debugging for the vero phase 45 - "final" report, sound clip

[barret77]

Hello all

I've built the torchy's/maillet's univibed phase 45 in veroboard, and couldn't get it to work so far.

My problem is similar to another recent post, but I wrote down the voltages

-> diagnostic:
-sound does not change when effect active/bypassed;
-rate led stays on all the time; does not oscillate - stays on when effect is active or bypassed.
-pots do not change almost anything, very very very subtle change.

- > basic checking performed:

FET and IC pinouts checked;
solder traces; cut veroboard traces;
pot/switch/led wiring.

-measured voltages:
( negative reference: battery (-)  pole)

IC1:
1 - 3.23
2 - 2.57
3 - 1.92
4 - 0
5 - 3.39
6 - 3.37
7 - 4.05
8 - 4.71

Q1:
d 3.13
s 3.13
g 0.59


IC2:
1 - 4.52
2 - 3.32 to 4.00
3 - 2.93 to 2.54
4 - 0
5 - 3.11
6 - 3.13
7 - 3.60
8 - 4.44

Q2:
d 3.01
s 3.04
g 0.56

zener diode:  3.26
rate led:  1.81


any ideas based on that? I don't know what to do...

Thanks a lot

[petemoore]

IC1:
1 - 3.23
2 - 2.57
3 - 1.92
4 - 0
5 - 3.39
6 - 3.37
7 - 4.05
8 - 4.71 All other pins should be about half of supply, this is the supply pin. excepting of course pin 4 which is Gnd [0v].

Q1:
d 3.13
s 3.13
g 0.59
 

IC2:
1 - 4.52
2 - 3.32 to 4.00
3 - 2.93 to 2.54
4 - 0
5 - 3.11
6 - 3.13
7 - 3.60
8 - 4.44 This should be about battery voltage [iirc there's a smallresistor between battery and V+] including battery voltage makes all the other numbers relevant...these voltages aren't biasing this opamp though.

Q2:
d 3.01
s 3.04
g 0.56
  I'd go with trying to fix the opamps biases, and wait for feedback on the Q voltages, I can't expound or elaborate on Jfet's bias voltages, used as variable resistance units.
  Anyway, hopefully this place to start may help sovle the ph45 problem.

[barret77]

thanks, I'll investigate that tonight...

[barret77]

hey, just one more thing: is it normal that the rate led stays on even when the effect is bypassed?

Thanks!

[slacker]

yeah the rate light flashes (or not in your case) all the time whether the effect is bypassed or not. You can use  it as a status LED as well you want. If you're using a 3PDT switch just wire the cathode of the LED up to the switch like you would for a normal status LED. Or you can use the millenium bypass with it, to do this remove the LED from the board and connect it to the millenium instead then take the power line of the millenium and instead of connecting it to 9volts connect it to where the anode of the LED was connected to the board.

[barret77]

Hello all

Starting to lose the hopes around here...

I checked the IC voltages at pin 8 - turned out that my battery was worn out, therefore the voltages were weird. A fresh battery showed the same voltage in the source and on both pin 8.

There's still no oscillation of the led, and no change in sound when the effect is active (sounds pretty much the same than when bypassed, very subtle change).

My question is: can this be due to bad jfet matching? Is this a symptom of non matching - no effect at all? I've built the jfet matcher, and despite the low voltages both fets matched... there's a change that I've built the matcher wrong, but it is remote...

Can anyone give me some hints of debugging procedures from this point? I pretty much don't know what to do...

Thanks a lot!
Anxious to play with the phaser...

[MartyMart]

Torchy's layout is a worker, 100%
There are two connection "jumpers" direct from the 9v connection ( Red strip on layout )
to both op-amps pin 8 ....
So before worrying about the jfets, you must go through the build very carefully.
Make sure of those connections, all values used and all the "copper trace cuts" are good
in the right place and "clean".
It's quite easy to leave a "tiny trace" of copper which will cause a bad connection, it's also
quite a big and packed layout, so take some time off, then come back for a serious check.
Do you have the correct "zener diode" at Z 1  ?  as this is part of the "V-ref" supply, for creating
4.5 volts .. ?

Stay calm 
Check all parts/values/orientation/cuts and jumpers, against the layout ...

Marty.

[barret77]

Hello all

So, here's my phase drama again. I'm very thankful to the feedback I had so far, and I'd like to thank in advance any insight, idea or suggestion for debugging my little child.


update: after cleaning the vero, I removed a really tiny copper fragment and activated the rate led. That's the progress so far.

Here's the pin voltages and description:

BATTERY:
9.46 before being connected
8.95 after

IC1:
1: 4.32
2: 4.30
3: 3.09
4: 0
5: 4.58
6: 4.67
7: 4.78
8: 8.84

IC2:
1: 4.40 - 4.54
2: 4.54-4.55
3: 4.11-4.15
4: 0
5: 4.57
6: 4.67
7: 4.57
8: 8.66

Q1:
D: 4.57
S: 4.66
G: 0.83

Q2:
D: 4.57
S: 4.66
G: 0.83

Z1:
A: 0
K: 4.66

RATE LED::
A: 0
K: 1.60

present status:

-Led blinks and speed pot works, adjusting blink speed;
-There are subtle (very subtle) changes when adusting bias and mix.
-The audio probe revealed that the sound is the same on all the mix pot ends. (is this normal?)

-The audio is basically the same when bypassed or not. the difference is in the tone, very subtle. No phasing whatsoever.

