger days, and "blocking" the path to ground, that output stage will not get the slight boost in bypass mode and especially the additional boost in effect mode.

My gut sense is that you should change out C6, paying attention to the orientation/polarity of the cap.

[Rob Strand]

I have inherited an original Boss Ph-1 which didn't pass any signal in either bypass or on mode. I reviewed the schematic and decided it was almost certainly an issue with the first IC chip. I replaced this and the pedal works, however I notice there is a huge tone suck even in bypass mode. Its a big loss of treble and some signal even when bypassed or with the effect on.

I've got my doubts it's the JFET switching since the Dry signal does not involve any switching.
The original fault may not be fully fixed.

One thing to check for is bad contacts on the 6.5mm sockets.  That  can cause what you are seeing.  I have seen a couple of cases where the socket had a marginal connection on some very old Boss units.

If you have a multimeter you can check the voltage on the output of each of the opamps.   That will tell you a lot. The DC voltage on the zener D3 is also a useful thing to measure.


While low on the list, the output cap C9 could have gone dry.

[Rand]

Thanks guys for your helpful responses.

I tried replacing C6 but not change there. I tried cleaning the contacts and sockets but no change either.

I measured the RC3403's and get the following:

IC1   Pins 1-3 = 5.1v
        Pin 4 = 8.83v
        Pins 5-7 = 5.1v
        Pins 8-10 = Jump around - ie the meter won't stay still and gives somewhere between 3.9 and 5.7v
        Pin 11 = 0v
        Pins 12-13 = ie the meter won't stay still and gives somewhere between 3.9 and 5.7v
        Pin 14 = also jumps around but ends up with a top measure of 10.8v

IC2 - Pins 1-3 = 5.1v
        Pin 4 = 8.83v
        Pins 5-7 = 5.1v
        Pins 8-10 = 5.1v
        Pin 11 = 0v
        Pins 12-13 = 5.1v
        Pin 14 = 5.1v

I also note that D3 measures 4.5v

Thanks for your time:)

[Mark Hammer]

I have inherited an original Boss Ph-1 which didn't pass any signal in either bypass or on mode. I reviewed the schematic and decided it was almost certainly an issue with the first IC chip. I replaced this and the pedal works, however I notice there is a huge tone suck even in bypass mode. Its a big loss of treble and some signal even when bypassed or with the effect on.

I've got my doubts it's the JFET switching since the Dry signal does not involve any switching.
The original fault may not be fully fixed.

One thing to check for is bad contacts on the 6.5mm sockets.  That  can cause what you are seeing.  I have seen a couple of cases where the socket had a marginal connection on some very old Boss units.

If you have a multimeter you can check the voltage on the output of each of the opamps.   That will tell you a lot. The DC voltage on the zener D3 is also a useful thing to measure.


While low on the list, the output cap C9 could have gone dry.

Yes and no, re: dry signal.  You are most certainly correct that the dry signal is always connected to the output/mixer stage, in both effect and bypass modes, but switching is involved in adjustment to the effect-bypass volume balance via Q6.

It is difficult to see how IC1-pin 14 gets up to 10.8V when the supply voltage at pin 4 on both chips is 8.83V.  Something is amiss in measurement, there.

[Rand]

Ok thanks Mark

When you say "It is difficult to see how IC1-pin 14 gets up to 10.8V when the supply voltage at pin 4 on both chips is 8.83V.  Something is amiss in measurement, there." - could you offer me any clues as to what to do to troubleshoot this difficulty?

[Mark Hammer]

Well, you at least might want to measure things again, especially that pin.

[Rob Strand]

Yes and no, re: dry signal.  You are most certainly correct that the dry signal is always connected to the output/mixer stage, in both effect and bypass modes, but switching is involved in adjustment to the effect-bypass volume balance via Q6.

I was just saying I had my doubts - so possible but unlikely until new info comes about.   There was a rationale to thinking JFETs weren't involved: *if* the DC voltage at the opamps were OK then the only way to lose signal would be for Q5 to short to ground or Vref and there's not much around there which would do that.

It is difficult to see how IC1-pin 14 gets up to 10.8V when the supply voltage at pin 4 on both chips is 8.83V.  Something is amiss in measurement, there.

Yes I don't undestand that one the supplies of both opamps are at 8.8V so the 10.8V could be the multimeter misreading the variable signal.


[Remove incorrect part of post.]


EDIT:

Actually Mark I think your line of thought as being a involved is looking more likely.

What I don't get is the volume drop for the Dry signal.   I mean there's nothing to kill the signal.

*If* Q5 is not turning off it could affect the Dry signal by you would expect to hear some phase effect when in non-effect mode.

[Rand]

Ok thanks guys so much for your info.

I'll try to chance R21 and see if this gives me the subtle improvement.

Just for my own knowledge, why is it that pins 8-14 (not including 11) jump around with the meter ? ie I cant get a stable reading with them. Is this normal?

[Rob Strand]

Ok thanks guys so much for your info.

I'll try to chance R21 and see if this gives me the subtle improvement.

Just for my own knowledge, why is it that pins 8-14 (not including 11) jump around with the meter ? ie I cant get a stable reading with them. Is this normal?

Sorry, I screwed up before.   Put R21 back to the old value.

There's got to be more to it.

I guess the key question is can you hear any Phasing when bypassed?  the best inducation would be that the speed control has an effect.

[Mark Hammer]

I'm not totally persuaded that R20's value, relative to R21, is the issue.  It is a non-inverting stage, such that gain for any given input resistor is set by R21/R19 in bypass mode and R21/R18 in effect mode.  If anything, maybe change R20 to 10k, though I'm not hopeful about that.

[Rob Strand]

I'm not totally persuaded that R20's value, relative to R21, is the issue.

