Rebote 3 delay refusing to cooperate

[HeavyFog]

http://www.tonepad.com/project.asp?id=74
Anyone else have any trouble with these? Working on one now and so far no luck getting it to work properly. The problem seems to stem from the Time knob, when I turn it to max I get very grainy delays at about 3-4 seconds apart, much longer than the 2 seconds it's supposed to work at. I was expecting the max delay time to be somewhat grainy, but this seems way too much. Turning down the time pot makes it self oscillate and distort all over the place, and past a point the repeats and feedback cut out all together and all i get is dry signal. Every other knob seems to be working fine, although I do have the modulation disconnected for the sake of debugging.

I've tried replacing the PT2399 chips to see if one was fried, but no fix. I've also triple checked my parts and all solder joints and i cant seem to find any mistakes, missing jumpers, or faulty parts anywhere, so I'm kinda at a loss. Did some digging and i read on some foreign forums that some people have had this same issue, so i'm not sure if there's something i'm missing, or if i've just got this wired wrong.

I did notice though that the lm7805 (used in place of a 78l05) gets somewhat hot, which it shoudn't given the delay chips only draw about 30mA each. Other than that the only thing i could see being a part of it is Q1-3 which control the delay time. I used a set of central semi 2n5087s for this, and i have them put in according to same pinout these usually use. Unless the central semi ones in particular have a reverse pinout i have no idea, the datasheet just says "EBC PINOUT" which i assume is the same as it usually is in most other transistors (2n3904, 2n5088, etc...). Il get some voltages next chance i get, but for now i could use some help, since i'm pretty stuck 

[antonis]

the lm7805 (used in place of a 78l05)

Tonepad BOM calls for L7805, which is a TO-220, 1.5A device and NOT 78L05 (TO-92, 100mA)
file:///C:/Users/user/AppData/Local/Temp/l78.pdf

For about 100mA 5V current draw (rough estimation due to PT2399 current draw reliance on delay time ), it should only get moderately warm..

You could interpose an Ammeter and take actual current draw measurement..
(or a 1R resistor and take voltage drop measurement..)

[ElectricDruid]

If an LM7805 TO-220 package regulator is getting hot, it sounds like there might be a short on the board somewhere that's causing a hefty current draw.

I mean, there are things you can build that will draw that kind of current, but the Rebote3 delay shouldn't be amongst them, even with three PT2399s.

[duck_arse]

if only we could see some photos of the build.

[ElectricDruid]

...Sigh!...if only...

[antonis]

Uh-oh..!! You eye-minded guys.. 

[HeavyFog]

Sorry, was at work and i haven't had access to get a pic until now. Here's some shots of the board. The wiring is messy i know, i just haven't cleaned it up yet. Otherwise I've checked it all over and the offboard wiring looks good.





Forgot to mention aswell that i am powering it on a 12vDV supply, and regulating it with an LM7809. Also using a relay for bypass switching.

[HeavyFog]

I realize that the lm7809 is standard for this build. The PCB has the symbol for 78L05 so i had assumed that was the norm here, my bad. The regulator doesn't get hot for say, it gets warm but definitely not enough to get too hot to touch. I guess this is normal and i had just assumed it would get warm at around 100mA if it was meant to hand 1A

[anotherjim]

I just found out that Rebote is Spanish for a rebound. I'm probably the only one here who didn't know that.
Carry on...

Oh, don't forget a TO-220 7805 pinout is opposite (input on left reading the part number) to a little 78L05 (input on right reading part number).

[HeavyFog]

I made sure before the regulator went in that i got the pinout right. Checked the datasheet and layout and it seems to work fine. I'm getting the correct 5vDC on its output and on pin 1 of each pt2399, and the regulator seems to be as hot as it should be for a 100mA draw.

What's got me sketched out is that you couldn't even view the project pdf for a while. Something about copyright or the previous rebote 2/2.5 being cloned by a bigger company, don't remember. Tonepad stuff is great and I've used it time to time for years now, but the nature of this project specifically makes me wonder if there's a flaw in the design or an error in the PCB, especially since others have reported very similar problems, and i cant seem to really find anyone who's actually built one of these successfully.

