Pleeeease Help Me! PWM/Rambler Problem

[Jaicen_solo]

Ok guys, I've been here a while but this is the first real design i've posted.
It's actually a bastardisation of two of Tim Escobeda's designs (rambler and PWM), minimised to suit my needs.
I perfed it and it works fine, but I just can't make my pcb version work, I was hoping someone here could spot where i've gone wrong!

This has been hanging over me for some time now, i'm just about pissed off with it!
If anyone can help, i'd be greatly appreciative!

*b

Layout is here: http://storage.msn.com/x1p3q6ujd6xcOpOfW-G0f53Of6axOUXKnK79AStM-yR3ZWWKd55JTn9rFOKsYHhJEGQix6jzR_VDBKQ96-ikdZjPekNt5Z6VeP_Os5OzlyfKc5BJ3OG3Z_DR22MiJxy_X_L8znfY2uJ_F9DuVV8woe-7g

[bioroids]

Hi!

You should post a schematic for the circuit. Otherwise, how could we tell if the pcb is good or wrong?

Luck

Miguel

[Hal]

also a more desctriptive title...

[Jaicen_solo]

Right, there is no schematic at the minute. Like I say, it was a design based on two of Tim's circuit snippets, found here:

http://www.geocities.com/tpe123/folkurban/fuzz/snippets.html

I used the Rambler as a front end for the PWM. The Dual op-amp buffers and splits the signal for the blend control.
The problem appears to be around the 386, as the signal disappears after the input.
The schematics i'm working from can be found here:

http://www.geocities.com/tpe123/folkurban/fuzz/rambler.gif

http://www.geocities.com/tpe123/folkurban/fuzz/pwm.gif

[Jaicen_solo]

Nobody?? 

I'm really frustrated by this project guys, I'm sure someone here can lend some advice!

*bob

[bioroids]

I think is really important to make a schem first with the definitive circuit, and then making the pcb following it. That way you can check easily if everything goes where its supposed to. There's even software that can aid you this way. By the way, how did you do the pcb?

If you do the schem and upload it, then we can follow the layout and try to spot some mistakes. Also the layout is too small from my eyes, you should upload a bigger image to make it easier to debug!

Luck

Miguel

[Jaicen_solo]

Well, I made the layout by hand using a piece of paper, a pencil and my original perfboard design. I pretty much bypassed the schem stage, as I didn't really think it was necessary. As it is, it would appear the schem is in fact essential, so i've drawn one up. It can be found here:

http://storage.msn.com/x1p3q6ujd6xcOpOfW-G0f53Of6axOUXKnK79AStM-yR3ZVLf7bHHeVRkMeSBRt8btcdg8L8r7JXndDGRjUS_jJlYoBF4tpLcs1bmVVQqyl0YCCZOCeD8IKWgJkFEJd38O3-TYxHtWB_f_8Cs4jP48q6Fw

FWIW, if you save the images and load them in windows imager or whatever, they'll be much larger. Windows seems to resize things without permission. Anyway, I really need help with this one, I seem to be getting no signal after the opamp now.
Any ideas guys??

[bioroids]

I think you're not biasing the opamps correctly?

If you're using a single supply (like most stompbox circuits) you need to bias the +input to the middle of the supply (called Vref usually) instead of ground.

So you should make a voltage divider (two 10k resistors in series from +V to GND, a 47uf cap with the + side to the junction of the 2 resistors and the -side to GND. The junction of the resistors and the cap is the Vref point)

You have to connect the 1M resistor from the +input of IC1a to the Vref point.

You need also a capacitor at the output for blocking DC. I see some danger on connecting the side of the Blend pot to GND when the Blend switch is off.

Also the standard way is to put another resistor from the -input of the opamp to Vref, and this sets the gain of the stage to the value of the feedback resistor (1M) divided by the resistor to Vref, plus 1.

If this is too complicated to follow (my poor english doesnt help either) I try to draw it over your schem if you want. Just let me know.

Luck

Miguel

[Jaicen_solo]

Ok, so is there any way to modify the current layout to make it 'work' ?? Perhaps by creating a daughter board for the power supply? If so, how do I go about doing it?? I'm not sure I fully understand what you're saying.
The problem is, i've already built and populated the board and I was hoping I could make it work somehow. I was hoping to give the pedal to a friend of mine next week, guess that's not gonna happen!

