General Guitar Gadgets - Phase 90 Clone issue

[bettsaj]

So I've just finished a phase 90 clone, General guitar Gadgets and I have a problem that hopefully someone can assist with. I think i know what the issue is, but just want clarification. The issue i'm getting is no effect at all.

The phaser has been built with 2N5457 Jfets, and a 2N4125 Transistor. All the IC's are LM741's

The voltages are as follows, and were taken with the rate knob in the fully counterclockwise position:

All JFets Q1 thru Q4 are reading

Gate = 5.05 to 5.06
Source = 5.16
Drain = 5.16


IC1 pins are reading

1 = 0.002
2 = 5.15
3 = 3.52
4 = 0
5 = 0.002
6 = 5.15
7 = 9.25
8 = 0

IC2 thru to IC5 pins are reading

1 = 0.002
2 = 5.16
3 = 5.06
4 = 0
5 = 0.002
6 = 5.15 to 5.17
7 = 9.25
8 = 0

The issue comes when we get to IC6.

1 = 0
2 = 5.14
3 = changes between 3.92 to 5.38 constantly
4 = 0
5 = 0
6 =  Changes from 1.85 to 8.61 constantly
7 = 9.25
8 = 0

I suspect IC6 is at fault as according to GGG's info the voltages on pins 3 & 6 should be around 4.7 & 4 volts respectively and shouldn't move around on their own. Would that all sound correct?.... I believe so, but just wanted it clarified.

[bluebunny]

I suspect IC6 is at fault as according to GGG's info the voltages on pins 3 & 6 should be around 4.7 & 4 volts respectively and shouldn't move around on their own. Would that all sound correct?.... I believe so, but just wanted it clarified.

IC6 is the LFO, so I would expect the output to vary.  I don't know why JD would suggest otherwise?  He does say that the JFET gates should vary (otherwise there would be no phasing) and they're connected to the LFO... 

[bettsaj]

He says on the pedal info page that the voltages will vary on IC6 when you turn the rate knob, but the voltages on pins 3 and 6 are changing by themselves when you hold the multimeter on the pins.

All the other voltages on the other IC's are constant and don't change.

[bettsaj]

Also, before anyone asks... The Jfets were bought as a matched quad set from www.stompville.co.uk

The rest of the components were purchased separately... i only bought the PCB from GGG

[bluebunny]

pins 3 and 6 are changing by themselves

IC6 is the LFO

[Rob Strand]

IC1 pins are reading
6 = 5.15

IC2 thru to IC5 pins are reading
6 = 5.15 to 5.17

What you will find is the zener voltage might vary (say 4V to 5.1V) and that shifts many of the other voltages up or down.   The pin 6 voltages on all but IC6 are indicators of the zener voltage and the fact the opamps are probably working.   So the 5.15V vs ggg 4.4V (or whatever) is simply caused by variations in the zener voltage.   That should not cause the device to misbehave.

It is normal that IC6 doesn't follow the pin6 of the others since it is operating in a completely different way.  The pin 6 voltage should swing say roughly 1V to 8V.   Off-hand 1.85V looks a bit high but it's not unreasonable.

3 = changes between 3.92 to 5.38 constantly

So when I bashed some numbers on my calculator those voltages look a little lower than expected but again not-unreasonable and perhaps your meter is loading the voltage a bit.  The span 5.38-3.98 looks reasonable.

Perhaps you should check the range of voltages on your trimpot.   

Also can you find a setting on the trimpot that sounds good at all?
If you can find such a setting but it sounds weak it could be a JFET matching issue.

[bettsaj]

I'll check the trim pot at both extremes and post the voltages, I'll rig it up again and report what I'm getting (which isn't much in the way of phasing) but i'll still triple check.

I'll do all this in an hour of so as i have other things need to be getting with....

[duck_arse]

the ggg circuit I looked at, and the board I looked at, were layed out for 2N5952 jfets. if you are using 2N5457, from any supplier, you would fit them reverse to the board silkscreen, as revealed by the datasheets.

might we see photos of your built board?

[bettsaj]

I thought i had checked the pinout of the Jfets!! i'm sure they're correct, looking at the end with the pins with the flat uppermost they should be D S G. So looking at the photo reading the pins from left to right they are G S D.............. I think you're correct, they're back to front

[bettsaj]

Swapped them around.... Sorted!

I'm an idiot.

[GibsonGM]

Swapped them around.... Sorted!

I'm an idiot.

Actually, less an idiot than many of us   As many times as we see the same build error - we'll still make the same one ourselves, ha ha!!! 

[duck_arse]

Swapped them around.... Sorted!

I'm an idiot.

easy. I like these ones. and as someone once said:

datasheets.

