Mu-tron Phasor I - schematic & a few technical questions

[aion]

I've got an old Mu-tron Phasor I in pretty rough shape. It's in dire need of standard electrolytic cap replacement, testing solder joint and offboard wire integrity, etc. - but while I have it open, I'm interested in trying to figure out how it works. There's not a lot of good info on this one, just some low-quality factory schematics, and not much analysis in DIY since it's often been overshadowed in favor of the more popular Phasor II (the optical one).

I have two schematics for the unit. The first is a factory schematic that's mostly legible except for the power section. The second is a pretty good redraw, but from a much lower-quality version of the factory schematic, so while it's more readable it likely has a lot of inaccuracies.

High level, it's a 6-stage OTA phaser, using RCA CA3080S (the "S" suffix denotes a metal can with DIL formed leads). However... what sparked my curiosity was the inside:



Each of the 3080s is marked with a sticker from 377 to 380. This is something I haven't seen in other Phasor I gut shots, but these are brittle enough that they would fall off with a light breeze so probably most of the stickers just haven't survived on other units.

The transformer is also marked with a large sticker that says "380 #21". This could be referring to something about the transformer, but I'm guessing it was just a convenient place to put the unit sticker, with "#21" being the serial number and "380" being the nominal OTA spec used in the unit.

A few questions:

1. What parameter is most likely being marked? I didn't see anything in the datasheet that is measured in hundreds and would be anywhere near 400, but it seems more likely it's just something measured in smaller units for better precision. For instance, the 3080's typical input offset voltage is listed at 0.4mV, which is 400 microvolts.

2. Several later phasers such as the EHX Small Stone use similar OTAs (CA3080, CA3094, LM13600) without sorting. Is Mu-tron's sorting just a side-effect of the early metal cans with less consistent manufacturing specs, similar to how we don't really need to use the 2k balance trimmer on Dyna Comps anymore? This is one of the earliest OTA phasers, perhaps the first, so maybe it was just an abundance of caution without having other examples to draw from.

3. It appears to run on +/-12V. Am I right that there shouldn't be any significant difference in operation (beyond just reduced headroom) if it was powered with +/-9V?

If there's any critical info missing from the schematic that can't be inferred, I can check my unit and try to fill in the blanks. I'm planning on doing a full trace at some point but was hoping to understand it a bit more before diving in.

[CheapPedalCollector]

I repaired one of these a couple years back, it still had all the stickers on the OTAs. I believe they were binned for offset voltage as you suggest.

I found the caps had all failed, and some of the 4558s also failed, as well as the 110 ohm resistors being stressed in the PSU and I replaced the zeners for good measure as they were also giving off readings.

I replaced all the 4558s with NE5532s for a significant noise reduction with no change in sound, bonus is the opamps have input protection diodes so hopefully it won't fail again any time soon, even if overdriven by an effects loop.

The older schematic is the correct one for this unit, but the part values are hard to read, I was under time constraints or I would have drawn a new higher quality schematic for it.

I found that the standoffs become brittle, and also flex a lot which caused a lot of cracked solder joints. I replaced them with aluminum ones of the same depth. I also resoldered every joint to help make it more bulletproof as it's being used at gigs. The foot switch had become intermittent too and was replaced.

Yes it's +/- 12v. It's designed for use with effects loops and stuff so I think it's good to stick with the original PSU. For guitar/bass +/- 9 should be fine I would think.

[Rob Strand]

I have two schematics for the unit. The first is a factory schematic that's mostly legible except for the power section. The second is a pretty good redraw, but from a much lower-quality version of the factory schematic, so while it's more readable it likely has a lot of inaccuracies.

Yes, I can see the cap at the output of the phase shifter is 0.05u on the original and 0.01u of the redraw.

Each of the 3080s is marked with a sticker from 377 to 380. This is something I haven't seen in other Phasor I gut shots, but these are brittle enough that they would fall off with a light breeze so probably most of the stickers just haven't survived on other units.

The transformer is also marked with a large sticker that says "380 #21". This could be referring to something about the transformer, but I'm guessing it was just a convenient place to put the unit sticker, with "#21" being the serial number and "380" being the nominal OTA spec used in the unit.

The red sticker on the transformer might be normal but only with the simple number not the 377 to 380 marking.

As for the 377 to 380 I've got no idea.   
The odd things are:
- 380 on the transformer and the CA3080 means it's not just related to the chips.
- repeated use of the numbers
- different pen used on 377

At first I thought it might be the date code on the CA3080 device, that would explain multiple markings of the same number but it doesn't explain the transformer marking.   It also means the 3xx number is unlikely to represent a measured parameter.

It's possible there was some sort of in-house testing done and people marked the parts after checking.    However a large number like 379 seem too high for an employee number.  Also the handwriting looks the same for those 3xx numbers.   If it was some sort of test/check marking the CA3080's would be expected to be checked pre-assembly as different number parts got different numbers.  The meaning of the number is still a mystery, especially due to the transformer marking.

These things are pretty old and lots of techs have probably looked inside so who knows the markings might have been done by a tech during debugging.  In this context the transformer marking is still weird.

3. It appears to run on +/-12V. Am I right that there shouldn't be any significant difference in operation (beyond just reduced headroom) if it was powered with +/-9V?

Changing the supply will change the LFO sweep, perhaps swing and frequency.   You would have to change some parts to compensate.  The IABC current depends on the IABC pin voltage so at 9V supply the IABC pin voltage will cause a more rapid decrease in the IABC at the bottom of the LFO swing.    It's all details that would need thorough checking.

