Phase 45 clone troubleshooting

[cnspedalbuilder]

@RTK:

is everything that is supposed to be at ground really at ground? With power off, check for continuity from case to battery negative, wiggle the wire to make sure. Then move on to check around for things that should be at ground, especially the ground pins in the ICs. Another thing is check that the IC pins have a good connection through the board. I think this may seem very basic and you probably already did most of this. But sometimes it's easy to overthink a problem. gl

- At my level, I'm not sure I'm capable of overthinking it
- ground wire for battery cable is currently disconnected (it broke during other tests, but feedback was there before anyway) but ground points at jack, switch, PCB terminals, and pin 4 of both ICs show continuity with the enclosure. Now when I checked voltages I found a small, oscillating voltage at pin 4 at IC1a and 2a. About .6 to 1v but up to 3.0 at times. 
-Good idea with IC pin continuity. To check I ran continuity from pin on IC to socket solder joint. All of them checked out.

@duck_arse:

also, as RtQ suggests - get a resistor leg cut-off or something similar, and poke into each pin of your IC sockets in turn. it sometimes happens that socket pins don't actually contact the IC legs.

-sorry to be dense, but when I poke the IC socket what should I look for? Continuity? Would the test I described above address this?

[robthequiet]

My curiosity about the sound changing when you wiggle wires and adjust the trimpot etc. is to make sure that the components are physically OK. The fact that you hear some guitar signal is actually a good thing. Without hearing what's going on, my next question is: If you connect guitar and amp (with volume not too loud) and apply power, does the sound change if you adjust the speed pot?

And if you set the device down and take a toothpick or some such non-metallic object and gently tap on various components to see if something might be loose, that might give us a clue.

I think you already checked the legs of the jfets to make sure they're in the right hole... 

Hm, and if everything goes dead if you put a meter lead on pin 3 of one of the ICs, is it shorting against pin 4 and taking evrything to ground?

Back to overthinking things, if the LFO circuit has a kink in it somewhere it might be acting not as a low freq oscillator but as a full blown audio freq oscillator, and the signal somehow radiating into the circuit. So I would make sure that everything about pins 5,6,7 & 8 of IC2B is happy, along with what is happening at the base of Q1 and Q2. Carefully, though, when the circuit is alive. If I'm reading the schematic correctly, you should get a very slow change in voltage at the base of each jfet that would be readable with a voltmeter on DC voltage. I would take the readings not from the actual transistor leg but from where it connects to another component, i.e. at the junctions of R17/C5 for Q1, and R18/C6 for Q2.

Please let us know what you find, and kudos for sticking with it!

[duck_arse]

the leg pokings - the idea is, sometimes the IC pin sits in the socket pin, but the socket has gorn loose, or is carp to start, and there is no actual electric connection being made. so, you get a skinny wire, and poke it into that gap you can see between the round of the socket and the flat of the IC pins, and see if the overall operation changes. if it does ......


.... and so you know you need to check the part # of the fets and their datasheets for pinouts, rather than just plug them in and solder, yes? yes.

[cnspedalbuilder]

Just wanted to let you know I'm still on the case, but have been working my day job and have not had time to run tests suggested by RTQ and DA. Will get back on this when I can play hookey at lunch

[cnspedalbuilder]

OK, so this AM I resoldered some joints in which there was a ball of solder on the top of the board. AT this point I think everything has been resoldered and some maybe a few times. the solder joints now all look good (i think) but the PCB is not pretty my friends

My curiosity about the sound changing when you wiggle wires and adjust the trimpot etc. is to make sure that the components are physically OK. The fact that you hear some guitar signal is actually a good thing. Without hearing what's going on, my next question is: If you connect guitar and amp (with volume not too loud) and apply power, does the sound change if you adjust the speed pot?