Now its most likely i've screwed something up along the way but i cant be sure. Maybe il send an email to tonepad just to check with them 

[ElectricDruid]

I realize that the lm7809 is standard for this build. The PCB has the symbol for 78L05 so i had assumed that was the norm here, my bad. The regulator doesn't get hot for say, it gets warm but definitely not enough to get too hot to touch. I guess this is normal and i had just assumed it would get warm at around 100mA if it was meant to hand 1A

The PCB shows a LM7805, T0-220 package. The LM78L05 is the smaller TO-92 package.

You've clearly got the LM7805 on the PCB, and it's the right way around, and you've got 5V, so that all looks good. Warm but not too hot to touch sounds ok.

I've been able to see the project pdf this morning, so if it was down for some reason, it's back up now:
http://tonepad.com/getFile.asp?id=151

[ElectricDruid]

The schematic shows the interesting way they've got around the problem of controlling three delay times at once. Each delay has a 2N5087 PNP tranny where the delay pot would go if it were a single delay, and then the single delay time control controls all three transistors.

As such, that part of the circuit is going to be very sensitive to the values of R4, the pot itsalf, and R37 and R31, plus the base resistors of each delay (R43, R6, R41, according to the schem). I'd check those.

One question though - why have they used a PNP? I don't get that.

[HeavyFog]

Just checked, i've got all the correct values in for those parts, so it wouldn't be that. But it does seem to have something to do with the time pot and the related circuitry.

No clue why the pnp route though. Might have something to do with the availability and inconsistency of through hole JFETs, since that's how I would have tried it.

[HeavyFog]

I also just read this on another post (https://www.diystompboxes.com/smfforum/index.php?topic=108526.0); "I cut the trace between R4 and C13, bridged R4 to IC2 as in the fix on the website." I couldnt find any other evidence of this anywhere, and there's no mention of it on the tonepad site. I checked the traces on mine and that specific connection looks good to me, but this only makes me a little more sketched out at this project.

[ElectricDruid]

Just checked, i've got all the correct values in for those parts, so it wouldn't be that. But it does seem to have something to do with the time pot and the related circuitry.

No clue why the pnp route though. Might have something to do with the availability and inconsistency of through hole JFETs, since that's how I would have tried it.

What I can't understand is that PNPs are supposed to have the base *lower* than the emitter to conduct. So in this situation, they're being operated in a reverse-bias mode. Which is pretty unusual and I don't know why you'd choose that over a typical NPN in forward-biased mode - to control the current flow from collector to emitter, that seems like the obvious choice. JFETs would work but are more expensive and more complicated for the same simple job. So what's going on? Is this deliberate? Why? Is this an error?

I don't get it.

[HeavyFog]

Yeah its kinda odd. Any pics i've seen of completed ones the view of the specific transistors isn't clear, so who knows. I would have figured though that if the case was that the PNP ones dont work someone would have pointed out by now, but idk. Wonder if a set of 2n5088 or simmilar might give better results with a bit of modifying

[PRR]

....that PNPs are supposed to have the base *lower* than the emitter to conduct. So ....I don't get it.

The base drive is AC coupled and may go low of ground. The reverse connection traditionally gives lower offset voltage. I don't say it is right as drawn. I do say it is reasonable for somebody.

[HeavyFog]

Tried a few different transistors, BC557, 2n3906, 2n5087 from a different maker, still nothing, i get the exact same problem. I'm pretty much out of ideas at this point, nothing so far has worked and i'm starting to think the design itself must be flawed.

I'm considering scaping this all together and just building a rebote 2.5, this just doesn't seem worth the trouble

[anotherjim]

Note the ground for the transistor emitter is pin4 "digital ground". This is connected to pin3 internally via a small resistance in the chip. This may let the emitter voltage rise to some positive value?

[ElectricDruid]

....that PNPs are supposed to have the base *lower* than the emitter to conduct. So ....I don't get it.

The base drive is AC coupled and may go low of ground. The reverse connection traditionally gives lower offset voltage. I don't say it is right as drawn. I do say it is reasonable for somebody.

Thanks PRR.

Note the ground for the transistor emitter is pin4 "digital ground". This is connected to pin3 internally via a small resistance in the chip. This may let the emitter voltage rise to some positive value?

Yeah, I wondered about that ground too. It's about the only place it's used on the whole schematic. If (say) the digital ground connection had been forgotten, that might explain why the whole thing doesn't work.

The basic idea of using a transistor to control the current out of pin 6 is fine, so it *can* work in principle. The question is why it doesn't in practice.