[DEBUG] GGG 6 band EQ turned into 10 band

[JRB]

1.What does it do, not do, and sound like?
It hums like crazy when in bypass. The hum gets amplified when the circuit is activated and the hum changes when turning the 31,3Hz 62,5Hz and 125Hz. There is no difference between using a power supply or a battery.
The hum is so bad it makes the EQ unusable.

2.Name of the circuit = GGG EQ

3.Source of the circuit (URL of schematic or project) =
http://www.generalguitargadgets.com/projects/18-eq/46-6-band-graphic-eq

4.Any modifications to the circuit?
Turned it into a 10 band.
Runs on a single 9V battery/power supply
Using the schematic as reference I also changed the following according to some tips I found on this forum like half a year ago(don't have a link).
IC1b pin 5 is hooked up to point C
150pF is hooked up to the - terminal of the battery
10K at the output jack is hooked up to the - terminal of the battery
220k resistors are hooked up to point C



Jacks bypass
In: 0V
Out: 0V

Jacks active
In: 3V   <----------------------Is this normal for this circuit?
Out: 0V

Battery
8,09V

IC1:TL084
1: 4,2V     Input buffer
2: 4,2V     Input buffer
3: 4,17V   Input buffer
4: 8,3V
5: 4,1V    Output buffer
6: 4,1V    Output buffer
7: 4,1V    Output buffer
8: 4,1V
9: 4,1V
10: 3,41V
11: 0V
12: 3,41V
13: 4,1V
14: 4,1V

IC2 and 3: TL084
1: 4,1V
2: 4,1V
3: 3,41V
4: 8,3V
5: 3,41V
6: 4,1V
7: 4,1V
8: 4,1V
9: 4,1V
10: 3,41V
11: 0V
12: 3,41V
13: 4,1V
14: 4,1V

[JRB]

Managed to get the DC of the input jack by turning the inputcap around and lowering it to 4,7uF. It stills makes noise like crazy.

[PRR]

Somthing not grounded, or it is not boxed-up.

[JRB]

It is boxed up so that couldn't be it.

When I have some free time this week I think ill just redraw the schematic from the board, seen someone suggest doing that in a other topic.

[Mike Burgundy]

That's always a good idea.
did you wire it up for true bypass? The published schem has no bypass setup, so no idea what it looks like. Also, since you've altered (expanded) the schematic, it's more difficult to see which bit goes where, without having a schem on *what you built*. Most likely there is one or more bad solder joints, dry or bridged. If it is wired for TB, start with the jacks and switch wiring. DC on the input should not be possible. If it was cured by flipping the input cap (which as per schem was the wrong way round - IC side is the + side), that suggests a solder error (bridge) more than an extremely leaky cap. It might be a good idea to replace that cap anyway since it has been reverse biased and they don't like that - it will probably develop a problem in the future. Oh wait, you changed the value - so that's a new cap.
The 3.41 voltages look wrong too. All IC in/out pins should be at 1/2Vbat.

[JRB]

That's always a good idea.
did you wire it up for true bypass? The published schem has no bypass setup, so no idea what it looks like. Also, since you've altered (expanded) the schematic, it's more difficult to see which bit goes where, without having a schem on *what you built*. Most likely there is one or more bad solder joints, dry or bridged. If it is wired for TB, start with the jacks and switch wiring. DC on the input should not be possible. If it was cured by flipping the input cap (which as per schem was the wrong way round - IC side is the + side), that suggests a solder error (bridge) more than an extremely leaky cap. It might be a good idea to replace that cap anyway since it has been reverse biased and they don't like that - it will probably develop a problem in the future. Oh wait, you changed the value - so that's a new cap.
The 3.41 voltages look wrong too. All IC in/out pins should be at 1/2Vbat.

For the bypass I just put the switch between the jacks and the input/output of the circuit board like you would do for every circuit with true bypass.

When measuring the IC's I started wondering if they might have been fried or not because of those voltages you mentioned. I just don't really know of a quick and easy way to check if they are fried or not.

