Phase 100 GGG build - still not working!

[axg20202]

I built this board ages ago (using a board bought from GGG) and it didn't work properly. Seems like lots of people have trouble getting this board to work. After numerous attempts to bug fix it gathered dust for a few months. Well, I thought I'd have another go at fixing it over the weekend. It seems that the board has been revised earlier this year (Jan 2007), so I decided to etch the new board myself and repopulate it. Result: the problem is exactly the same with the repeat build even after wiring using toggle switches instead of the rotary switch to keep things simpler. There is no phasing at all on any setting, but one of the toggle switches causes the pedal to produce a high pitched squeel that can be pitch-adjusted using the speed pot. Bizarre. Has anyone else had this experience? Does someone know a good checklist for debugging this particular pedal? For example, how do I check that the LED units are working properly?

I know bug fixes are difficult to help with, but if aynone can help I'd really appreciate it.

Thanks,

Andy.

[Processaurus]

the squeal is suspicious, check all the components in the LFO section to see they're the right value.  To check the LFO you can hook a meter up to ground and one of the LEDs, and on the slowest speed setting you should be able to see a voltage going up and down as the LEDs get turned on and off.  Or sub in a regular LED and see if it pulses.

[R.G.]

At the risk of repeating myself...

Read the thread stuck to the top of the forum: "Debugging Thread: What to do when it doesn't work" and supply that info.

[axg20202]

Yep, will do. Just thought I'd see if there was anyone who had experienced the same before doing so.

Andy.

[axg20202]

Hi,

well, I've made some probes of the circuit. I can immediately tell that it isn't right. To start with, the LED units are not showing any oscillation in voltage as one would expect. If someone can find some wisdom out of the following I'd really appreciate it.

Board description: MXR Phase 100 Replica Project  from the GGG website.

Layout: http://www.generalguitargadgets.com/pdf/ggg_mp100_lo_2.pdf (revision Jan 2007)

I have made the board as shown with toggle switches to select the various phase flavours. I have not included the battery connector or the input jack switching, only the boss style 9v socket. Everything else is as shown in the PDF link above. I know it's against 'the law' but I have performed the following tests using a 9.5v DC mains adapter as power source.

As per the layout and build instructions, the ICs (qty 6) are all TL072. Pin out for this is as follows: 1 = 1OUT; 2 = 1IN neg; 3 = 1IN pos; 4 = V neg; 5 = 2IN pos; 6 = 2IN neg; 7 = 2OUT; 8 = V pos.

All four transistors are 2N5088.

The three LED/Dual Photo Resistors units (abbrev. as OC1, OC2 and OC3 below) are all VTL5C3/2.


So, here are my readings:

Q1:
C = 0.5v
B = 0.5v
E = 0v

Q2:
C = 0.5v
B = 1.06v
E = 0v

Q3:
C = 0v
B = 0.6v
E = 0v

Q4:
C = 9.56v
B = 4.9v
E = 4.4v

IC1:   1 = 4.71v
   2 = 4.73v
   3 = 4.26v
   4 = 0v
   5 = 4.4 – 4.55v
   6 = 4.96v
   7 = 4.7 – 5v
   8 = 9.56v

IC2:   1 = 4.62v
   2 = 4.71v
   3 = 4.58v
   4 = 0v
   5 = 4.27v
   6 = 4.77v
   7 = 4.78v
   8 = 9.56v

IC3:   1 = 4.78v
   2 = 4.79v
   3 = 4.27v
   4 = 0v
   5 = 4.57v
   6 = 4.71v
   7 = 4.45v
   8 = 9.56v

IC4:   1 = 4.81v
   2 = 4.81v
   3 = 2.28v
   4 = 0v
   5 = 4.27v
   6 = 4.79v
   7 = 4.8v
   8 = 9.56v

IC5:   1 = 4.78v
   2 = 4.78v
   3 = 4.27v
   4 = 0v
   5 = 4.59v
   6 = 4.79v
   7 = 4.79v
   8 = 9.56v

