Help - Rebote Delay 2.5 REPEATS problem

[Jaan127]

Hi everyone

So... I'm building the Rebote Delay 2.5 pedal found on Tonepad
Page info: http://www.tonepad.com/getFileInfo.asp?id=98
Schematic: http://www.tonepad.com/getFile.asp?id=98

And I'm having some trouble right now with the repeats. Before reaching the 'Repeats POT' there is this Resistor which should give the option to get endless repeats, It's set to 15k and for getting the endless repeats you should change that to 10k and below.
Thing is, I don't want endless repeats, I'm ok with what the 15k "should" do, BUT every time I try to go about 60% above of the repeats POT the pedal just goes nuts, I start getting an endless repeated noise that isn't even from my guitar, it seems to be auto feedbacking over and over, and that noise is there even if you haven't plug the guitar.

So I thought the PT2399 IC was not working right, but I have 10 of these ICs and the same thing happens with everyone
I also thought on changing the 15k resistor to a higher value (20k) but it will just CUT the repeats to about a maximum of 3 repeats which is too little.

Anybody has an idea of what is happening? the PT2399 has a digital and analogic ground which are both connected kind of as if they were the same in this schematic. But that connection seems to work for everyone else so I'm out of ideas.
Maybe the OP Amp is the one not working right?

thank you in advance

[slacker]

The first thing I would do is check the connections to the repeats pot, make sure pin 1 of the pot is connected to ground. After that check the values of the resistors around pins 13 and 16 of the PT2399, wrong values could be boosting the signal too much causing the problem you're having.

Welcome to the forum

[Jaan127]

I checked and Pin 1 of the POT is connected to ground
Resistors around pins 13 and 16 of the PT2399 are correct too, although that 5% tolerance is awful in some of them.
I'll see if I can get a 1% tolerance resistors and try again

Thanks for the welcome 

[ElectricDruid]

The first thing I would do is check the connections to the repeats pot, make sure pin 1 of the pot is connected to ground. After that check the values of the resistors around pins 13 and 16 of the PT2399, wrong values could be boosting the signal too much causing the problem you're having.

Welcome to the forum

+1 agree. The only way I can see that you'd get runaway feedback like you describe is if one of those two filters is actually providing some gain. That shouldn't happen, so I'd expect there's an error somewhere in one of those. Are the caps grounded correctly? Are the 47K/24K feedback resistors going to the right place? Are the caps from the output back to the inverting input connected correctly?

HTH,
Tom

[PRR]

> get a 1% tolerance

This does not sound like a 4.7K versus 4.747K error.

More like 47K or 470r where should be a 4.7K.

Resistor-reading was tough when makers kept their paint-pots clean and I had natural daylight in my shop. Now parts are small, paints are muddy, and it's got hard to read these things.

Note that many 1% resistors have way too many too-skinny stripes of odd colors. Until proven otherwise, I'd rather work with 5% parts.

[ElectricDruid]

Note that many 1% resistors have way too many too-skinny stripes of odd colors. Until proven otherwise, I'd rather work with 5% parts.

Nar, that's what multimeters are for. I either recognise a resistor value, or I measure it. I don't ever actually *read* them. There's no point when errors are so easily avoided.

Totally agree with your point about the values though - this is not a problem caused by 1% tolerance.

T.

[Jaan127]

The first thing I would do is check the connections to the repeats pot, make sure pin 1 of the pot is connected to ground. After that check the values of the resistors around pins 13 and 16 of the PT2399, wrong values could be boosting the signal too much causing the problem you're having.

Welcome to the forum

+1 agree. The only way I can see that you'd get runaway feedback like you describe is if one of those two filters is actually providing some gain. That shouldn't happen, so I'd expect there's an error somewhere in one of those. Are the caps grounded correctly? Are the 47K/24K feedback resistors going to the right place? Are the caps from the output back to the inverting input connected correctly?

HTH,
Tom

Is not a resistor-reading problem at all.

I double-checked and resistors are correct /values and connections
the caps are grounded correctly too

"Are the caps from the output back to the inverting input connected correctly?"
if you mean all the caps around pins 13 and 16, they are

Dunno where the problem is 

[ElectricDruid]

Ok, let's go back to square one. There must be some assumption we've made that's wrong somewhere.

You said when you turn the Repeat pot past 60%, the pedal makes a noise.

Does the noise change when you change the delay time? If it's feedback in the delay, the pitch of the sound will vary with the delay time, coming down and down as the delay gets longer.

Does it make any difference whether there is a signal or not? What happens when you play into the pedal when it is making this noise? Can you hear the guitar, or only the noise?

Thanks,
Tom

[Jaan127]

Ok, let's go back to square one. There must be some assumption we've made that's wrong somewhere.

You said when you turn the Repeat pot past 60%, the pedal makes a noise.

Does the noise change when you change the delay time? If it's feedback in the delay, the pitch of the sound will vary with the delay time, coming down and down as the delay gets longer.

Does it make any difference whether there is a signal or not? What happens when you play into the pedal when it is making this noise? Can you hear the guitar, or only the noise?

