Dan-Echo (PT2395) may be dead, or just a switching problem?

[robmay]

Hi all, so I'm working on recasing my beloved Danelectro Dan-Echo pedal (Schematic here: https://www.dropbox.com/s/mnp42vknn9wb61a/Dan%20Echo%20Schematic.pdf?dl=0)right now so that I can make a few things more accessible (footswitch for hi/lo switch, footswitch for infinite repeats, and a couple of mods). So far all I've done is wire new pots off the PCB, take the high/low switch off the PCB, and replace both with better components.

I think I may have killed it somehow though, either that or it's not switching correctly. The signal is currently not changing at all when I hit the switch (aside from the LED coming on), so the signal going in is exactly what's coming out, regardless of the position of any of the four pots. Using a probe, I'm hearing input at the 4558, as well as at the PT2395. I'm reading a proper +5V DC voltage everywhere including the 4558, 2395, CD4013, and 2.5V DC at the couple of pin outputs where it should be. Everything appears fine.

Alas, with proper voltages and inputs, no output awaits. I do hear the expected tone change when probing pin 2 of the hi cut knob, yet I hear no delayed audio anywhere. 

I'm thinking it's the 2395 that's fried somehow, since it's getting the right voltages and I'm getting no delayed output. My main question (that I hope will fix this or show me what's really wrong) is how the 4013 switching is working in this circuit. From looking at it, my guess is that it energizes the transistor that's either grounding out the delayed audio coming from the 2395 or letting it pass. That's my guess, am I on the right track? Because if so, I know the 2395 is dead since it's not outputting anything. If not, (what I'm kind of hoping) is that the 4013 is affecting the powering or something about the 2395 and it's the switching that's broken.

Related, I would like to completely remove the switching system and make it true bypass (if there was ever a tone-sucking pedal...). I've seen posts of people just jumpering the 4013 so that it's always active, but I'd like to simplify and take things out. How would I remove that part of the circuit, and where would I wire up the bypass switch? It does seem a little complicated just because it seems that the input passes through as the dry signal normally.

Thanks in advance!

[marmora]

Were the pots grounded? Maybe you need to reconnect those ground points.
Hope that helps, let us know how it goes.

[robmay]

Thanks for your response! The pots were grounded to the PCB, and those points connect to other grounds. I checked and all ground points that I can find are properly connected.

I am noticing something that may be odd, but may not. I get a +2.5V DC reading at pins 36, 37, and 38 of the PT2395. Has anyone worked with this chip before that could give me some advice?

I'm tempted to just replace the 2395, a spare would only cost ~$6. The thing holding me back is how annoying it would be to replace a 40-pin IC...

Any ideas? Thanks all!

[robmay]

SOLVED! 

Alright, so here goes. The PT2395 is fine. Everything is fine on the board, voltages are fine, all confirmed except the switching system, which I wanted to get rid of in the first place.

If you too would love to make your similar Danelectro pedal true bypass, here's how I cut out the switching system altogether.

Apologies if some of this information seems basic, I'm hoping it will help some people like me, and I'll keep notifications on in case others have questions.

Find whatever point on the schematic has the "MIX" knob, and near it you should find a transistor linking the MIX control to some components that comprise the CD4013 switching system. If you pull the transistor out (or just snip the legs, if you're into that) it will kill the switching as well as the LED. Wire the LED up to your 3PDT (or Millennium Bypass) switch, and you're good to go.

Note: this will change how your mix knob works! If you are looking for a more wet mix than what's provided, or some amount of killing the dry signal, you'll have to play with the opamp summing stage. What I did was add a 200k trimpot across R25, then adjusted it until I was happy with the mix control. You can change more about this if you want, but essentially R25+mix knob and R22 create the inputs to the 4558 to be summed, and varying the resistance will change how it adds the signals and which it favors in the mix. For more info about how opamp summing amps work, look here: http://www.electronics-tutorials.ws/opamp/opamp_4.html

I know it is possible to hotwire the CD4013 and avoid having to do all that, but If you're going to hotwire it anyway, why not take it out and get a better understanding of the basic circuit?

Danelectro Dan-echo schematic here for reference:
https://www.dropbox.com/s/mnp42vknn9wb61a/Dan%20Echo%20Schematic.pdf?dl=0

[anotherjim]

How the 4013 switching works?
4013 is a D-type flip flop. This moves the logic level on the D input to the Q output when the Clock (Clk) input is pulsed. This one is wired as a clock divider - or "alternator" (my term) if you will. The clue to this is that /Q (Bar-Q or Not-Q) output is wired to the D input. /Q output is the opposite state to whatever the Q output is.
Assume Clk =0 and D = 0 ;then /Q is 0 and Q is 1.
When Clk changes to 1 (switch connects to +5v), Q changes to 0, because D was 0. Then /Q becomes 1 and so does D.
When Clk changes to 0 (switch open), nothing happens. 4013 Clk ignores changes from 1 to 0 - It is "positive edge triggered" which is 0 to 1.
Now, since D last changed to 1, the next time Clk changes 0 to 1, Q will become 1 and /Q0 and D also 0 - it's changed back. It will flip or flop every time the switch changes state, always the opposite of the previous state. In other words, it alternates.

Q2 is the thing that actually kills the signal. When /Q = 1, the transistor is switched on causing the signal feeding the mix and repeat pots to be grounded out - silenced. I'm not sure if that transistor has been correctly identified, it may be a J-Fet (2sc1815 is BJT), but if it is BJT, then forward biasing the base makes the transistor switch on.

So that killing the Delay input does not also kill existing repeats, I would think I'd disconnect Repeat pot 1 from R9-C23 and use an extra 10k resistor to connect the pot to R9-C8 instead.