IC6:   1 = 4.79v
   2 = 4.79v
   3 = 4.59v
   4 = 0v
   5 = 4.27v
   6 = 4.79v
   7 = 4.79v
   8 = 9.56v

OC1:   PIN1 = 0.5v
   PIN2(LED+) = 9.56v
   PIN3(left) = 4.26v
   PIN4(middle) = 4.68v
   PIN5(right) = 4.26v   
   
OC2:   PIN1 = 9.55v
   PIN2(LED+) = 9.55v
   PIN3(left) = 4.27v
   PIN4(middle) = 4.68v
   PIN5(right) = 4.27v

OC3:   PIN1 = 9.55v
   PIN2(LED+) = 9.55v
   PIN3(left) = 4.27v
   PIN4(middle) = 4.68v
   PIN5(right) = 4.26v


Hope someone can help me sort this mess out!

Cheers,

Andy.

[R.G.]

Q1:
C = 0.5v
B = 0.5v
E = 0v

Q2:
C = 0.5v
B = 1.06v
E = 0v

There are certainly problems here. The LEDs are strung from +9V down. At even 2V each, that's only 6V, and the collector of Q1 and Q2 should not be able to get below about 3Vdc. They're down at 0.5V, so something is wrong with the LED string.

OC1:   PIN1 = 0.5v
   PIN2(LED+) = 9.56v
   PIN3(left) = 4.26v
   PIN4(middle) = 4.68v
   PIN5(right) = 4.26v   
   
OC2:   PIN1 = 9.55v
   PIN2(LED+) = 9.55v
   PIN3(left) = 4.27v
   PIN4(middle) = 4.68v
   PIN5(right) = 4.27v

OC3:   PIN1 = 9.55v
   PIN2(LED+) = 9.55v
   PIN3(left) = 4.27v
   PIN4(middle) = 4.68v
   PIN5(right) = 4.26v

From this set of numbers, it appears that OC3 and 2 might be tied to the power supply properly, but there is probably an open between OC2's LED cathode and OC1's LED anode. OC1 LED anode/cathode may be tied to something it should not be, as well.

IC1:   1 = 4.71v
   2 = 4.73v
   3 = 4.26v
   4 = 0v
   5 = 4.4 – 4.55v
   6 = 4.96v
   7 = 4.7 – 5v
   8 = 9.56v

Your LFO is probably not oscillating. Something is wrong witn the B side of this IC.

[axg20202]

Wow, that's great! Thanks! I'll do some digging with the circuit based on your comments and report back.

Cheers,

Andy.

[axg20202]

Hi RG,

please excuse my ignorance, but can I just confirm that I understand what you are suggesting and also give some preliminary results on tests based on your comments.

1) So, the voltages I see at Q 1 and 2 are due to problems with OC1 - I should be seeing ~9v at the cathode of OC1?

2) Not sure what you mean by an "open between OC2's LED cathode and OC1's LED anode. OC1 LED anode/cathode may be tied to something it should not be, as well."  There is a trace directly linking these and it is good. If I probe the leads of the LEDs I get the same readings as I do along the trace and at the soldered nodes, unless I'm completely misunderstanding you, which is more than likely. :-)

3) "Your LFO is probably not oscillating. Something is wrong witn the B side of this IC." - could this be as a result of my odd readings at OC1 - OC3, or could this maybe the cause? i.e. should I consider that IC1 is faulty? Is there a simple way that I can confirm whether OC1-3 are good?

Thanks for your help, it's much appreciated. Sorry for my possibly stupid questions, but this is the first board I've had to debug - other circuits I've built and modded etc with no problems. I guess I have been lucky before, but they were simpler builds to be fair.

Andy.

[R.G.]

please excuse my ignorance

No problem! People have been excusing my ignorance for a long time. 