Answer: I re-tested after all the resoldering, and this time, I could not hear the guitar. But it's possible that I just couldn't hear it over the feedback. Sadly, there is still feedback and The gain is still absolutely ridiculous--at volume=5%, it cranks at speaker-blowing volumes, so that makes extensive amp testing difficult. Tt appears that feedback tone changes depending on input/output. If I only connect amp cable either to input line or output line feedback is low-pitched (yes I get feedback from both in and out jacks). If I connect instrument line to input jack and amp to output jack, feedback is noticeably higher in pitch (but still relatively low) and it still sounds vibe/phasey and synth like. Speed pot indeed affects oscillation speed as expected. Moving around the PCB affects the feedback pitch, and sometimes it affects depth of vibe/phase oscillation but feedback is continuous (except if bottom of PCB touches and shorts to enclosure).

And if you set the device down and take a toothpick or some such non-metallic object and gently tap on various components to see if something might be loose, that might give us a clue....Hm, and if everything goes dead if you put a meter lead on pin 3 of one of the ICs, is it shorting against pin 4 and taking evrything to ground?

-I re-checked voltages and readings are:
on IC1a: 4.0 4.0 3.82 .6 IC2a: 4.0 4.0 4.0 .6
-Yes, it looks like everything goes dead when I short to ground, and that happened when lead shorted pins 3-4 
-So I did not get any effect of tapping most of the components with a pencil. There were some weird changes in feedback when tapping in zone of Q 1 but I could not replicate them consistently.

About pins 5-8, what does it mean for them to be happy? Should I be checking for continuity? There aren't any useful voltage references for IC2b, but I can certainly check voltages at IC1b. I'll also check for very slow oscillation at JFET base, though I might get some false alarms because my hand is not steady

[robthequiet]

Basically "happy" means everything connected soundly, so my test for signals or voltages at _junctions_ would possibly reveal a clue, as in does each component have connectivity across the trace. HM, so it sounds like you've built a really great oscillator. I'm going to toss out a suggestion that should be taken with a grain of salt and maybe corroborated by a couple of experts, but if it were me I would de-solder and disconnect all jacks and switches, including battery cables, clean everything up and reassemble after inspecting everything with a magnifying glass. I get the feeling that the components are mostly OK if you get some response from the controls. The change in pitch as you plug in input and output is interesting, although most circuits will radiate signals here and there which can be affected by objects moving around. My thinking is that the audio and LFO are secretely co-mingling so maybe a resistor or cap is bridging through a solder blob. Yup, the magnifying glass would be my preferred tool at this point, before even applying power.

[M.A.P]

If I were you I'd replace Q1, Q2 and just to be sure also IC1 and IC2.

If you're sure all the solder joints are ok and all the connections between the jacks, 3PDT switch and PCB are done right you must have a broken semiconductor device. To be sure it's one of the JFETs you could use your audio probe and check IC1 pin5 and IC2 pin 3. They're connected to the JFETs. You should have no oscillation on IC1 pin 3 when nothing is connected.

Just to be sure you could recheck the wiring of the potis, jacks, switch etc. I had issues with one pedal of mine and when I looked closer I saw that the ground connection from the jacks to the pcb was missing

[cnspedalbuilder]

@RTQ, you mentioned IC2b and base of Q1 and Q2. The layout says "D S G" for pins of FETs. novice ?: is base=S? Q1 rapidly oscillates at G from 1.2-1.5ish. S is solid at 3.99ish. I get inconsistencies on D pin some times it oscillates very slowly and others faster in 3.9 range, others it slowly veers between 3.99-4.0 the variability might be due to my unsteady hand, but worth thinking about. Connections are consistent with these pins. at Q2 it's 3.8-3.9, 3.98, 1.4-1.8 (D S G) and it looks like connections are consistent. I need to re-check actual voltages 5-8 of IC2b, because the rapidly changing numbers were giving me a headache but they looked consistent with connected nodes.

Gotta get back to day job and band practice tonight but will attend to other sggestions when I can.

[robthequiet]

Good catch... the equivalent of the base is the gate -- my bad -- so if I'm looking at the same schematic, looks like the LFO is doing it's thing as expected. The gates of Q1 & Q2 should pretty much be the same, and under the control of R26 and R27.

