Rebote 3 delay refusing to cooperate

[HeavyFog]

That's a good catch, I noticed it but didn't think anything of it before. Measuring each emitter to ground gives continuity, with a resistance of just less than a single ohm. This definitely isn't enough to cause any issues, plus on the pcb the emitters and pin 4 of each chip are all directly connected to the main ground trace. Why it's drawn this way on the schematic i have no idea, but it doesn't look like it would be the source of this issue. 

[anotherjim]

I'd be inclined to try NPN transistors (2N5088?) in reverse beta. As already suggested, this may be the original intent. I wonder if somebody didn't see a scheme with reverse NPN, thought it was an error and decided it must be PNP?

[HeavyFog]

I think the way it's implemented is probably the way it was intended to be. The echo base uses basically the same method to add modulation to the delay signal, so it seems like a pnp transistor should work the way it's drawn. There's got to be something i'm missing, i couldn't see tonepad putting out a non functional project

[anotherjim]

Yes, the PNP can work like that.
OK, is the basic problem, that if the LFO Depth is zero, then the only delays you get are stupidly long & lo-fi?

Each delay chips pin6 should be about 2.5v.
Have the Depth and Repeats controls set at minimum. For the transistor to conduct, the voltage at the wiper of the delay pot must not exceed that 2.5v minus the base-collector junction voltage (about 0.7v) and whatever voltage is dropped across R37. Turning the Delay pot CCW should increase the transistor action and reduce the delay time as more current will be drawn from PT2399 pin6.

If the transistor isn't doing anything, the 100k resistor across it is setting the delay time. The delay quality is already bad if that resistor was 50k. Try soldering a smaller resistor across the 100k, maybe 27k, and see if you get shorter delay with decent quality.
Not much point in getting the LFO & repeats working if the basic audio delay function is bad.

[HeavyFog]

The issue happens regardless of the modulation settings. Right now i've actually got it disconnected all together for debugging. The delays being long and grainy happens regardless of how the modulation is set.

Il get some voltage reading next chance i get

[HeavyFog]

Measuring the delay pot's wiper gives me 0v at minimum, and 2v at max, more than the 1.8v I would expect given what you're saying. Also the voltage across r37 (1K) is only about 0.02v, which seems a bit small to be taken into account.

I could definitely try swapping out the 100k resistors for something smaller, but I don't think this would do all that much. When I turn the delay time pot down the delay gets only slightly slower before it starts distorting and feeding back, then cuts out all together, it seems like changing out the 100k timing resistors would only minimize the narrow delay time where it doesn't feedback or distort

[anotherjim]

Delay pot voltage depends on the voltage divider formed by the pot and a fixed upper 15k resistor. It expects a 10k pot as a lower resistor to make the voltage range from the pot. Pot's have large variation in value anyway. 0-2v isn't too bad just that part of the range. That's just the start of the checks for there. Negligible volt drop across R37 could mean nothing working after it! You ought to be able to see voltage vary when turning the delay pot at each resistor feeding the base of each transistor. At the same time, the pin6 of each PT2399 should hold about 2.5v only falling slightly in the event the transistor should switch fully on.

The only thing setting the delay time of the Pt2399 is supposed to be the resistance of the path to 0v out of its pin6. This design is using a transistor as a variable resistor. A fixed resistor across the same path sets the maximum delay when the transistor is effectively switched off.
A chart in the datasheet tells us that at 1% distortion, the pin6 resistance is about 27k for about 350ms delay. It doesn't show data above that, presumably because the distortion gets too much to brag about.

[HeavyFog]

For each base resistor im getting a reading of roughly 1.6v (delay time min) to 2v (delay time max). On each pin 6 there's 2.49-2.5v, falling as low as 2.3v when delay time is set at minimum. Weirdly enough measuring resistance from pin 6 to ground gives me no reading at all. There's no continuity short to ground, but i get no reading all together. Wonder if that has something to do with it

[ElectricDruid]

Weirdly enough measuring resistance from pin 6 to ground gives me no reading at all.

Not so weird. If it's powered up, the power will screw up the multimeter, since it uses a small voltage to test the resistance. And if the power is off, the transistor will be too, and you'll just measure the 100K.

[anotherjim]

I don't think the transistor base control is moving far enough. Are you sure all the resistors around the delay time pot and transistor bases are the right values? Any misread colour codes?
As Tom just said, you can't always measure circuit resistance directly. To do it, you could insert a fixed resistor (say 1k) in series with the emitter leg to 0v then as the control varies, measure the voltages. You can calculate the resistance of the transistor as with the added resistor it forms a voltage divider between 2.5v & 0v. If the emitter measures 1.25v, the transistor is worth the same as the added resistor. To get down to a short slap-back delay, the resistance needs to be as low as 1k.

[HeavyFog]

I'm absolutely positive that the values I have in are the correct ones. I went back and checked all of that when it wouldn't work the first time, especially those associated with delay time control.

I will have a look at the resistance though to see if that gives me any clues. Otherwise I'm starting to think this might be a lost cause. Everything seems right but it refuses to work.