I also managed to find the post with the suggested changes I used.
http://www.diystompboxes.com/smfforum/index.php?topic=69194.msg554490#msg554490

[Mike Burgundy]

Still a bit of a treasurehunt. for example: the schem is for a 6-band, as you stated. You have a 10-band. How did you do that? Did you build the 5-band version twice, and put those in series using the first boards input buffer and the second boards output buffer? Did you use the board design on GGG? did you use another layout? Draw it up, exactly as you have it there (including pin numbers), and preferably take good HQ pictures of the PCB's.
That, combined with voltages, will make debugging a little easier.
In the mean time: did you socket the IC's? Did any voltages change after you changed the input C? Do you know how to trace the signal through the circuit? (If not, search "audio probe" on this forum - wait, it's under the debugging link top left, right under the transistor logo)

[JRB]

Voltages are the same.

This is the schematic off the board as it is right now I couldn't really find anything that was out of order.



And here are the photo's you asked for the links are more detailed shots. IC1 is the one on the right, IC2 the middle and IC3 the left.




http://i41.photobucket.com/albums/e294/jtheunknown/2013-01-21203948.jpg
http://i41.photobucket.com/albums/e294/jtheunknown/2013-01-21204013.jpg
http://i41.photobucket.com/albums/e294/jtheunknown/2013-01-21204002.jpg
http://i41.photobucket.com/albums/e294/jtheunknown/2013-01-21204058.jpg
http://i41.photobucket.com/albums/e294/jtheunknown/2013-01-21204053.jpg
http://i41.photobucket.com/albums/e294/jtheunknown/2013-01-21204047.jpg


EDIT: Seems the schematic picture didn't turn out that great ill see if I can fix that.

[Mike Burgundy]

Looking forward to it. Okay - sockets, that helps. Gently take out the IC's and remeasure voltages. You should get + voltage on pin 4, ground on 11, Vref (half supply voltage) on 3/5/10/12, except one single opamp (so 1x either 3/5/10 or 12 on either IC1,2 or 3): the output buffer

[JRB]

Here are the new voltages. almost exactly like you predicted, so I guess the IC's are fried right?

Jacks bypass
In: 0V
Out: 0V

Jacks active
In: 0V
Out: 0V

Battery
9,4V

IC1:TL084
1: 0,12V     Input buffer
2: 0,12V     Input buffer
3: 4,6V   Input buffer
4: 9,3V
5: 0,11V    Output buffer
6: 0,11V    Output buffer
7: 0,14V    Output buffer
8: 0V
9: 0V
10: 3,9V
11: 0V
12: 3,9V
13: 0V
14: 0V

IC2 and 3: TL084
1: 0V
2: 0V
3: 3,9V
4: 9,3V
5: 3,9V
6: 0V
7: 0V
8: 0V
9: 0V
10: 3,9V
11: 0V
12: 3,9V
13: 0V
14: 0V

[Mike Burgundy]

There's something leaking around IC1. pins 1/2/5/6/7 should not be connected to anything passing DC.
Glad to see there's a new battery involved ;P (quick thought: if that is the same battery something is shorting!)
Pins 3/10/12 on all i cs and also 5 on IC2 and IC3 should be at 1/2V. Only pin3 on IC1 has that, rest is low - bad connections? Partial short? Looks like a bit of both?
What is the voltage between the two 10k's of the Vref voltage devider? All those pins should carry that. Aside from the power pins, all others should carry nothing.

For a start, make sure you have the correct voltages on those pins, and nothing where it shouldn't be. This will probably mean selectively disconnecting stuff to see wether that changes everyting. I would take out the two leads to the EQ gyrators (the two going down from between IC1a and IC1b) and concentrate on getting IC1 in good working order first.

There's a LOT of solder on that board - I can see many opportunities for shorts. You don't need this much solder! See if you can suck some of that off (solder pumps and solderwick, definite must-have), this will greatly reduce risk of shorts/conduction paths. Do make sure all joints are good. Inspect under good light for any bridges/hairline connections etc. Run an exacto knife between the traces to cut anything you may have missed.
Am I correct that the IC numbering you referred to in the pictures does not correspond with the schematic? IC1 is input/output buffer on the schem, which on the board is to the left, not right. I'm sticking to the schem, preferably.

