Understanding the Maestro Phase Shifter PS-1

[VintageGear]

As I am loving old gear, I won a Maestro Phase Shifter in auction. Psyched for its arrival, but unfortunately it did not work. Sigh. Ah well, worked with more pedals, so started out with the standard list of work: Replaced the el. caps and resoldered all solder joints. Hurray, I got sound. However, the phase is not audible. I hear the tone changes upon engaging ("slow phase/off"). Messing with the trimpots as per adjustment procedure gives me the phasing sound I am looking for, but only if I turn with my little screwdriver, hehe. So I am suspecting the LFO to be at fault.
Got myself a schematic, but turns out I don't really understand how it works... On top of the schematic, at the switch, we have two 2N3638A transistors (Q8 and Q9) and Q12 Q11 (2N4303). LFO? If so, what is the bottom part where the medium and fast phase switches are?

Who can clarify please?

Schematic: http://s000.tinyupload.com/?file_id=30481914756933457671

[balkanizeyou]

the LFO section is the bottom of the schematic (IC4, Q14, Q13). Check whether you have +-15V rails at the junctions of Zeners and 270R resistors. If you do, try socketing in another op-amp and see if that helps.

[PRR]

No-brain stuff first. Are the +/-15V rails correct? Without thinking, I can imagine that with say +3V/-15V supply, audio would pass but LFO would not go. Quicker to check rails than to think about it. And power problems happen, and may be easy to find and fix.

https://s12.postimg.org/lev37hykd/Maestro_PS_1.gif

> Q8 and Q9 and Q12 Q11 . LFO?

No timing cap. No loop feedback. And what do Q11 Q12 apparently do? Connect the buffered signal, or the effected signal, to the Output. This is bypass switching.



Which as you (and Mikolaj) say, leaves Q10...IC4. It has fast/slow notes. It goes TO the same point as the "bias pot" which you say does wobble the signal. You could tie a windshield-wiper to the bias pot!(?) Probably the signal off R2 does the same thing more conveniently.



> I am suspecting the LFO to be at fault.

Agree.

Q10 has a monster cap on its Base. This is a clue that we might have only a DC signal here, varying as slow as the user's knob-finger. This suggests Q13 Q14 IC4 are some sort of voltage-control oscillator. The 0.33uFd is logically the timing cap. IC4a is some kind of integrator, making a triangle wave to feed R2. Also feeds IC4b which is a Comparator with positive feedback, a Schmitt Trigger. This makes a square wave which bangs complementary N and P FETs, which switch resistors to Vcontrol or ground. There's a bazillion forms of tri-wave generators like this. Not interesting enough to analyze for free. Approach paths:

1) Get all the voltages. See what is "impossible". If an opamp's +in pin is positive of the -in pin, but the output hasn't gone positive, something is wrong with either the opamp or its connections. Likewise observation of FET voltages (watch polarity) will tell if the FET is on or off, and what that should feed the integrator. A bit of brain-pain may point directly to the problem.

2) Brain-pain hurts. Also this device must be OLD. Which means some "never happens" problems may have happened. R2 trim pot wiper may have grit or warped-wafer or crack in the track. I usually figure opamps never fail, but Mikolaj's suspicion may be spot-on. First I would glare at the 220uFd at Q10 Base, and probably replace it without hesitation. While there I would get it out in bright light with a magnifier and study ALL solder joints for perfection. If not visibly perfect, at this age I might *gently* suck the old solder off and do it right. Film caps are tough, but low-price ones can fail, consider the 0.33uFd.

[VintageGear]

That's why I love this forum. Thanks, all makes much more sense now! I will report back with my findings!

[VintageGear]

Turns out I had the wrong version of the schematic. It is version 2 / 3 instead of v1 of the phaser. My unit however runs at 15V, although the schematic says 12V. My bad, did not look hard enough at the actual LFO and missed it did not have two FETs in place! Luckily found the correct version on the web as well.
https://s3.postimg.org/edrbcjuib/maestro.png

So far: Resoldered all solder joints and replaced electric capacitors out of precaution. (Bonus question: Why are those radial capacitors so large in size compared to my modern ones?)

Voltages are at +15 and -15 volt.
LFO Q9 PNP: E: (+1.72v) B: (+1.08v) C: (-15v)
Q10 FET: G: (2.30v) D: (1.72v)  S: (1.72v)
IC4 (MC1458): 1: -13v 2: 1.1v 3: 0.86v  4: -15v 5: -6.68v 6: 0v 7: 14v 8: 15v
http://circuits.datasheetdir.com/37/UTC-MC1458-pinout.jpg

All voltages are fixed. If the LFO would be working, I would see the voltages differ over time, right? So.. Theory time.. No more complementary banging on the FETs, as one of em has fallen off the wagon in this (accurate) circuit. I understand the combination of Q9 Q10 and IC4 will generate some sort of waveform.
But the Q10 G is high, allowing S->D, right? Therefore shunting the signal to ground all the time? Thus, IC4 does not give a varying output with a constant high 14v?

