Guitar LED circuit sucking tone

[slacker]

Get rid of the blue jumper between the 220uF cap and D1, that is connecting the output straight to ground so it won't do anything. Apart from that, taking into account Gurner's comments it should work.

Pin 5 of the buffer should be connected to pin 3 to keep the unused half of the opamp stable but that won't stop it working.

[Gurner]

Get rid of the blue jumper between the 220uF cap and D1, that is connecting the output straight to ground so it won't do anything.

Good spot (& might account for why the LM386 is getting hot - all AC signal is getting shunted to ground!)

Mr Stab, try removing that jumper as per slacker's suggestion & seeing if it works, before trying my 'experiment' out.

[MrStab]

i should probably have built each circuit separately....

just before reading either of your posts, i removed the buffer-side output cap (10uF). i've removed the 220uF as well, and cut the jumper to ground just after where it was. now the two blue LED's stay on. dimly lit (no resistor yet, for some reason...). it's a start at least! i had one short which caused no voltage readings but since fixed it, no more obvious shorts from continuity test.
worth replacing the LM386? i have 3, i'm guessing the first one is toast, though. only had one IC socket spare so the TL072 is "perma-installed", but fingers crossed i dont have too bad a track record with op amps.

my only unused battery's reading 6.3 but im gonna go score some more in about an hour.

[Gurner]

Well unless you have a scope I'd be inclined to put a loudspeaker on the output of the lm386 *with* the 200uf output cap in place - until you get some AC signal out of the lm386 this cct is never gonna work (& how will you know unless you put a speaker on there?). If I were a betting man, I'd say you've likely nuked your LM386 with that AC short to ground (i.e. that jumper link you had in place....you did say it got very hot)

If you post your DC reading for your opamp pins, that'll lend a view as to whether they're working or not.

[MrStab]

i replaced the original chip after it overheated, was unsure if maybe the 2nd one had screwed up as well, though.

any ratings i should look out for, or would some generic old computer speaker suffice? i guess i just hook it up as though it were one of the LEDs, in this circumstance. i'll have a look around just now, in the meantime here're voltages taken from a fresh battery:

TL072
4.68
4.68
2.23
0
0
8.71
8.71
9.31

LM386
3.3
0
4.64
0
7.55
8.23
9.25
0


cheers!

[MrStab]

so i hooked up a little speaker (8ohm / 2W rating, tested with "ye olde battery trick") in place of the LEDS, restored the 220uF cap, and now the LM386 is getting hot again.

in an hour or so i'm gonna try removing that IC & try an audio probe to test if the buffer's working, at least

[Gurner]

TL072
4.68
4.68
2.23
0
0
8.71
8.71
9.31

LM386
3.3
0
4.64
0
7.55
8.23
9.25
0

Do you have a schem vs. just a board layout? (it hurts my eyes & brain trying to map/visualize the circuit in my head).

As a starter,  pin 3 of the tl072 aint right...it should be about half your supply voltage ....are you sure you measured it right? If so, I'd be inclined to desolder your tl072 & see if that voltage on pin 3 rights itself (if so, your tl072 is probably bust). In turn pin 1 & 2 should replicate whatever DC is on pin 3...so something wrong there too (although it is sitting near 1/2 VCC, which makes me wonder if your pin 3 reading is right?)

[MrStab]

okay, the whole thing is screwing up now. but that's part of the "fun".... lol

observations!

-the TL072 voltages had become way weirder, 1.42 on pins 1 & 2 and 0.98 on pin 3, so i replaced it to no avail. could a failure on the other IC be affecting readings on this one? i only have one LM386 left!
-the speaker being connected made the LM386 overheat, but when i hooked up the LEDs again instead, it seems fine.

Edit: sorry, Gurner - forgot about the schematic. i don't have one per se, but for all intents & purposes it's just meant to be a combination of this basic buffer: http://www.muzique.com/images/buff8.gif (w/ 2.2 instead of 1M for divider resistor values), and this artistic masterpiece i found on the web: http://img.photobucket.com/albums/v17/un_zipped/guitopam.jpg (except for the speaker). output from the former into input of the latter, all common grounds  etc.

