Phase 90 perfboard help

[aefpv]

Stompboxers,

I could use some help.  I have attempted to make a Phase 90 based off of the Effects Layouts perfboard plan  (https://effectslayouts.blogspot.com/search?q=phase+90).  However, I am not getting any sound.  My voltages for the IC and Qs are off.  I have looked for solder bridges, especially between the gate and source of Q2, 3 and 4, but I cannot find one.  Also, with my voltages, I query if pin 6 and pin7 connections are switched (see schematic: http://www.generalguitargadgets.com/pdf/ggg_p90_sc.pdf).

Before I attempt to switch the connections for pin 6 and 7 for the ICs, I wanted to see what you all thought.  Below is a picture of the build and the values for the voltages.  To note, I did not have a 15 uV cap for C7, so I used a parallel 10 uV and 4.7 uV.  Also, Q5 is a 2N3096 and Q2, 3, and 4 are 2N5457.  I did not "match" the transistors.  I actually do not know how to do that (may need a little more education).

Thanks for your time and expertise!
AE

Q2, 3, 4
Gate:  4.18
Source: 4.18
Drain: 4.08

Q5
Collector: 2.5
Base: 3.58
Emitter: 4.17

IC 1
1: 0
2: 4.16
3: 2.79
4: 0
5: 0
6: 9
7: 4.16

IC2, 3, 4, and 5
1: 0
2: 4.18
3: 4.09
4: 0
5: 0
6: 9.17
7: 4.20
8: 0

IC 6 (voltages do vary)
1: 0
2: 5.3
3: 3.13
4: 0
5: 0
6: 9.17
7: 1.29
8: 0

[r080]

I looked through your pictures just enough to confirm you do indeed have power going to pin 6. I did not verify whether you just have pins swapped, or some other larger problem.

It might help to create a reversed version of the layout to line up everything on the back.

[aefpv]

Thanks for verifying.  I did switch the connections of the pins (6 and 7), but did not have any luck with audio.  Also, my transistors continued to demonstrate the voltages noted above.  To note, when I look at the PCB from GGG, the power seems to be routed to pin 6 and not 7, so I am not sure which document is the correct wiring. 

I did use my audio probe.  I traced backwards from the output.  Audio is lost after R27, 150 K resistor.  I replaced the resistor, but no luck. 

Questions:

1.  Since I did not match the FETs, is that the issue?
2.  The voltages for the FETs suggest a solder bridge or could other issues cause the gates to be too high in voltage?

Lastly, Effects Layout's website (via link below) does have a reverse layout.

Other thoughts?

Thanks again,
AE

[r080]

Now I see the reversed layout, and I see that I am completely confused about which pin is the power pin looking at the back of your board.

The voltage on your gates should be varying with the LFO. I agree, you could have a solder bridge from one of your sources to the LFO output.

Without matched FETs, you should still get audio. If I understand right, you are not getting any dry signal through, as well as phase shifted. Your Q5 voltages seem okay.

[duck_arse]

Stompboxers,

I could use some help......
Also, Q5 is a 2N3096 and Q2, 3, and 4 are 2N5457. .....

is this still true? do you have a datasheet for your 2N5457's, and one for the 2N5952 as specified on the layout?

also, supply pin for a single opamp like the LM741 is pin 7. don't be fooling with supply on pin 6, that's the output.

[aefpv]

I used the 2N5457s based on the recommendation from the layout creator. I do not have the data sheets.

I attempted to switch the pins 6&7 connections again, but no avail. The voltages also do not meet the desired values. Pin 7 would measure near 7 and Pin 6 would measure a little over 6.

I think I am going to recreate the layout on a breadboard and see if I can figure it out better.

[duck_arse]

http://pdf.datasheetcatalog.com/datasheet/fairchild/2N5457.pdf
http://pdf.datasheetcatalog.com/datasheet/fairchild/2N5952.pdf
http://pdf.datasheetcatalog.com/datasheet/fairchild/LM741.pdf

the layout creator will have had the datasheets, otherwise would not have known they can substitute. note carefully the pinouts, and compare with the markings on the tagboards layout.

a single opamp, as shown in the datasheet, has supply to pin 7 and ground to pin 4. there is no reason or use messing about with trying different connections. get the datasheet and compare the connections on the layout and the layout and the all important circuit diagram, which has to match the layout, and they should all match.