OK, so I tested and:
-Yes, the noise changes with the delay time, it actullay matches the delay time, until it just goes nuts as it seems to be feedbacking itself
-No signal at all and Input jack connected but guitar unplugged would just make a lot of noise
-I can hear both the guitar and the noise, but eventually the noise is so heavy it overcomes the guitar

[ElectricDruid]

Ok, so we've definitively established that the noise is coming from feedback around the delay like you thought initially.

Since the only route for feedback is from the Repeat pot via that 15K resistor, it has to be the circuit within that loop.
That's the two filters made with the op-amps on the PT2399 chip. Are the filter caps correct? Getting two caps the wrong way around could give you a resonant filter instead of a flat response, and that would produce feedback.

I'm running out of ideas, because there aren't that many options. The problem has to be somewhere there - cracked trace, maybe? Bad joint? Tiny bridge of solder making a short?

Tom

[Jaan127]

Ok, so we've definitively established that the noise is coming from feedback around the delay like you thought initially.

Since the only route for feedback is from the Repeat pot via that 15K resistor, it has to be the circuit within that loop.
That's the two filters made with the op-amps on the PT2399 chip. Are the filter caps correct? Getting two caps the wrong way around could give you a resonant filter instead of a flat response, and that would produce feedback.

I'm running out of ideas, because there aren't that many options. The problem has to be somewhere there - cracked trace, maybe? Bad joint? Tiny bridge of solder making a short?

Tom

I set up everything on breadboard, so there isn't the tipical problems with the PCB

The caps for the OP amps around pins 13 to 16 are 560pF (code561) and 1nF (code 102) so they are correct

However the caps for the OP amps (integrators) around pins 9 to 12 are 100nF instead of 82nF as I dont really have the 82nF ones. If that is the problem then I'm really dumb for assuming the 100nF would work instead

[Jaan127]

pls 

[slacker]

100n caps are fine instead of the 82n. Have you checked your pots are the correct value and are log ones, if you used linear pots it would probably self oscilate too early in the pots rotation. Are you sure just making the 15k resister larger won't fix the problem?

[anotherjim]

One thing I find that makes repeats easier to maximise without hitting runaway feedback, is to reduce the bass content of the signal feedback path. That's as easy as reducing the value of that 0.1uF cap in series with the 15k. Try 47nF.
It can still runaway if the signal gets large enough, but there's a bit more travel on the repeat pot and the repeats sound clearer.

[Jaan127]

100n caps are fine instead of the 82n. Have you checked your pots are the correct value and are log ones, if you used linear pots it would probably self oscilate too early in the pots rotation. Are you sure just making the 15k resister larger won't fix the problem?

B25K, that's logarithmic right?
I tried making the 15k larger, to 20k, but it cut out the repeats, I could get a maximum of 3 repeats and the feedback noise althought decreased, was still there. Making it even longer would leave me with even less repeats =/

[Jaan127]

One thing I find that makes repeats easier to maximise without hitting runaway feedback, is to reduce the bass content of the signal feedback path. That's as easy as reducing the value of that 0.1uF cap in series with the 15k. Try 47nF.
It can still runaway if the signal gets large enough, but there's a bit more travel on the repeat pot and the repeats sound clearer.

Alright, I'll try that

[Jaan127]

One thing I find that makes repeats easier to maximise without hitting runaway feedback, is to reduce the bass content of the signal feedback path. That's as easy as reducing the value of that 0.1uF cap in series with the 15k. Try 47nF.
It can still runaway if the signal gets large enough, but there's a bit more travel on the repeat pot and the repeats sound clearer.

Alright, I'll try that

It did help just a little bit, but the problem is still there-

[Lothric]

100n caps are fine instead of the 82n. Have you checked your pots are the correct value and are log ones, if you used linear pots it would probably self oscilate too early in the pots rotation. Are you sure just making the 15k resister larger won't fix the problem?

B25K, that's logarithmic right?
I tried making the 15k larger, to 20k, but it cut out the repeats, I could get a maximum of 3 repeats and the feedback noise althought decreased, was still there. Making it even longer would leave me with even less repeats =/

B25k is not logarytmic, it is linear. Doubble check the pot, measure it's resistance. In my opinion maybe You use even rev-log instead of Linear pot?

[Jaan127]

B25k is not logarytmic, it is linear. Doubble check the pot, measure it's resistance. In my opinion maybe You use even rev-log instead of Linear pot?

Then what does the B stands for?
it's in the back of the POT (Schematic)

I'll specifcaly get a log one then. 

[Lothric]

B25k is not logarytmic, it is linear. Doubble check the pot, measure it's resistance. In my opinion maybe You use even rev-log instead of Linear pot?

Then what does the B stands for?
it's in the back of the POT (Schematic)

I'll specifcaly get a log one then.

Sorry, You must get me wrong. I thought that You think that B on your pot stands for logarythmic. Don't change anything, stick to what You have on the schematic.
B- stands for linear A-for logarythmic - C - for reverse logarythmic

http://obrazki.diyaudio.pl/15781/434thi8kfalszknxz4pq_1396113618.gif