[bioroids]

I think there must be a way to make it work ok, i'm not sure how it will sound by the way.

There are some capacitors that you should add too, is the board too tight or is there place to make some new holes and slide some caps in?

I made some changes to the circuit, I think it will work this way, and you have to make minor changes (assuming the layout is following the schem you posted)

Check it out: http://www.dedalofx.com/bioroids/guitarsynth_v2.jpg

How much gain did you need on the first stage? I changed it to a buffer, but if you want gain there are more changes to be made. Also there is a track that you'll have to cut and add a capacitor (marked with the red cross)

Try this and post what happens!

Miguel

[Jaicen_solo]

Ok, I tried your mods but it still doesn't work 

I get a small signal after the opamp, but it is lower level than the passive signal and sounds somewhat mis-biased.

I also swapped out the 10uf electro for a 0.47uF cap to make sure it wasn't a polarity thing. I just don't understand why it doesn't work now, when it worked on perfboard!?

[bioroids]

There must be something wrong on the layout or the construccion. Let me check it.

Miguel

[bioroids]

Please can you post the voltages reading at all the Opamp pins?

Also, there seems to be too little space between tracks on the pcb image, are you sure there are no short circuits because of this?

Miguel

[Jaicen_solo]

Yeah, the PCB image doesn't fit op-amps, as I had not templates so I drew the layout by hand using a sharpie. All the traces are cool, definitely no shorts as i've been over it a million times! 

I'm not sure if the voltages are going to be helpful. I did them all wth a PSU as the only battery I had was in my multimeter 
I'm afraid i'm not too knowledgable about IC's so i'm hoping you guys are gonna spot any problems!

Voltages are as follows:

TLO72

1: 7.3v
2: 7.3v
3: Ramps down from around 3v to 0v when probed?! Strange
4: 0
5: 3.6v
6: 4v
7: 4v
7: 12.3v (unregulated supply I guess)


NJM386

1: 3.85v
2: 1.39v
3: 4.7v
4: 0v
5: 11.73v
6: 12.35v
7: 7.2v
8: 0v

40106 Schmitt

1:10.7v
2: 0v
3: 0v
4: 12.3v
5: 0v
6: 12.34v
7: 0v
8: 12.34v
9: 0v
10: 12.34v
11: 0v
12: 12.34v
13: 0v
14: 12.34

[bioroids]

I cant help much with the 386 and the 40106 as I have no experience with those ICs.

But the opamp does luck strange. You should have very close too Vcc/2 in all pins except 4 and 8.

I see in the layout the 1M resistor to ground is on the wrong side: it should be between pin 3 and 4, and you have it on the other side of the cap! And it should be there too the new 1M resistor that I added to your schem (between +vcc=pin 8 and pin 3)

Miguel

[Jaicen_solo]

Hey guys, I just got back to this after a bit of a break. Seems that Miguels' ideas about the opamp being mis-biased are spot on!
With a 12v supply, the pins are all at half voltage, resulting in a strong buffered signal at the output.
Unfortunately, there's still something very wrong with the 386/40106 section. Using the probe, I seem to be getting 50hz hum on the output, so perhaps the 386 is dead. I'm going to get some more chips soon and see if that helps, but in the meantime could somebody re-check my layouts see if i'm missing anything else??
For reference, i've now posted the new pin voltages for the dual opamp below.

1: 6v
2: 6v
3: 4.4v
4: 0v
5: 6v
6: 6v
7: 6v
8: 12v

EDIT: Well, i got it working tonight, turns out it was something to do with the diodes. Strange, but I took them out since I didn't really need the octave function. Anyway, as I say, it works but it doesn't sound as good as the original. There's very very little sustain off the PWM. Running it through my Whammy in bypass helps, but not much. Perhaps the 40106 has been damaged, it's been messed with plenty. Anyway, I'll get a new one and report back.

[bioroids]

Good, hope it works OK with the replacement chips.

I really have no idea how the pwm and the other circuits work, so I would be of little help at that 

Luck

Miguel