Tonepad phase 90 issues

[patricks]

Hi all,

I've put together a Phase 90 clone from tonepad and it's almost working but won't phase. Tweaking the trimpot will get it to phase a bit, but it won't sweep like it should.
Here's a link to the schematic: http://www.tonepad.com/getFile.asp?id=42
I haven't modified anything, but I have substituted the 0.05uF caps for 0.047uF and the 15uF for a 10uF
The pedal's powered by a 9V OneSpot, here are the measurements:

Tl072
IC1                              IC2                            IC3
Pin                              Pin                             Pin
1: 4.77    5: 3.25       1: 4.76     5: 4.66      1: 4.73    5: 4.66
2: 4.76    6: 1.27       2: 4.76     6: 4.76      2: 4.73     6: 4.73
3: 3.24    7: 0.00       3: 4.66     7: 0.00      3: 4.66     7: 4.77
4: 0.00    8: 9.55       4: 0.00     8: 9.54      4: 0.00     8: 9.53

5.1 Zener Diode=  4.76
1N914 = 9.54

Matched 2N5952
Pin = all 4
1: 1.49
2: 4.76
3: 4.66

2N4125
C: 2.93
B: 4.17
E: 4.76

Could it be something to do with the fact that the voltage at pins 3 and 5 of IC1 is 3.25V, whereas for IC2 and IC3 the voltage at pins 3 and 5 is 4.66V?

[armdnrdy]

The 15µf cap is in the LFO.

It has direct influence on the sweep.

[patricks]

Cool, I'll solder a 4.7uF cap in parallel - the 10uF is electrolytic, so the 4.7uF should also be electrolytic, shouldn't it?

I just noticed something about the 250k trimpot, too - if the trimpot is in the wrong way around, would that make a difference?. The schematic (http://www.tonepad.com/getFile.asp?id=42) says that pin 1 of the trimpot is connected to the source pins of the JFETs and to the cathode of D2. Based on the datasheet for the trimpot I bought (http://www.taydaelectronics.com/datasheets/RM-065.pdf), I've got it in the wrong way around.

[armdnrdy]

"the 4.7uF should also be electrolytic, shouldn't it?"
Yes....a 4.7uf poly cap would be huge!  

The trimpot is a voltage divider...so if pin 1 & 3 are reversed from the schematic/layout it won't make a difference. The voltage will increase/decrease the opposite then if oriented "correctly".

[jimilee]

Are the jfets matched and oriented correctly?

[patricks]

Great, thanks! I picked up a 4.7uF electrolytic from Ratshack this morning, so I'll solder it in and see how the pedal sounds.

I'm certain that the jfets are oriented correctly, I got the ones specified in the bill of materials and double-checked their orientation. I'm pretty sure they're properly matched; I built RG's jfet matcher, but all the Vgs measurements came out at between 0.286V and 0.648V. The four I chose had reading of 0.373V, 0.377V, 0.379V and 0.383V, so they're within 3%.
I haven't investigated why the jfet matcher's reading low, because I posted in this forum about it and another member said that they'd experienced the same thing and their matched jfets worked fine.

[Jaicen_solo]

Which Zener are you using? You might need to drop it to something like a 3v Zener with those Jfets.

[patricks]

Ah, I wondered about the zener. I've tried 4.3V, 4.7V and 5.1V diodes and it definitely changes the character of the sound, but it didn't actually start to phase with any of these.

Back to ratshack...

[Seljer]

I had to lower the value of the resistor that supplies current to the zener to get the voltage high enough to be able to bias the FETs into the proper range. Though your voltages seem to be ok, mine were like 3.3V at first 

[Mark Hammer]

The zener's function is to turn a changing voltage source from a battery into a stable and consistently valid bias voltage for the FETs.  So, the zener gives the trimpot 5.1Vdc for as long as the battery is within range of owering the rest of the circuit, and the trimmer divides that down to whatever the FETs need.  You could make a P90 without any zener whatsoever, but as the battery voltage changed over time, you'd need to keep rebiasing the FETs.