[HeavyFog]

Also i checked the resistance to ground from each pin 6 unpowered and i get 100k as expected. I'm thinking this issue may not have to do with the rate control like i though. There's no question that i am getting a delay of sorts, even if it is lofi and messy, and turning the time control does affect the time to a degree, i just cant turn it down much before it feeds back and then cuts out all together. I can't seem to find a reason for there to be any feedback as the delay time goes down, so maybe there's a short somewhere i'm unaware of. Even weirder why would the delay cut out altogether so suddenly as i turn the time down? I'm trying to think of a reason this could possible happen so i can work backwards from there, i just need something to work off first. Maybe a good place to start is to trace the whole pcb and check for errors since I have had this problem with one tonepad pcb before.

In the meantime il order parts for a rebote 2.5, and if i cant get this working before they get here in a few weeks il probably just admit defeat for now

[ElectricDruid]

Since the circuit is really just three equal copies of the same PT2399 part, could we remove two of the PT2399 chips and connect the outputs to the inputs so that we can temporarily turn it into a single PT2399 delay?
That removes a lot of circuitry and if it starts working, then perhaps there's a fault in the circuit around one of the other two chips. In which case, move the one chip that's working to a different socket and rinse and repeat.
If it still doesn't work, at least the problem is reduced significantly. You could pull the transistor out and replace it with a 100K pot and see if it works then (100K + 100K pot would be equivalent to the typical 50K pot).

[HeavyFog]

Maybe what ill do is disconnect the transistors and replace them with several variable resistor pots to ground. If i get the same issue as i turn down the delay time i can pretty much rule out the pnp transistors being the problem. From there i should be able to at least isolate the issue down to signal path or delay time control

[HeavyFog]

Spent some time at it tonight and I think I've made some progress. Pulling out the transistors had the same effect as having the delay time at max, and the oscillations did not stop. Using a 100k pot on all 3 stages sortoff worked, for the first stage it did nothing, the second stage had some slight delay time change, and the 3rd slowed the delay time no problem, so maybe it has something to do with the first or second stage. I also reconnected the modulation and i noticed that with it turned up i do get modulation on the delay time and on the little dry signal i can hear, so it looks like the transistors are actually doing their job correctly. This seems to rule out the, being at fault, and since i know the modulation and dry signal work that only leaves the delay path and PT2399 chips to look at. 

EDIT: Even though the modulation does work i also get a clicking from the LFO, even with the wire from depth 2 disconnected. Not sure if this is relevant but i just noticed

[HeavyFog]

Alright, so i finally got this thing running good, turns out when i swapped the original chips for the new ones i bought, i must have accidentally placed 2 of the same ones back in, because I put 3 fresh ones I knew hadn't been used yet in, and the damn thing worked! So it means i either fried the ones i got from smallbear somehow, or they're fakes. I kinda doubt the latter, but il send smallbear an email regardless just incase they have a bad batch and they don't know it.

Sound wise? Its a pretty cool sounding delay, not overly digital sounding and pretty dark, with a lot of delay time to work with.I might have a go at modding it to get the delays louder, and maybe to brighten it up a bit, but overall i like it!

Thanks for the help everyone, if there's a moral to this story it would probably be to check your parts and buy extras incase you fry them!

[HeavyFog]

Mods are in and WOW this thing is good. Some changes for anyone else building one of these:

-Change R12 from 100k to 24k
This will bring the delay volume on par with the dry volume

-Lower R20 from 27k to somewhere around 20-15k
Doesn't say on the doc, but lowering this resistor will increase the total amount of feedback possible. Get it just right and you can get endless repeats without oscillation!

-Lower R27 from 240k to 100k
Takes the modulation from being very subtle to full on vibrato

-Add a 2.2uf cap to ground from IC1 (LFO) pin 7 to ground
-Cut trace between IC1 pin 4 to ground and bridge with 220ohm resistor
This is brings it into spec with the echo base v2 LFO, and really helps suppress LFO ticking

And just for fun...
Add an LM386 sound 2 light circuit (deadastronaught) to lug 3 of the level pot so you get an LED that blinks with each delay!

[Yupnitski]

Hello folks!
Just digged up my build from my "fix this" box and checked it thoroughly for any errors in component values etc. Only error was C21 which is supposed to be 4,7pF and I had put a 47pF in. That didn't do anything much I guess since it still has a "rubberbandish" effect to the delay. All the pots work as intended. I've tried different PT2399's without any audible change, so I guess I could have bad chips all the way through OR there's something else going on
Also, I didn't realize that regulator error discussed before (I have a TO-92 reg there now). So that's going to be something I have to fix 1st. After that, I'll test out the mods HeavyFog kindly presented above Will keep this post updated if I have any success...and even if I don't

[Yupnitski]

Nah. Couldn't get rid of the wobble in the sound. The "vibrato" is there even with the lowest Rate setting (yes, the Rate pot is the only control that I have over the wobbling). Got to go over mu PT2399 stash with a known working delay to scratch the bad ones. I do think I did that in the past before I populated this one, but...aww, that's like 3yrs ago

[ElectricDruid]

Nah. Couldn't get rid of the wobble in the sound. The "vibrato" is there even with the lowest Rate setting (yes, the Rate pot is the only control that I have over the wobbling).

Why is the Rate control the only control you've got over the wobbling? What happened to the Depth control? And if you haven't got a depth control, how did you expect to get zero depth?

More explanation, please.