1) So, the voltages I see at Q 1 and 2 are due to problems with OC1 - I should be seeing ~9v at the cathode of OC1?
2)Not sure what you mean by an "open between OC2's LED cathode and OC1's LED anode. OC1 LED anode/cathode may be tied to something it should not be, as well."  There is a trace directly linking these and it is good. If I probe the leads of the LEDs I get the same readings as I do along the trace and at the soldered nodes, unless I'm completely misunderstanding you, which is more than likely. :-)

That's my guess, based on the voltages. You show voltages on the string of LEDs which start at 9.55V and ... stay at 9.55V until you get to pin 1 of OC1. What should be happening is that each OC LED should drop about 1.5 to 2.2V, depending on the LED inside. So if they are 1.5V LEDs and if the voltage starts at 9.5V at the anode of the LED in OC3, you should see about 8V at the cathode of OC3/anode OC2, and 6.5V at the cathode of OC2, anode of OC1, and about 5.0 at the cathode of OC1. What we see instead is that all the points are at 9.5V except the last one at 0.5V. So something is wrong in the chain. If the LED in OC1 were open, that would explain it. More likely, there's an open solder joint on OC1's LED.

3) "Your LFO is probably not oscillating. Something is wrong witn the B side of this IC." - could this be as a result of my odd readings at OC1 - OC3, or could this maybe the cause? i.e. should I consider that IC1 is faulty? Is there a simple way that I can confirm whether OC1-3 are good?

Well, the squeal indicates that it may be oscillating, but it's oscillating at way higher frequencies than it should. You should see it waver up and down on your meter. A squeal might be too fast for the meter to follow.

Check out the LED voltage thing first.

[soundcollage]

I  had a problem that took quite some time to fix using the ggg phase 100 board. There was an tiny thread of copper that was not etched that was shorting out the signal. Use your continuity tester to make sure traces are not touching that should not be.

[axg20202]

Thanks! I will explore again when I get home from work. Although the premise that the OC1 cathode is not connected to the board properly makes sense, I'm not sure how this would explain how I am able to read the same 0.5v at any point along the path from the actual cathode lead of OC1 up to Q1. Surely if the OC1 cathode lead is disconnected e.g. because of a poor solder joint, and I probe the lead to ground, I would see the 9.5v at the lead, but would only see 0.5v at the actual solder point and up the trace to Q1.  I wonder if a dead LED in OC1 could cause what I'm seeing? What dya reckon?

The fact that the same symptoms have been reproduced on two different boards with slightly different layouts (components were transferred from one to the other) could obviously mean perfectly repeated user error in assembly (better to be a consistent tw@t than a random one, I say!), but it could well mean the transferral of a faulty part(s). I think maybe I will swap OC1 and OC2 and see if the problem moves with OC1.

I'll report back - I'm sure you can't wait! 

Andy.

[R.G.]

I wonder if a dead LED in OC1 could cause what I'm seeing? What dya reckon?

If the LED in OC1 were open, that would explain it.

Yep - ah reckon.

It's just that dead new components are quite rare. It's many, many times more likely to be a solder or other construction problem. But dead components do happen... rarely.

[axg20202]

Who said anything about dead NEW components - they were probably fried during my continuous fiddling to get the sodding pedal to work!     
I've ordered replacement LED/LDR units, and transistors for good measure, and will swap them out. They're not that expensive so no big deal. I have already replaced IC1 (I have loads of TL072s kicking around) and got the same readings. I'm hopeful that the new LED units will get things on track.

Thanks for your help RG - much appreciated.

Andy.

[R.G.]

Just for grins, since you're going to remove them anyway, can you test out the "bad new component" theory?

Remove OC1 and OC2. Hook up your ohmmeter to the LDR end and get yourself a 9V battery and a 4.7K resistor. Read the ohmmeter with the LED open, should be big, maybe 100K to many megs. Then connect the 9V through the resistor to the LED. Does the resistance go down? If not, reverse the battery polarity on the LED/resistor. Now does it go down?

You should see the LDR resistance go much lower in one orientation. If the two OCs act different, then one is bad.

If OC1 is bad/new, you should see high resistance both ways.

For a further verification, hook your voltmeter across the LED part of OC1. Now apply battery+ resistor. If the LED is open, it will be battery voltage both polarities. If the LED is OK, you'll see only about 1.5-2.2V one way, and much larger, maybe full battery the other way.