I would assume that the zener voltage selected was chosen to strike a balance between providing as broad a range of potential bias voltages for the FETs as possible, and a maximum voltage as far away from the battery's starting point as possible to provide the most long-term stability.  A person could probably use 6.2V or whatever, though you may have to rebias.  You could also use 3.3V but may not find the bias voltage you seek within the 0-3.3V range.

[patricks]

Hmm, so are you saying that changing to a lower value zener isn't likely to have an effect?
I don't know a lot about biasing FETs, but if the 5.1V zener is OK, do I need to change the value of the trimpot or another resistor?

[armdnrdy]

Great, thanks! I picked up a 4.7uF electrolytic from Ratshack this morning, so I'll solder it in and see how the pedal sounds.

What was the outcome of adding the cap?

[patricks]

No luck with an extra 4.7uF capacitor in parallel, unfortunately. A funny thing happened, though. I thought that maybe the FETs weren't seated properly in their sockets, so I wiggled them around a bit. When my thumb came into contact with the underside of the circuit board, I picked up radio signals!  

If changing zeners is the right way to go, I'll order some from Small Bear tonight. Any other suggestions are definitely welcome, though!

[armdnrdy]

Does the voltage change on pin 1 (the gate) of the JFETs when you adjust the bias trimmer?

[Seljer]

Something is iffy with your LFO. Look around pins 5, 6 and 7 of IC1.

If you set the rate knob to the lowest settings you should notice the voltages on those pins moving around. Pin 7 should jump rather abrutly between about 2v and 7v. On pin 6 the change should be a bit less intense and smoother (basically a triangluar waveform if you were to hook it up to a scope).


Using a 10uF instead of a 15uF just make the LFO slightly faster overall.

[Mark Hammer]

Hmm, so are you saying that changing to a lower value zener isn't likely to have an effect?
I don't know a lot about biasing FETs, but if the 5.1V zener is OK, do I need to change the value of the trimpot or another resistor?

The zener value will only have an "effect" to the extent that it does, or does not, provide a suitable range of bias voltages to tune from/to.

The trimpot value provides some limiting of the current fed to the FET gates.  Apart from that, all it does is divide down the voltage provided by the zener.

I know that one, or maybe even two, of the P90s I built from earlier Tonepad layouts had the FETs oriented backwards.

[patricks]

It's alive!

I rechecked the voltages at the gates of the FETs and they changed when I turned the trimpot, so then I checked pins 5, 6 and 7 of IC1. All the voltages were constant, but the "look at the LFO" advice got me thinking of the nearby components. It turns out that the wire connecting the board to pin 1 of the rate pot was shorting out against the box!
I clipped away the excess, put down some electrical tape for good measure and tried again; no luck with the 5.1V zener, but the 4.3V zener got the voltages on pins 6 and 7 dancing around 

Only a couple of issues remaining now. The first is that the "rate" pot doesn't have much effect for the first third of its range and the second is that when I dial in a fairly lush sound I can hear the LFO ticking. Come to think of it, the sound isn't all that lush, really. Nowhere near as swirly as the other pedals I've seen in youtube clips. I'll try the 4.7V zener and see whether that makes a difference...

[Jaicen_solo]

Do a search on this board for Phase 90 ticking and you'll find the solution I used. It's fairly simple and mostly eliminated the problem I was having.
I would suspect that the swirliness could be a consequence of the Zener, and the mixing resistors. However, the phase 90 is fairly subtle compared to some thing like a Smallstone.
The lushness is perhaps a consequence of the pedal itself, it does have a trebly character with a lot of lows frequency roll off. Try raising the input and output caps.

[patricks]

Apologies for posting without searching, that was a dumb move!

Anyways, the 4.7V zener did the trick. A nice, lush sound and no LFO ticking, hooray!! I'm going to solder in the zener and the FETs, then I'm done with pedal building for today

[armdnrdy]

Good to hear you got it going. Changing the zener is probably the most common fix for a non phasing issue.

But what about your reply (#7) where you stated that you tried different zeners (including a 4.7V) and it still didn't work?