Echo base delay switching problem, cannot disengage effect

[Soebam]

Hey there!

I have known this site for a while, but I decided to register an account because I'm lost debugging an Echo Base delay, and I really hope you can help me. Until now I've built and debugged much simpler circuits, and using an audio probe to follow the signal path could usually identify the problem. This time however, the effect stays engaged, cannot be switched off, and the switching is done by a 4066 chip which I'm not familiar with. Also, the effected signal is heavily distored with pops and cracks entering the delay. If anyone is interested I could make a recording and upload it to soundcloud.

The effect itself seems to work just fine, all the controls behave like they should and it puts out some fun delay sounds and self-oscillating feedback, yay! Also, with the volume pot all the way left, the clean guitar signal sounds fine. I think I have thoroughly triple checked the offboard wiring, component placement, bridges, trace cuts, solder joints, cleaned in between traces with a knife and all that, and although I might overlook something, I simply cannot find a faulty connection.

So here I need your help . I've followed the forum guide to categorically describe my problem, I hope I did well. Thanks for reading!

1. Effect: Echo Base delay
2. Built on veroboard: http://www.sabrotone.com/wp-content/uploads/2011/05/EchoBase.gif
3. Pics: http://imgur.com/a/D1wYB
4. I followed the layout exactly, with two exceptions:
    - R6: A 220k and 20k resistor in series instead of 240k (didn't have one)
    - D1: 1N4007 instead of 1N4001. As I understand, I just used a 'beefier' version, capable of handling more load but less efficiently, but it should work fine. I do mean to replace it for the 4001 as soon as I have my new diodes in, though.. Could this be the problem?..
5. Negative ground effect
6.

- The problem: it doesn't bypass, and the effected sound is distorted, like there are multiple cracks/pops included in the delay. These pops, as well as guitar input, can be controlled with the pots as expected: delay volume, time, feedback, and modulation rate and depth all work as expected. However the bypass does not work, the effect is always engaged. The bypass switch itself is fine (measured resistances), and manually connecting the bypass SW wire to ground doesn't work.

- The switching is done by a 4066 chip (http://upload.wikimedia.org/wikipedia/commons/thumb/b/b1/4066_Pinout.svg/2000px-4066_Pinout.svg.png). The switching diagram is straightforward, but I must admit the way it's connected in the layout confuses me. Switching happens when the Bypass SW 1 wire is connected/disconnected from ground. This wire leads to the base of Q2, and through a 220k resistor to the collector (why is this?). The collector receives 9v through a 10k resistor (R3), and I can measure 8.9V at its collector. The emitter is then connected to pin 10 of the 4066, where I suppose the switching starts. Connecting/disconnecting the 'Bypass SW 1 wire' to/from ground does not change the voltage on pin 10, which is always 0. In fact, removing Q2 doesn't change the behaviour of the effect (still always on), so I suspect the problem to be 'further in' the circuit, and beyond my limited comprehension..


Some extra info:

- With the volume rolled off, the dry guitar signal is clean and crisp, so the buffer/in out works fine (also, the dry switch kills the signal)

- I'm using a 9V DC pedal power supply (powerplant junior from thomann.de). I can measure 9.02V on the effect input (effect engaged (that's the problem!)), and it powers other pedals just fine.

- The LED indicator oscillates with the mod rate control. This seems a bit odd, I'd expected it to oscillate with the delay time, but since it's connected through a switch in the 4066 and R4 to the modulation part in lane b, I suppose it works like it should..


Now, all the voltage measurements:
+9 in: 9.02
Chassis, in/output tips: 0.00

Q1 (78L05 5V regulator):
In: 9.02
Gnd: 0.00
Out: 5.05

Q2 (2N50089):
E: 0.00
B: 0.00
C: 8.94

Q3:
E: 0.00
B: 2.01
C: 2.52

IC1 (TL072):
1: 5.05 (with modulation engaged, oscillates 4-6)
2: 5.05 (with modulation engaged 3.82)
3: 5.05
4: 0.78 (with modulation engaged, oscillates 0.65-0.80)
5: 2.81 (with modulation engaged, 3.87)
6: 5.05
7: 7.55 (with modulation engaged, oscillates 2.2-7.5)
8: 8.16 (with modulation engaged, oscillates 8.15-8.30)

IC2 (TL072):
1: 5.05
2: 5.05
3: 5.03
4: 0.00
5: 2.51
6: 5.05
7: 5.05
8: 9.02

IC3 (4066):
1: 2.88
2: 2.88
3: 2.88
4: 2.88
5: 7.39
6: 0.80
7: 0.00
8: 2.56-2.60
9: 2.54-2.58
10: 0.00
11: 0.00
12: 0.00
13: 8.86
14: 9.02

