Vintage MXR phase 90 script model clipping low notes on guitar

[newperson]

Would anyone know why it would clip / distort on lower guitar notes?  Could an bad op amp cause this?
Or where should I look first. The bypass is clean throughout any of the notes. Not a reissue. Thanks,

[newperson]

I took some readings.  I could not find any information about which IC is labeled to which, ie IC1, IC2, IC3, IC4, IC5, IC6.

There are 6 ICs, 4 grouped together that appear to be the phase shifting stages and two grouped that appear to be the input buffer and LFO. 

Looks like bottom right hand corner is the LFO and top right hand corner is the input buffer.

All measurements are VDC with the ground clipped to the chassis of the pedal.

Battery 9.022
diodes 9.001/0  and 4.42/0
trimmer 4.42/1.356/0

transistors 4 that go with the phase shift all the same
4.41/4.42center/1.76-1.8

single transistor
2.81/3.84 center/4.42

LFO
1_.002
2_4.2-4.7
3_3.9-4.5
4_0
5_.001
6_0 to 6.1
7_9.01
8_.001


Buffer
1_.003
2_4.35
3_4.138
4_0
5_.0032
6_4.35
7_9.01
8_.001

phase stage one top left corner
1_.0039
2_4.42
3_4.415
4_0
5_.004
6_4.49
7_9.01
8_.001

all of these are almost 100% the same the only difference is pin six on 1 and 3 reads 4.49VDC
and pin 6 on 2 and 4 reads 4.36VDC.  Assuming they are ordered directly in a row from top to bottom.

Any thoughts on what might be causing the high about of distortion / clipping on the lower E string?  I resoldered all the PCB to make sure the connections where good and swapped out the buffer input just to see if that was the issue.  The distortion goes away in bypass mode.


From reading around some people say that the ref voltage should be 5.1 and it appears that mine is closer to 4.42.  Which is correct and would that lower voltage lower the 'headroom' and cause the clipping on lower notes?

[robthequiet]

That's a great box -- Ok, so you only get distortion on low frequencies. If the box is an original I might think about any electrolytic capacitors that might have worn out. Some can vary over time and go out of tolerance. Does the distortion sound like a regular guitar overdriving an amp? Is it only on the lower string or can you make it distort by playing the othe strings hard? Is it just the lowest note or does it also happen if you play higher notes on the same string?

When you say that you swapped out the buffer input, did you replace the original IC with a new one?

[newperson]

There are a couple TANTs in there.  I did not remove them but did a quick test with my ESR meter and they both gave good results.  10uf and 15uf. 

Yes, it is a distorting sound that comes with a heavy hit.  It might be in the higher strings also but it is just harder to hear.  I know it will happen on the A string also.  It will distort/clip on any note on the low E string.  It is worse on the attack and then goes away after the note/string settles a bit.  Mis bias?

And I swapped out the input buffer 741 IC with a new one without any change.  It has a socket at that position at the moment. 

Thanks for the reply,

[Mark Hammer]

I'm going to suggest lifting one end of the feedback resistor.  The circuit provides no DC blocking or bass-trimming in the feedback path, and uses fixed rather than variable feedback.  So, as it sweeps low, it is easy to imagine that in some units the additional resonance provided by the feedback could exceed what one or more of the FETs is comfortable with.  If removing all feedback cures the problem then clearly the solution is to somehow curtail low end in the feedback.

At the same time, you want to retain at least some feedback.  Assuming too much bass in the feedback signal IS the source of the problem, there are two things you can try.

1) Nudge the sweep range up a bit so it doesn't sweep down quite as low and introduce too much bass feedback.  This is a simple matter of adjusting the trimpot to shift the sweep range upwards.  Downside is it may not provide3 as much gurgle as you like.

2) Insert a small cap in series with the 24k or 22k feedback cap.  Something in the .039uf-.047uf range will still yield audible resonance across much of the spectrum, but less for your heavier wound strings.

[newperson]

I have fiddled with the trim pot just to see and it has a very narrow working range. It did not change much, since the effect drops out quickly.

Does this model have a feedback resistor?  I have read about that for newer reissues.

[Mark Hammer]

Script and block issues differ in two ways.  One has a wider sweep than the other (3M3 resistor, rather than 3M9), and one has more feedback than the other (22k feedback resistor, rather than 24k).  I'm not aware of any issue that did NOT have a feedback path (apart from the 2-stage Phase 45).

[newperson]

I have been working off this,
http://www.electrosmash.com/images/tech/phase-90/mxr-phase-90-script-logo-schematic-parts.png

http://www.electrosmash.com/mxr-phase90

Which R* is the feedback to pull a leg up on?

[Mattnezz]

I once had a problem with the output mixer not providing enough overall headroom for humbucker input.
Can't remember how I solved the issue though

[newperson]

It is definitely worse on humbuckers

[Mark Hammer]

I have been working off this,
http://www.electrosmash.com/images/tech/phase-90/mxr-phase-90-script-logo-schematic-parts.png

http://www.electrosmash.com/mxr-phase90

Which R* is the feedback to pull a leg up on?

Interesting.  The schematic doesn't have one; which is why you can't find it.  I wasn't aware there was ever an issue without any feedback.  MInd you, the Boss PH-1 had no feedback path, and then added one in the PH-1r, so I suppose MXR could have had a non-feedback issue and then added feedback at some later point.   News to me, but I've been underinformed about other things before.

Apologies for the needless detour.

On the other hand, that may be the schematic you're working from but may not be the actual schematic.  Here's a script logo unit which does have a feedback resistor, visible as the "flying" 22k with the tubing on its leads.

[Blitz Krieg]

block?

[newperson]

That should be a newer one with dual amps instead of the 741.  This one does not have the extra 22K resistor that Mark's picture shows.

[Mark Hammer]

Correct.  It has 24k in the same position, which is slightly less feedback.  Note that the TL061 was not in production when the first P90s were being made (which is why they used a sextet of 741 chips).  I may be wrong, but I think it came later than the TL071.  The first FET op-amps I'm aware of were the National Semiconductor LF series, like the 351 and 356 (single) and 353 (dual).  The TI chips didn't come around until later.

Perhaps someone with more expertise in the history of op-amps would like to chime in ?