[R.G.]

The LFO circuit in this unit is a variant of a voltage controlled oscillator. The integrator ramps up and down depending on the difference between the input + and - pins. The + input is always sitting at 1/2 of the input voltage from the bipolar because of those two 10K resistors. The - input is either the full voltage from the bipolar or ground, always through 150K. So the integrator ramps either up or down at a rate of

dv/dt = (Vin/150K)/C

The output is monitored by the second opamp running as a Schmitt trigger for its peak and valley voltages, and switches the input voltage to make it reverse directions when it hits either a peak or valley. Note that the JFETs are an N-channel and a P-channel so while one is on, the other is off, and the same gate signal turns one on and the other off no matter what.

Paul's comments about debug are right. The key is making sure that the integrator integrates and the Schmitt trigger switches. Check the voltage at Q10 emitter and ensure that it is constant and non-zero. Then verify that the voltage at pin 3 of the first opamp is half of that. Then verify that the Schmitt trigger flips when the voltage on its input exceeds + and - boundaries.

The Schmitt section has positive feedback through the 33K to the input 10K. So when its input goes positive by more than the trip voltage, its output goes positive. The output then sits as near the positive power supply as the 1458 can go, perhaps V+ minus 2V. Call it 13V. The 33K/10K feeds back 13V* 10K/(10K+33K) of this or 3V to the + input, so the input signal has to go below the reference voltage on the inverting pin (zero in this case) by that much to make the output flip the other way. So the output stays positive until the input from the Schmitt trigger gets to -3V, then flips negative. Works the same on + excursions, so the output of the integrator ramps up and down between +3V and -3V.

[VintageGear]

Thanks a lot RG for tuning in as well, very educational read! Prior to this pedal I never heard of Schmitt triggers, so it is already earning itself back in terms of knowledge :-)



I've intensively looked at the trigger as you described. In v2 there is one JFET less. Makes sense as it would always switch to ground as the other one would come on, right?
I sucked and resoldered all joints at the height of the LFO. Replaced the 220uF. Unsoldered the trim pots and cleaned them, no troubles there, they measure alright over the entire turn. I unsoldered the 0.33uF and placed it in my components-tester (which i got from ebay not too long ago). Says it is alright. So the timing should be alright.
Q10 (Q9 in the latest schematic) is always on +1.72v and indeed, it is half that at the opamp through the 10k/10k rig. But the problem is that there is no voltage fluctuation at all... So the Schmitt will not trip into firing from +14 to -14 either, right?
Furthermore, pin 5 (-6.68v)< 6 (0v) --> wouldn't that result in 7: -14v instead of +14v?
IC4 (MC1458): 1: -13v 2: 1.1v 3: 0.86v  4: -15v 5: -6.68v  6: 0v 7: 14v 8: 15v

Am still bit puzzled, but getting there I think :-)

[R.G.]

I've intensively looked at the trigger as you described. In v2 there is one JFET less. Makes sense as it would always switch to ground as the other one would come on, right?

This one saves one JFET by letting the - input be either at full input voltage (and therefore 2x the voltage at the + input) or ground depending on the JFET switch. You could test this by removing the JFET and manually shorting or un-shorting the JFET source and drain. The integrator will rise all the way toward V+ in one condition, ramp down to nearly V- in the other. Notice that for manual, the Schmitt trigger won't do anything. But you can manually test that the integrator runs to both extremes, and that the Schmitt does click full up and full down. If that happens but the LFO does not run with an input voltage, then the Schmitt is not turning the JFET on and off, so it may be a problem between the Schmitt output and the JFET gate, or the JFET could be hosed or some earlier genius may have replaced the JFET with a different type/pinout.

Q10 (Q9 in the latest schematic) is always on +1.72v and indeed, it is half that at the opamp through the 10k/10k rig. But the problem is that there is no voltage fluctuation at all... So the Schmitt will not trip into firing from +14 to -14 either, right?
Furthermore, pin 5 (-6.68v)< 6 (0v) --> wouldn't that result in 7: -14v instead of +14v?
IC4 (MC1458): 1: -13v 2: 1.1v 3: 0.86v  4: -15v 5: -6.68v  6: 0v 7: 14v 8: 15v

Hmmm. It does look like the Schmitt side ought to be full negative. Try the manual ramp up/down process by removing the JFET and manually shorting/unshorting the JFET source and drain pads. IF the Schmitt never flips when you do that, either there is a short on the output of the second opamp, or the opamp itself is hosed.

[VintageGear]

Thanks a lot once again for the elaborate reply. I will continue next weekend when I have the time for it and report back!

[VintageGear]

Hurray!! 10 points to everybody! Substituted the MC1458 with one I got from a local electronics shop and wooshh, there it was, that delicious warm lush phasing sound. Loving the way it ramps up if you flick in the fast and medium phase switches.
Learned a lot. Definitely worth the hassle and hours of measuring/thinking/resoldering/recapping!