IC4 (PT2399):
1: 5.05
2: 2.51
3: 0.00
4: 0.00
5: 2.04
6: 2.41
7: 0.02
8: 0.80
9: 2.52
10: 2.52
11: 2.52
12: 2.52
13: 2.52
14: 2.52
15: 2.52
16: 2.52

D1:
A: 0.00
C: 9.02

D2:
A: 0.00
C: 0.00

D3:
A: 0.00
C: 0.00

Thanks again for reading, I really hope I'm just being silly and someone will point out what I'm doing wrong . If more measurements are needed, please ask! Again, I can record te cracks and pops if that'd be of any help, just ask.

[slacker]

Hi

Hope you don't mind if I point you at this thread where someone was having similar problems, a lot of the debugging info in there relates to your bypass problem, have a read through that and see if it helps.

http://www.diystompboxes.com/smfforum/index.php?topic=110128.0

Welcome aboard

[slacker]

.

[canman]

I had similar problems and I believe slacker pointed you to my thread where we spent a few days working it all out.

Ultimately, my problem came down to two things:  I had the feedback switch wired up backwards, so it was constantly self-oscillating (not your problem here), and my 4066 chip wasn't making a solid connection.  I had the same voltage readings you got at pins 1-4 of the 4066, so try jamming in some cut off resistor leads into the socket and then insert the chip, so it makes a really tight connection. 

With regards to heavily distorted delays, I had to replace a couple of resistors...that thread linked above specifies which ones. 

[Soebam]

Thank you both for your responses. I've read canman's debug thread, and although I enjoyed the read and learned some more about this effect and the components, those problems are different from mine I believe. Anyway, I managed to solve the switching problem on a 1 hour train ride staring at the schematic and my own post.. Here's how:

I read my own question why the 5089's base was connected to the collector through a 220k resistor. But I reasoned backwards: 9V is carried from the +9 rail through R3 to the collector, and then through a bridge and R7 to the base and the switch wire. Connecting this switch wire to ground (through the switch) depletes the base of current and stops the transistor from emitting. Since I couldn't measure a voltage on the base on either switch setting, either the switch was somehow grounded (it wasn't, I tested before), or the voltage never reached the 5089's base. So I traced the voltage to R7, and it turned out I misplaced R7, bridging lane 'h' and 'i' instead of 'g' and 'i'. I replaced R7 and the switching works now ! These things are easily overlooked, apparently..

But... I sort of hoped that putting the 0.62 volts after R7 on the wrong rail caused the noisy clicking problem, but this didn't go away so something else is wrong, still. I realize I didn't describe the noise adequately, it's not like a distortion, it really feels connected to the delay clock. Like there is a repeated sequence of loud 'pops' at every delay cycle, the pops are somewhat irregular but the repeats are regular and matching with the delay time. Both modulation pots alter the pop sequence a bit, too, even when the mod switch disengages the modulation. The pops are super loud and oscillate quickly. Only setting the feedback to 0-10% or so will eliminate this pop oscillation and get the delay quiet. In the meantime, the signal comes through in the background and behaves perfectly like it should. Also, the guitar signal doesn't self-oscillate at low feedback levels, and the level pot can put the delay signal higher than unity gain, so I feel that all of this is in order. It's just that the guitar signal cannot be heard because all those pops are ridiculously loud and feedbacking..

I'm going to make a recording since I feel that will be much clearer than my attempted description. Will post back when I have it on soundcloud. In the meantime, anyone with an idea? Thanks again!

[Soebam]

Here's a sound demo: https://soundcloud.com/soebam/echo-base-noises

The waveform sometimes looks like this:


It looks like it is skewed a bit, as in the wave baseline is not centered at 0. Could I have some DC in my signal? Or am I talking nonsense?
Or is this the kind of distortion you had, too, canman?

[canman]

I know this is probably something you've already tried, but...have you tried a different PT2399?  I hear they are pretty temperamental.

[Soebam]

If I had one I'd try. Sadly I only bought a single chip. If all else fails I'll order another one and give it a try.

[slacker]

I've never heard noise like that before. I would double check the values of all the caps attached to the PT2399 and that the connections are good especially the 100n caps.

[Soebam]

Ok slacker you are officially a wizard. I had C9 bridging Q22 to T22, instead of R22 to T22. Shame on me for two misplaced components in one circuit. It works now! I'm really impressed, wonderful pedal, great sound, overall really cool! So more props to you for that, and thank you very much for helping me out.

[slacker]

Excellent, glad you got it fixed.