[Gurner]

okay, the whole thing is screwing up now. but that's part of the "fun".... lol

observations!

-the TL072 voltages had become way weirder, 1.42 on pins 1 & 2 and 0.98 on pin 3, so i replaced it to no avail. could a failure on the other IC be affecting readings on this one? i only have one LM386 left!
-the speaker being connected made the LM386 overheat, but when i hooked up the LEDs again instead, it seems fine.

Edit: sorry, Gurner - forgot about the schematic. i don't have one per se, but for all intents & purposes it's just meant to be a combination of this basic buffer: http://www.muzique.com/images/buff8.gif (w/ 2.2 instead of 1M for divider resistor values), and this artistic masterpiece i found on the web: http://img.photobucket.com/albums/v17/un_zipped/guitopam.jpg (except for the speaker). output from the former into input of the latter, all common grounds  etc.

I know your  tl072 is soldered in, but still IMHO that puppy needs to come out (or at least the pin 3 leg pulled out) to test the dc voltage on pin 3 with the tl072 removed ....you need to see the junction of the two 2.2M resistors be at half your supply ....this ain't gonna work until you have have about 4.5V on pin3, 4.5V on pin 1 & pin2  ....something is dragging the voltage on pin 3 down - easiest way to get to the bottom of it, is to remove the tl072 & see what you got then....

[MrStab]

when i replaced the TL072, i put an IC socket in, so it being soldered in is no longer a problem. i've just removed pin 3, bent it upwards and nothing's touching it, but it still reads 0.98 and pins 1 & 2 are 1.4. ...aaaand i've just snapped off pin 3. damnit. lol

a huge clue as you suggest is the fact that the voltage isn't being halved.  i'll have another wee look & try to get a photo(s) maybe

[Gurner]

to be clear....you are measuring the vacated pin 3 hole (not pin 3 on the opamp itself)...  re why you aren't getting 4.5V at pin 3, well there's not a lot of components involved so it ought to be easy to find. ....you could try unsoldering the cap feeding to the junction of your two 2.2M resistors.

[MrStab]

my bad - just tried that and got a clear 0 on that pin socket :S . i accidentally broke the 3rd leg off another opamp in the process, that's commitment! lol

tried disconnecting the 100nF input cap, if that's what you meant - while the voltages did seem to fluctuate higher for a moment, pins 1 & 2 settled back down at the same 1.4-ish. i replaced the 10uF electrolytic which comes after that with a jumper early-on, unless you meant that one.

you're right, this is probably a lot easier than i'm making it... here's a pic anyway! some sketchy hookup wiring, but i'll fix that at some point... (still no shorts obvious to me or a DMM)

[Gurner]

I'm now not following what you're saying/doing ...but now that I can see your board - just remove the tl072 totally....measure the voltage on pin3 of the (empty) IC socket - do not proceed or put the tl072 back in place until that pin 3 on the IC socket measures approx 4.5V!

If you are now saying pin3 is 0V, then that isn't right either. There's only two resistors involved....one resistor is connected to 9V, the other to 0V ...they meet in the middle where there should be 4.5V....but you are saying there's now 0V? (or are you saying there's something else?!)

[MrStab]

my bad, i'm a generally confusing person but i've been ill the past coupla days so twice the loss of brain functionality! i'm a bit slow on the uptake.

removed the TL072 altogether, pin 3 voltage is still 0.

power goes in at full voltage (8.4), measures 2v after the first 2.2M resistor, then the voltage drops to the culprit 0.98 on the jumper connecting that to pin 3. looking back on my previous readings, i don't understand how the whole thing has changed now, only thing i can think of is that 1-2 of the LM386's were fried, but could that even affect the voltages on the other side of the board?

the whole thing just makes no sense to me anymore, as simple as it is. i have no clue how the hell those voltages have changed & why i can't fix pin 3.

[Gurner]

removed the TL072 altogether, pin 3 voltage is still 0.

power goes in at full voltage (8.4), measures 2v after the first 2.2M resistor, then the voltage drops to the culprit 0.98 on the jumper connecting that to pin 3.