[bettsaj]

This phaser doesn't sound like I expected it to.... It's not a smooth phase at all. I'll try and upload a video of it so you can hear what I mean. When you set the trimmer so that you get a "full phase" sound there is an audible noise at the top of the phase.. It's hard to explain, but when compared to a real Phase 90 it's totally different. It's like there's a bump at the top of the phase.

When I set the phase to the slowest on the speed, instead of being smooth and flowing, it's quite jarring to listen to.

As I said, I'll try and upload a video or audio clip tonight so you can hear it.

[bettsaj]

Video as promised

[Rob Strand]

Yes, there's something wrong with the LFO.  The duty cycle isn't 50%.  In the sound, there's a slow rise and the sharp fall.

Perhaps measure the voltage across the zener.  I think the original was around 4.8V. 
OK I see you already have measurements above.   Are you sure the 9V rail and the zener voltage are still those values?

Beyond that check the parts around the LFO and the polarity of the polarized cap.

BTW, that video is very helpful.

[Rob Strand]

I checked your photo against the info on the ggg site and I couldn't see any part problems.
I think your 15uF is around the right way but I can't quite see the markings.

The issue comes when we get to IC6
6 =  Changes from 1.85 to 8.61 constantly

The wonky LFO could be related to this.  When I listen to the clip the duty cycle extremely skewed.  Perhaps even more than those voltages alone would imply.

Check the soldering on pin 4 of IC6.

As a diagnostic check you could try temporarily soldering a 2k2 to 4k7 resistor from pin 6 of IC6 to ground.
See if that improves the situation by listening to the unit.  Then perhaps measure that pin 6 voltage again to see what it is.

[bettsaj]

Ok.... When I get home tonight i'll remeasure all the voltages as before and post what they are now.

I'll also re-check the soldering on all the IC pins and re-flow as necessary. I'll also recheck the orientation of the polarised tant capacitors

[bettsaj]

Ok, here are the latest voltage readings.

IC1 pins are reading

1 = 0.002
2 = 5.15 (According to GGG this should be less than 0.01v)
3 = 3.52
4 = 0
5 = 0.002
6 = 5.15
7 = 9.25
8 = 0

IC2 thru to IC5 pins are reading

1 = 0.002
2 = 5.15 to 5.16
3 = 5.07 to 5.15
4 = 0
5 = 0.002
6 = 5.09 to 5.16
7 = 9.25
8 = 0

IC6.

1 = 0
2 = Changes between 4.58 to 5.82 constantly
3 = changes between 4.80 to 5.25 constantly
4 = 0
5 = 0
6 =  Changes from 1.89 to 8.20 constantly
7 = 9.25
8 = 0


I've rechecked the soldering  on IC6, especially pin 4 and all seems OK..... I'll fire up the soldering iron though and re-flow all the pins just in case there's a dry solder joint I can't see.

[bettsaj]

Tant caps orientated correctly, and as far as I can see all the pins on IC6 are soldered OK..... I'm at a loss.

I've actually bought another PCB, a Fuzzdog PCB for his Faze90 clone.... I started soldering resistors onto it tonight. if I can't get this Phase 90 clone working i can just swap out the PCB's. I'll at least then have one working effect pedal, and I can tinker around on this one that's not working right.

[Rob Strand]

Tant caps orientated correctly, and as far as I can see all the pins on IC6 are soldered OK..... I'm at a loss.

Yes, I couldn't see anything wrong.   Like I mentioned before the swing on the opamp isn't enough to cause the problem.

However, I analysed your audio file, and it is fairly obvious now that your JFET's VP (or Vgs_off) parameter is on the low side and the LFO is driving them too hard.    You have adjusted the trimpot by ear to get the best result and so it sounds *nearly* right, *but*, there's no way to get it 100% right by just using the trimpot.   

What you need to do is reduce the value of R17 (1M).   I'd try soldering another 1M on the back of the board across R17.  Keep in mind that's only a guess.   Too small a value will make the phasing weak too large will get the problem you are seeing now.  When you change R17 you *must* tweak the trimpot for best sound.   If you want you could put 1M trimpot in place of R17, adjust until you get it right then use fixed resistor later with the same value.

Since you bought the JFETs you won't have any VP (Vgs_off) measurements.  If you did measure them then we would have a good idea what to set R17 to suit your JFET's VP (Vgs_off).

What people don't realize is there is only a narrow range of VP (Vgs_off) values that actually work like the real units.   If the VP (Vgs_off) is quite different from the normal 2N5952 values then it will sound different.  In many cases it's different to the original but doesn't sound bad after tweaking the trimpot.  In extreme cases like yours there is a side-effect which cannot be fixed without tailoring the circuit to suit the actual JFETs.