Looking at the fets, I might double-check orientation -- the datasheet at https://www.google.com/url?sa=t&rct=j&q=&esrc=s&source=web&cd=1&ved=0ahUKEwihj82eqJ7MAhUF9h4KHfu9DV0QFggdMAA&url=https%3A%2F%2Fwww.fairchildsemi.com%2Fdatasheets%2F2N%2F2N5952.pdf&usg=AFQjCNGmjZWDtFgAA0NAfXWjbHJozjeDdQ&sig2=FnbZyEZUxniSrwX7vEt1_A&cad=rja is confusing me a bit.

New idea, when you plug in your instrument, can you verify that the on-off switch actually disengages the bypass? If it was "on and off at the same time", it might create a feedback loop visible at both ends. Just a guess... Oh, and does the screaming stop if you activate the switch S1? And does the LED turn on and off? Sorry if you've already looked at this, and I don't mean to be snippy in any way, please feel free to disregard anything that might seem obvious or irrelevant. GL!!

[cnspedalbuilder]

@rtq: running to band practice, but very quickly, how can I verify whether: "the on-off switch actually disengages the bypass" or whether it was "on and off at the same time"?

[robthequiet]

Hey, sorry I didn't catch you earlier -- essentially, the switch will bypass the effect when you click it "off". At that point you should get nothing coming out of the box except for the guitar and the LED should turn off. If you're getting something other than guitar in both "on" and "off" there is probably a crossed wire somewhere. If the squealing is always there no matter what the switch does, that would be "supporting evidence." If I'm reading the schematic correctly, the guitar should come through if the switch is "off" even if there is no power to the box. HTH!

[cnspedalbuilder]

Hey, sorry I didn't catch you earlier -- essentially, the switch will bypass the effect when you click it "off". At that point you should get nothing coming out of the box except for the guitar and the LED should turn off. If you're getting something other than guitar in both "on" and "off" there is probably a crossed wire somewhere. If the squealing is always there no matter what the switch does, that would be "supporting evidence." If I'm reading the schematic correctly, the guitar should come through if the switch is "off" even if there is no power to the box. HTH!

@rtq: confirmed bypass works. clean guitar sound w/bypass on, even when pedal power is disconnected. Loud feedback when engaged. In earlier test I kept amp super low because even at zero volume the feedback is loud and that is why I couldn't hear guitar. When I turned it up today, I could actually hear faint fizzy guitar amidst the siren-like feedback. This is consistent with audio probe test run earlier.

If you're sure all the solder joints are ok and all the connections between the jacks, 3PDT switch and PCB are done right you must have a broken semiconductor device. To be sure it's one of the JFETs you could use your audio probe and check IC1 pin5 and IC2 pin 3. They're connected to the JFETs. You should have no oscillation on IC1 pin 3 when nothing is connected.

Just to be sure you could recheck the wiring of the potis, jacks, switch etc. I had issues with one pedal of mine and when I looked closer I saw that the ground connection from the jacks to the pcb was missing

@M.A.P. -thanks for this advice. I will re-check the connections. Speaking of 3PDT, is it possible that some failure or defect in the 3PDT switch could cause a problem like this?
-So are you saying that I should check IC1 pin5 and IC2 pin 3 when there is no line input plugged in? Or when the pedal is turned off or when there is no input or output whatsoever? 
-It is good to know I am not the only one who has done such things

[robthequiet]

Hm, I might try swapping the opamp ICs between sockets to see if it makes any difference. I think you already tried this, though. If one chip was blown I would expect a change. I'm leaning towards what M.A.P is suggesting and looking at possible faulty FETs. Or revisit the previous quote,

"The kit comes with a great sounding set of matched 2N5485 transistors.
The original units had 2N5952 JFET transistors. Note that the pin out of these
transistors may very. The 2N5952 will probably have to installed backwards from
what the layout shows.

Depending on who made the FETs and which part they are, you might have a reversed gate and drain which would definitely cause misbehavior. These instructions do not inspire confidence. I think they meant that the 5485 would have to be installed backwards from the 5952, so don't necessarily go with the layout. 