I'm convinced it's not R21.     My bad, wrong idea altogether.   The fact it restores some of the dry gain is suspicous in itself as that means Q6 might not be turning off.  If everything is working R21 should not affect the Dry signal gain.  And if Q6 isn't turning off then maybe Q5 isn't turning off either.

I guess another test would to measure the voltage at the collector of Q8 to make sure the flip-flop output is swinging correctly.  IIRC it should be switch between 0V and 6V.

The key problem to be solved is the drop in bypass mode mode.   I just shouldn't happen.

An ideal test would be to listen to the output of IC2 at pin 6/pin 7 to see if the level drop is present there - you would need to add a coupling cap to do the test.   If the problem exists there it narrows down the problem to a very small region of circuit.

The next step would be to temporarily lift Q5 but you need to be confident of your soldering skills.  Lifting one end of R17 might give some info but it's not as good as lifting Q5.

[PRR]

> why is it that pins 8-14 (not including 11) jump around with the meter ?

D3 and C10 not doing their job. Voltage on C10 should be dead steady and your observation suggests it isn't.

[Rand]

Thanks Rob and PRR

Rob  - I had a listen with a little audio probe I made and the output at pins 6-7 of IC2 is barely different to the direct signal. Maybe slightly but barely noticeably different.

Ive measured Q8 and the three pins measure 8.8v, 2,8v and 8.5v when the pedal is pressed on (for the life of me I cant find the pinout of this transistor to work out which pin is B,C and E) and measures 7.9v, 8.5v and 8.5v when the pedal is pressed off.

If I were to lift Q5, what would I test to determine things?

PRR -I've measured the voltage across c9 and that measures a nice stable 5.1v with no jumping around. D3 seems to measure a stage -9v on one side and -3.9v on the other side, also stable.

[Rob Strand]

Thanks Rob and PRR

Thanks to Mark as well for kicking me back on the right path.

Rob  - I had a listen with a little audio probe I made and the output at pins 6-7 of IC2 is barely different to the direct signal. Maybe slightly but barely noticeably different.

OK great so that means the input side is fine.    It's the output side which has the problem.   Either the flip-flip, the JFET switches, or the output opamp (not likely)

Ive measured Q8 and the three pins measure 8.8v, 2,8v and 8.5v when the pedal is pressed on (for the life of me I cant find the pinout of this transistor to work out which pin is B,C and E) and measures 7.9v, 8.5v and 8.5v when the pedal is pressed off.

Those look very weird and don't make sense at all.   The transistor base should not be going above 0V to 1.3V or so.   So if 2.8V was the base then it's like the voltage is shifted up.  But it's not just that the pedal off measurements are way too high.

When you connect the ground of the multimeter make sure you connect it to the ground on the input socket, or output socket, or the PCB.   Don't connect it to the DC jack as components D5 and R45 (390R) will affect the measurements; the components are at top right corner with the schematic orientation as in reply #1.   Maybe that's why you got voltages > 9V before?

The transistors follow the Japanese pinouts,



If you aren't sure how you measured the voltages maybe it's worth rechecking the measurements.

It's definitely worth getting all these measurements right because that process may actually expose the fault.
One possibility is there is a soldering problem on D4 (check this first), Q7, Q8 or around that flip-flip circuit.

If I were to lift Q5, what would I test to determine things?

Maybe hold off for now.    Given the flip-flop voltage issue there's no need to do this - not yet anyway.
The idea was to show the level rose-up.    To get a the full-picture that the output opamp is OK you might also have to remove Q6 or R18.     Your audio probe test shows the signal is at least coming in at full-level.

[Rand]

Thanks Rob

I did have my meter's negative probe on the ac jack instead of the output socket! Naive error so thanks for educating me with that.

Now re-measuring with the neg probe on the output socket,  I get measurements of Q8 as follows: Base 1.11v, Collector 0.54v, and Emitter 0.52v and after pressing the foot switch I get 0.19v, 5.98v, and 0.52v.

[Rob Strand]

Now re-measuring with the neg probe on the output socket,  I get measurements of Q8 as follows: Base 1.11v, Collector 0.54v, and Emitter 0.52v and after pressing the foot switch I get 0.19v, 5.98v, and 0.52v.

Those voltages look fine.

I guess we have reached the nasty point in the debugging process.    It's looking like either one or both of the JFETs have a problem.

If you are confident of your soldering skills it would be very helpful to remove both JFETs Q5 and Q6.    The aim then is to see if the signal loss problem is *completely removed* - in both Effect and Bypass modes, although both will sound like Bypass with the JFETs pulled.

*** Before removing the JFETs try to identify which JFET went into which position.  According to the schematic  Q5 should be K30A-Y  and Q6 K30A-O but that might not be the case in the actual unit.   They could be identical and that means you need to identify them by eye - eg put a pencil mark on one.

If removing the JFETs fixes the issue then the next step is testing the JFETs.

When viewed from the bottom, the pinout for the 2SK30ATM's is,


From,
http://www.datasheetdir.com/2SK30ATM+N-Channel-JFET

Some basic JFET tests are use the diode tester between G and D and G and S.  It should read about 0.65V in one direction and open in the other.   Check both directions.    You can then use the resistance test to measure the resistance between D and S with G open and it should be around 400 ohms.  Check both directions.

Maybe those measurements will be enough to detect the bad JFET.

To do more, you would need to measure VP and IDSS as per section 11 of this page.  I recommend using 10M instead of 1M in the test circuit.  Also a safer set-up would be to add another 100R or 220R in series with the switch on the test schematic.

http://runoffgroove.com/fetzervalve.html

Now, if you find a JFET is fried you will need to source new JFETs.    (If one is working it might be possible to get the unit working with one but it's better to keep the original circuit.)