[JRB]

There's something leaking around IC1. pins 1/2/5/6/7 should not be connected to anything passing DC.
Glad to see there's a new battery involved ;P (quick thought: if that is the same battery something is shorting!)
Pins 3/10/12 on all i cs and also 5 on IC2 and IC3 should be at 1/2V. Only pin3 on IC1 has that, rest is low - bad connections? Partial short? Looks like a bit of both?
What is the voltage between the two 10k's of the Vref voltage devider? All those pins should carry that. Aside from the power pins, all others should carry nothing.

For a start, make sure you have the correct voltages on those pins, and nothing where it shouldn't be. This will probably mean selectively disconnecting stuff to see wether that changes everyting. I would take out the two leads to the EQ gyrators (the two going down from between IC1a and IC1b) and concentrate on getting IC1 in good working order first.

If I am looking at the schematic shouldnt the voltage on Pins 3,5,10,12 be lower then Ureff cause of the 220k resistor between it and the pins? The Ureff is around 4,5V
One thing I noticed when measuring the pins of the in/output buffer was that it was slowly going down to zero. It starts at around 0,20V and then go's down rapidly to 0,10V where it go's slower and slower towards 0V.

There's a LOT of solder on that board - I can see many opportunities for shorts. You don't need this much solder! See if you can suck some of that off (solder pumps and solderwick, definite must-have), this will greatly reduce risk of shorts/conduction paths. Do make sure all joints are good. Inspect under good light for any bridges/hairline connections etc. Run an exacto knife between the traces to cut anything you may have missed.
Am I correct that the IC numbering you referred to in the pictures does not correspond with the schematic? IC1 is input/output buffer on the schem, which on the board is to the left, not right. I'm sticking to the schem, preferably.

I admit it has a lot of solder however using less I couldn't get the solder to flow over the perfholes to make the lanes. Ill see if I can do it with a bit less solder but using my mk1 eyeball it doesn't connect anything between them. I did gave the knife ting a go what a mess does that give all the yellow dust

The pins on the schematic I drew and the board match. If you look at the picture you see that the little dent indicating the starting point of the IC's is on the bottom not the top of the picture. Speaking of the schematic I can't seem to get the picture uploaded properly without losing the detail. Do you happen to know where I can get it uploaded properly so its readable?



Just tried putting in some other IC's LM324AN's now the low farty hum noise has changed to a highpitched one. If i look at it with a scope it changes with the 16kHz and 125Hz control. Gonna box it up and see if it makes it go away.



Update: Unhooked the wires going to the gyrators and the noise is still there, one thing I noticed was that if i turned the trimpot up it boosted the noise so that makes me think its in the input buffer.

[Mike Burgundy]

If I am looking at the schematic shouldnt the voltage on Pins 3,5,10,12 be lower then Ureff cause of the 220k resistor between it and the pins? The Ureff is around 4,5V
One thing I noticed when measuring the pins of the in/output buffer was that it was slowly going down to zero. It starts at around 0,20V and then go's down rapidly to 0,10V where it go's slower and slower towards 0V.

There's no current through those resistors, since there's nothing hooked up at the other end. The IC's aren't there - open circuit. With no current through them, the resistors cannot drop voltage. Also, with the IC's in place, the inputs draw so little current that the voltages should be very, very near to Vref.
Do you have a digital  multimeter or an analog one? Digital ones have a very high resistance on voltage measuring, analog ones a lot less.(edit)
If the voltages are dropping, you have a tiny leak somewhere that puts that voltage there, measuring that voltage bleeds it away again - your meter also has some finite impedance resistance, so it will pull a minute current. If that current is larger than the one supplying the voltage in the first place, the voltage will drop. You either have a leaky cap or a tiny, tiny almost-short. Does it do this on all pins with that voltage?

I admit it has a lot of solder however using less I couldn't get the solder to flow over the perfholes to make the lanes. Ill see if I can do it with a bit less solder but using my mk1 eyeball it doesn't connect anything between them. I did gave the knife ting a go what a mess does that give all the yellow dust

It's probably better to use wire or cut off component legs to make those connections. Doing it with just solder is messy, to say the least. It also leaves the board with lost and lots of leftover flux, wich is also messy, and can lead to conduction problems where there should be none. That's the yellow stuff: flux residue.