[Mark Hammer]

One can usually obtain datasheets for any semiconductor by entering the manufacturuer name, part number, and "pdf".  Some suggested links can lead one down a garden path, but others will take you directly to the datasheet.

Alternatively, one can simply go directly to a picture of the pinout.  For instance, entering "Fairchild  2N5457 pinout" into an image search brought me to this pic.

[aefpv]

Friends,

I got it to work!  There were several (bonehead) errors that I have unearthed.  Duck_arse and Mark Hammer's suggestions to review the datasheets was helpful for me to find an orientation error on the FETs.  Correcting that issue fixed the voltage issues on the FETs.  I also had made an incorrect connection with my off board wiring.  For all that responded, thanks for the help.

I have a few more questions.  I like a slow, throbbing phaser.  The phasing does not have the depth that I really wanted.  The speed knob is also very sensitive.  If one turns it a hair too much, the speed will greatly ramp up. 

Would matched FETs will I get a better throb and control with the speed knob or are there other ways to influence the depth and speed control?

Thanks again. 
AE

[PRR]

> I like a slow, throbbing phaser.
> The phasing does not have the depth that I really wanted.

Increase C5 to maybe 50uFd.

Increase R17 to 1.5Meg or 2.2Meg.

[aefpv]

Your suggestion is very tempting!  I added the feedback resistor (R15 on the GGG schematic) and it did enhance the throb. However, going further sounds even better. 

[ElectricDruid]

Did you manage to find a 500K reverse-log pot?

Using a linear pot might make the speed control over-sensitive.

[Mark Hammer]

All FETs will have a limited range of gate currents/voltages that will result in corresponding changes to drain-source resistance.  The ones we tend to use for phasers have a range of possible drain-source resistance that results in wide-enough "movement" of where the phase shift occurs and where the resulting notches are located.  If a given FET only allowed for a very modest change in that drain-source resistance, it would not provide for a very inspiring phaser. 

Using the GGG schematic you linked to as the reference drawing, note that what comes via R17 and R18 sums together to drive all the FET gates.  The bias trimmer sets the basic voltage fed to the gates and range where sweep will happen, R17 limits the current at that voltage, and the LFO adds on to that, with its current limited by R18.  Reducing the value of R18 (and don't go much lower than 3M) will increase the width of the sweep.  That is the contrast between the highest and lowest points in the sweep.  Of course, how wide that sweep can be will be determined by the trimmer setting.  In other words, there is a "window" of possible sweep, within which the settings of the bias trimmer and LFO (in conjunction with the properties of the FETs) result in systematic changes in drain-source resistance.

I build my P90s with a variable "offset", by replacing R17 with a 680K fixed resistor in series with a 500k variable resistance (pot), such that current-limiting resistances from 680k to 1180k are obtainable.  This moves the "range" of the sweep around from low and gurgly to higher and swirly.  Similarly, I will include a toggle to select between several sweep widths, by varying the resistance of R18.  For instance, making R18=4M7, and toggling in parallel resistances of 22M and 18M, gets you effective R18 resistances of 4M7, just under 3M9 and just under 3M2, which will provide narrow, stock and wide sweeps.

The feedback resistance of R15 can be reduced a bit more to yield more feedback and a more pronounced phaser sound.  This also makes slow and wide sweeps sound more powerful.  HOWEVER, R15 should not be reduced beyond 15k or so, since it risks oscillation.  Remember that all of those phase-shift stages are supposed to be unity gain, but due to resistor tolerances may introduce very small amounts of gain.  When all of that is multiplied via the feedback, it can produce some ugly artifacts.  One of those artifacts is also noise.  In anticipation of that, many manufacturers include a feedback capacitor in one of the phase-shift stages to roll off any highs  that are recirculated and potentially amplified.  In view of that, a 2200-2700pf cap in parallel with R14 will keep any high frequency content from getting out of control.

Putting all of this together, more "throb" will come from having a little more feedback, a little more sweep width, and a suitable bias setting.  More conventional control of speed WILL come with a 500K reverse-log pot, but one can always use a regular log (A-taper) pot and wire such that clockwise=slower.  This will distribute the change-of-speed across rotation i the same way as a reverse-log, except in the opposite direction.

[aefpv]

Wow!  These are great suggestions that I will look forward to implement. I also appreciate the explanation. Thanks again!