-Jfets were *kind of* matched (they also show same voltage readings). Tried a batch of diferent jfets for testing, no difference.

-adjusting the bias pot change the both jfet G voltage from 0.81 to 1.55

-If I remove the jfets from its sockets while the audio is playing, the difference in the audio is reaaaaally subtle - the jfets connected or not make almost no difference in the sound output.

-RDV posted his working voltages; his jfet G values are much bigger, the same as D and S on Q1 - Is this a hint of a problem? I tried different jfets and they behaved in the same way, ranging from 0.8 to 1.55.

With these hints, what you guys think?

Again, thanks a lot.

[MartyMart]

OK, getting there, now the op amps look good 
Sounds to me like you only have "one half" of the signal getting to your output, there are
"two halfs" - if you like, which will give you the "Phase" tone ..... so :
Check your "Mix" pot connections and all board parts/cuts related to those board connections.
This pot is going to mix these two halfs together so is where I would be looking ....
Once again, a not quite "cut" copper trace will work as if its not been cut at all !!!
If the fet voltages dont look right, check those connections/cuts/values also
MM.

[smccusker]

i know you said the differences of the mix pot are subtle, but this tripped me up a bit while building mine. The mix pot does very little before about 12o'clock, at about which the phase effect will quickly appear at its fullest, turning the knob any further will lessen the phase effect and increase the vibrato.

[MartyMart]

i know you said the differences of the mix pot are subtle, but this tripped me up a bit while building mine. The mix pot does very little before about 12o'clock, at about which the phase effect will quickly appear at its fullest, turning the knob any further will lessen the phase effect and increase the vibrato.

A good point, there's only a small area where the "phase" becomes obvious !

MM.

[barret77]

following the mix pot clue, I checked with my DMM and found out that instead of 22k, the max resistance was 15k... that's some tolerance! I added a 10k resistor on the central lug, , so now the resistance goes from 10 to 25k.
The only difference I got is that at max settings, there's some slight vibrato going on - still no phasing, at any bias pot position...

I also "twisted" the board in every possible way to try to find out any bad solder joint...

I'm not sure how to track the low G voltage on the jfets.. is there any spot I should look for?

I'll keep the investigation tonight... Thanks a lot for giving feedback.

[MartyMart]

Check values and wiring of R15 - R16, ( 470k's )  C7 and C11 these are coming of the Fet "gates"
also the electro cap off gate  Q1- C13
Check also for any small "bridges" between source and gate at both fets

I'm sure that you're almost there, dont get despondant as this is a BIG build !! 

Marty.

[slacker]

sounds like you've got problems with your Jfets have a look at this post, especially RG's reply
http://www.diystompboxes.com/smfforum/index.php?topic=37859.0

hope you get it sorted, it's a great pedal

[barret77]

Hello all

I'm here to post an update on the phase 45 debug...

1) the bias pot was out of spec. It's a alpha pot, from mouser, 100k - but was reading 80k max.
I've added a 22k resistor to the lug 1, since the area of operation is on the counter-clockwise beginning of the pot, according to many posts.

2) there's a steep voltage drop on the battery - even a brand new one is operating around 7v after connected for an hour. I don't know why, everything looks normal. So I connected a wall wart operating at 10.2 V - and it worked! I wonder why the pedal is sucking so much juice from the battery. I spent 3 new ones during the debug, before pluging the wallwart.

So now it's working! but there's still some work to do. The effect is still very subtle, I believe I should try different pairs of matched jfets (am I right?) - also, what's the deal with the bias pot? Why the schem requestes a 100k pot if it only works on the initial notches?

thanks for all the help! the feedback was precious.

[smccusker]

Why the schem requestes a 100k pot if it only works on the initial notches?

i BELIVE that the bias pot was included to make matching the fets a little quicker. you'd plug one in and sweep the bias pot until you can hear the phase effect. if this happened at around twelve oclock, put that one aside and find another, if not, try more until you found two that did.

[MartyMart]

If you've got battery draining like that, then you probably have a "short" somewhere and I bet
the battery gets a bit hot.
9v or 4.5 volts is getting pulled to ground by a tiny mis-wire ....
You're close though
Build a Jfet matcher, look at GEO above, invaluable for this kind of phaser.

MM.

[barret77]

Hello all

I'm writing to complete the information on the phase 45 saga, and then make this post more useful to other trouble shooters.

- I figured out that most of my problems were coming from the lousy pot tolerance; I balanced them with resistors, so the max (or min) value would be achievable.
- I rebuilt my jfet matcher, and this time it really worked.  I got a matched pair at 1.89 - but found out that another almost matched pair (1.95 / 1.96) was sounding better - less subtle.

The pedal now works both on battery / wall wart. But I think the slightly higher voltage of the wall wart makes it sound better.

The pedal is still subtle; I know that is the nature of the phase 45 being subtle, but my pedal is much  more subtle than RDV sound clip, for example. It still sounds more like a tremolo than a phase pedal, but anyway, I intend to record some sound samples and post to see what you guys think.

Thanks for all the help the development of this baby!

[barret77]

here's a sound clip... it isn't subtle when recorded after all. But the univibe still sounds more like a standard tremolo...

guitar > phase45 > preamp > sound card
no effects/edits added
first half, phase switch - slower then faster speed. Second half, univibe switch.

http://www.lunetta.com.br/musica/testp45_96.mp3

[petemoore]

  Sounds like a 45 clone to me, a bit more 'wobble' to it than the stockversion...kudos !!