You're going to have to go back to basics with the pin3 thing.

With the tl072 out of its socket, disconnect the cap feeding the junction of the two 2.2M ......& also disconnect the jumper feeding the junction onwards to pin 3, so basically all you are going to have left is this....

9V
|
2.2M resistor (top)
|
---->   voltage here should be approximately 4.5V
|
2.2M resistor (bottom)
|
|
0V


if after doing the above you've still got  0V at the junction of those 2.2M resistors then you've a solder bridge or short somewhere or other.

[MrStab]

okay, done those things - voltage going into the circuit is now 4.5, voltage at the junction is about 1V. it's just being illogical now. i'm thinking i should just cut the board in half and re-do both circuits separately.

[MrStab]

cutting power to the LM386 brought the input back up to full voltage. didn't even touch anything near that side of the circuit. dunno if my mind's just totally fuzzy lately, there's one ultra-tiny solder bridge i just can't find, or if merging these circuits in the way i've done is just not feasible.

[Gurner]

okay, done those things - voltage going into the circuit is now 4.5, voltage at the junction is about 1V. it's just being illogical now. i'm thinking i should just cut the board in half and re-do both circuits separately.

If you've the circuit as outlined in my last post yet still are seeing just 1V at the junction, it can only mean three things...

1. You resistors aren't the values you think they are (but they look fine in the photo - so I doubt it)
2. There's still something else connected to the track that those two resistors join.
3. The device (or person!) doing the measuring isn't working properly!!

Essentially you've a basic two resistor potential divider that has the wrong (unexpected) voltage in the middle ...if you are feeding it 4.5V into two equal resistors ...you should be seeing 2.25V at the junction.

edit: Just seen your latest post - it's still not clear - are you now saying that after removing the LM386 the resistor junction voltage is now reading half the supply voltage correctly? (if so see 'possible problem' No2 above)

[MrStab]

upon audio probing, the signal passes right through the buffer stage as it should, ofc that probably doesn't guarantee successful buffering. or does it..? lol

i'll try to clear things up: for some reason, the input voltage dropped to 4.5, so i cut the jumper taking the battery voltage straight to the LM386, and input voltage returned to full ~8v (due a replacement battery...). can probably ignore that part for now, though.

so, assuming full, unadulterated input voltage, pins 1 & 2 are back at 4.5 (or dying battery equivalent), and the junction and pin 3 are still 2v. i don't understand how the voltage is being halved properly elsewhere but not pin 3 or at the junction itself. cannot find any shorts at all! even cutting between strips with a stanley knife for the sake of it. *sigh*

[Gurner]

upon audio probing, the signal passes right through the buffer stage as it should, ofc that probably doesn't guarantee successful buffering. or does it..? lol

i'll try to clear things up: for some reason, the input voltage dropped to 4.5, so i cut the jumper taking the battery voltage straight to the LM386, and input voltage returned to full ~8v (due a replacement battery...). can probably ignore that part for now, though.

so, assuming full, unadulterated input voltage, pins 1 & 2 are back at 4.5 (or dying battery equivalent), and the junction and pin 3 are still 2v. i don't understand how the voltage is being halved properly elsewhere but not pin 3 or at the junction itself. cannot find any shorts at all! even cutting between strips with a stanley knife for the sake of it. *sigh*

You are overcomplicating...forget about the LM386 for now (take it out of cct too!)  - your input tl072 buffer will not work unless it's biased properly.

That's the job of the 2 x 2.2M resistors, but clearly you have a problem (refer to possible reasons in my last post) ...there should be half whatever your supply voltage is at the junction of the two resistors....if there's not, you've a problem & until you sort it, I wouldn 't waste anytime/energy trying to second guess whether your LM386 is affecting the circuit or not ... just like a house, you have to build up from solid foundations, & until you get the TL072 working...your house is shakey!

There's only two components (resistors) in play here - it really doesn't get any simpler! (actually, ok I relent....a one component circuit would be simpler  !)