If there's any critical info missing from the schematic that can't be inferred, I can check my unit and try to fill in the blanks. I'm planning on doing a full trace at some point but was hoping to understand it a bit more before diving in.

- I'd be confirming the caps values
- Maybe do a rough check of the number and values of resistor to see if it matches the schematic.
  If there's a mismatch try to work out which ones are different.

[Rob Strand]

I repaired one of these a couple years back, it still had all the stickers on the OTAs. I believe they were binned for offset voltage as you suggest.

380mV seems like a large offset.  Also the transformer marking confounds that theory.

[CheapPedalCollector]

I couldn't come up with another theory, I think the ones in the one I worked on were marked 390.

It also had all dark brown piher resistors in it. I swear I took a pic of the top of the board, still digging.

[Rob Strand]

I couldn't come up with another theory, I think the ones in the one I worked on were marked 390.

It also had all dark brown piher resistors in it. I swear I took a pic of the top of the board, still digging.

It could be that the marking on "transformer sticker" isn't normal.  I saw the red sticker on the metal shield plate in one unit.  Someone could have put it on the Tx.   That being the case the marking on the OTA might mean something, perhaps the transconductance at some low IABC current.  The current being some value in their
in-house test jig.

The schematic has a note that R53, R54, R55, R56 are selected based on the OTA spec.   That implies some sort of categorization of the OTA.

All a bit confusing.


I had a look at the caps and got this far.

Mutron Phasor I

1972 to 1974   ;red sticker metal shielding plate, or on transformer.
1975 to ?      ;red sticker on PCB

1800pF cer   12   C4 to C15      ; 6x OTA stages, 2 cap per stage ; large diam
220n poly   2   C17, C27      ; polyester, brown (Mat) or topical fish (Phil)
100n poly   2   C1, C26        ; polyester, brown (Mat) or topical fish (Phil)
47n poly      1   C3      ; polyester, brown (Mat) or topical fish (Phil)
10n/50n   cer   4   C21, C22, C23, C24 ; med diam
10n/50n   cer   1   C16            ; med diam
270p cer      1   C2      ;small diam

Electros:
470uF 25V   2   C19, C20      ;1972 to 1974 use axial, 1975 and after radial
50uF 25V    2   C25, C28???
4.7u?      1   C18

*** There seems to be an extra Electro on the PCB which isn't on the schematic.
I've called it C28, probably connects across A9/A10 +ve power rail.
However, it might really be C21 or C23 as there's too many caps.

[aion]

At first I thought it might be the date code on the CA3080 device, that would explain multiple markings of the same number but it doesn't explain the transformer marking.   It also means the 3xx number is unlikely to represent a measured parameter.

I read the sticker as "serial number 21, using OTAs from sorted bin 380", with bin 380 perhaps containing OTAs measuring from 375 to 385 (in whatever spec is being sorted, possibly offset voltage or transconductance at X current) and the transformer is just one of a few convenient surfaces for the sticker. But I would be very interested in CPC's photo of the top side if he manages to find it, just to be able to compare a second unit.

What's the easiest way to measure offset voltage in an OTA? Is there a simple test circuit for this? I would be more than willing to pull a couple of these to see if the measured offset voltage is anywhere near the marking, or if parts marked with the same measurement test the same.

[ElectricDruid]

What's the easiest way to measure offset voltage in an OTA? Is there a simple test circuit for this?

I dunno, but it would have to be at a specific Iabc current because of the control feedthrough. E.g. Offset voltage varies with Iabc. That was one of the things the CA3280 was supposed to do better.

[Rob Strand]

I read the sticker as "serial number 21, using OTAs from sorted bin 380", with bin 380 perhaps containing OTAs measuring from 375 to 385 (in whatever spec is being sorted, possibly offset voltage or transconductance at X current) and the transformer is just one of a few convenient surfaces for the sticker. But I would be very interested in CPC's photo of the top side if he manages to find it, just to be able to compare a second unit.

To me the 380 marking on the red sticker is anomalous.  I'm happy to ignore it until someone posts another unit with the same marking.   Marking the board with a serial number pretty common.  However it is usually done on the bare board, before the board and transformer meet.   I did see a pic with the red sticker on the metal shield so maybe that's how Mutron did things.

What's the easiest way to measure offset voltage in an OTA? Is there a simple test circuit for this? I would be more than willing to pull a couple of these to see if the measured offset voltage is anywhere near the marking, or if parts marked with the same measurement test the same.

See figure 1 for example,
https://www.experimentalistsanonymous.com/diy/Datasheets/CA3094.pdf

The offset should be small.   An offset 380mV isn't normal.  You would have to inflict some meaning on the number.  For example 380uV offset or 380mV at the output of some jig.    The thing is test jigs don't have to follow the datasheet so the number from the output of a test jig can change.   Also note it's best to test offset with a low IABC current and with small input resistors otherwise mismatches in the input stage transistor hFE's will add to the offset.   For a simple test jig it might be possible to have an imprecise number!

I'm not not too enthusiastic about the offset voltage theory.  *Matching* the offsets in a particular unit doesn't add any goodness.  What you want is outright *low* offsets to reduce thump.

The comment on the schematic about choosing the LFO resistors based on the OTA is probably the only evidence for sorting the OTA.   Those resistances are related to LFO sweep.   Sweep relates to transconductance more than voltage offset.  IMHO the measuring the transconductance at low IABC is a possible.  But even then a phaser would be pretty robust against a significant amount of mistracking when the 6 stages are put together.

If you look at the CA3094 datasheet, on-paper Transconductance can vary +/-30%.   Maybe Mutron decide that was too much.   

Very hard to second guess this stuff.  And who knows if the reason for sorting is even justified in the end.