The pins on the schematic I drew and the board match. If you look at the picture you see that the little dent indicating the starting point of the IC's is on the bottom not the top of the picture. Speaking of the schematic I can't seem to get the picture uploaded properly without losing the detail. Do you happen to know where I can get it uploaded properly so its readable?

Wasn't talking about the pins - I was talking about the IC numbers themselves (1,2 and 3) Haven't taken the time to trace the board, but by the look of it the left IC is IC1.

Just tried putting in some other IC's LM324AN's now the low farty hum noise has changed to a highpitched one. If i look at it with a scope it changes with the 16kHz and 125Hz control. Gonna box it up and see if it makes it go away.

Update: Unhooked the wires going to the gyrators and the noise is still there, one thing I noticed was that if i turned the trimpot up it boosted the noise so that makes me think its in the input buffer.

Keep the gyrators out for now and get the buffers working first. Probe the circuit, and see where problems arise.

[JRB]

There's no current through those resistors, since there's nothing hooked up at the other end. The IC's aren't there - open circuit. With no current through them, the resistors cannot drop voltage. Also, with the IC's in place, the inputs draw so little current that the voltages should be very, very near to Vref.
Do you have a digital  multimeter or an analog one? Digital ones have a very high resistance on voltage measuring, analog ones a lot less.(edit)
If the voltages are dropping, you have a tiny leak somewhere that puts that voltage there, measuring that voltage bleeds it away again - your meter also has some finite impedance resistance, so it will pull a minute current. If that current is larger than the one supplying the voltage in the first place, the voltage will drop. You either have a leaky cap or a tiny, tiny almost-short. Does it do this on all pins with that voltage?

It's a digital meter and it does it for all the pins that have 0,12V/0,11V on them.

It's probably better to use wire or cut off component legs to make those connections. Doing it with just solder is messy, to say the least. It also leaves the board with lost and lots of leftover flux, wich is also messy, and can lead to conduction problems where there should be none. That's the yellow stuff: flux residue.

Ill give that a try one day.

Wasn't talking about the pins - I was talking about the IC numbers themselves (1,2 and 3) Haven't taken the time to trace the board, but by the look of it the left IC is IC1.
Keep the gyrators out for now and get the buffers working first. Probe the circuit, and see where problems arise.

The IC's are numbered from right to left on that picture, so the on on the left is 3 the middle 2 and the right 1.
Gonna have a look with the scope to track the start of the signal.




Update:
Seems the signal gets picked up at the input wire by electromagnetic waves when the circuit is on and it disappears as soon as it turns off.

I had to resort to cutting the image up to get it to display properly
http://i41.photobucket.com/albums/e294/jtheunknown/eqschematic1.png
http://i41.photobucket.com/albums/e294/jtheunknown/eqschematic2.png
http://i41.photobucket.com/albums/e294/jtheunknown/eqschematic3.png

[JRB]

After some more testing I managed to narrow it down to the middle IC, IC2. Once that one is taken out it stops the noise being generated.

VICTORY resoldered the 125Hz part and the beep went away, guess it was shorting somewhere or one of the joints got screwed up.

[Mike Burgundy]

Good to hear!
Here's an example of what I what talking about using the leads to make channels instead of flooding the islands with solder to the point they're bridging the islands:
http://www.youtube.com/watch?feature=player_embedded&v=3N3ApzmyjzE (from 3:50).
It really helps keeping things clean and preventing possible bridges.
Happy soldering!

[JRB]

Well it worked and then I tried to make it more silent by changing the LM324's for TL074's funny ting is it started to make noise unfortunately one of the LM324's pins broke of when switching so I cant use that one any more.
What I don't get is that the LM324 makes more noise then the TL074 when the LM324 is a normal opamp and the TL074 is a low noise one. Could someone please explain whats going on? what parameter am I over looking in the datasheet that makes the LM324 make less noise then the TL074's?