[M.A.P]

@M.A.P. -thanks for this advice. I will re-check the connections. Speaking of 3PDT, is it possible that some failure or defect in the 3PDT switch could cause a problem like this?
-So are you saying that I should check IC1 pin5 and IC2 pin 3 when there is no line input plugged in? Or when the pedal is turned off or when there is no input or output whatsoever? 
-It is good to know I am not the only one who has done such things

Exactly. Check the IC pins when there is no input signal, no output connected and power on. Maybe you have to solder a link at the input jack for the connection of the battery with the circuit ground. The FETs are driven by the LFO in the right corner of the schematic. You could also check IC1 pin 3 and IC2 pin 7. Or just to be sure you could check all the pins with the audio probe and write down on which one an oscillation can be heared.

I don't think that a broken switch would create an oscillation. But never say never

[cnspedalbuilder]

Exactly. Check the IC pins when there is no input signal, no output connected and power on. Maybe you have to solder a link at the input jack for the connection of the battery with the circuit ground. The FETs are driven by the LFO in the right corner of the schematic. You could also check IC1 pin 3 and IC2 pin 7. Or just to be sure you could check all the pins with the audio probe and write down on which one an oscillation can be heard.

Done. So, with DC in, pedal powers up w/o input. There was no oscillation with the probe, just increased static. I checked again with an input, output, or both, and feedback comes back. Heres the weird part: If I have both inputs and then pull 'em out, running the audio probe on IC1a pin 3 still elicits feedback for as long as I want. But then if I touch one of the other pins, the feedback goes away and touching pin 3 again just gets static from the audio probe. In other words, there seems to be an after effect on IC1 pin 3, even after instrument cables are removed, but touching probe elsewhere makes the after effect go away.

Also, I had the battery ground hooked into the input stereo jack, as you suggested. There was no battery for these tests, though, so I'm not sure it would've made a difference anyway.

[cnspedalbuilder]

@RTQ, ok I can try switching orientation of the JFETs. I'm a little worried that resoldering the joint could damage them, but you think it's ok?
It was also suggested to replace the JFETs and ICs. I'd have to order them, which I can do. I've got a transistor testor coming in a few days, so thinking I'll test w/that before ordering new ones.

[duck_arse]

don't blindly reverse polarised parts. get the part # and manufacturers name/logo from the part, look the interwires for the datasheet. then compare the pinout shown on the datasheet to the board layout and the circuit diagram. if you don't know what it is you're looking at on the schematic, or don't understand the pinout diagram, just ask.

[cnspedalbuilder]

don't blindly reverse polarised parts. get the part # and manufacturers name/logo from the part, look the interwires for the datasheet. then compare the pinout shown on the datasheet to the board layout and the circuit diagram. if you don't know what it is you're looking at on the schematic, or don't understand the pinout diagram, just ask.

Got it, that makes sense. I probably will need help w/the translation between the datasheet and the layout on the board, so hope you don't mind more questions

[robthequiet]

Yes, you may just have it right with the FET's, but the two part numbers in question hava a diffeent pinout. The datasheets have pictures. As a consumate nitpicker, I always triple-check these things. If, hat tip to duck_arse, you determine that you need to, de-soldering the FET's should not harm them as long as you're using a heat sink on the legs (i.e. an alligator clip or such, I use small forceps) and small needle-nose pliers. Definitely feel free to ask questions any time.

The nature of the sound changing (aftereffect) is fairly normal. By the time we're done this is going to be one tired but strong little box.

[cnspedalbuilder]

@d_a+@RtQ:
So I found the datasheet: https://www.fairchildsemi.com/datasheets/2N/2N5485.pdf
I believe that the orientation corresponds to the GGG layout and the transistor's orientation in the PCB. If viewed from the back/flat side (front?) first, G wire is on right hand side right wire, S is middle, and D is left.

Is this all I should check out? Is there other useful info in the datasheet that I need?