Doubleheader - Power supply and Octavia Debug

[Gregory Kollins]

Today has been long. I have essentially done nothing but debug, all day. I succeeded in fixing a tremulous (posted about it earlier, about 5 min before solving it myself...) and a sort of self designed fuzz, and I gutted a briefcase and fitted it with velcro for use as a pedal board (as a therapeutic break from debug headaches). But I still have a malfunctioning power supply and a broken octavia.

Power supply:
I do not have a schematic. I will draw one out and post it tomorrow. 12v DC brick powering a 5 output board. Each output has a little circuit feeding it involving a lm317 and a 3904 transistor.
Problem:
each individual output works, when used alone. When a second pedal is hooked up, the lm317 connected to the output that was plugged up first gets blisteringly hot, to the point where touching the heat sink will burn you if it is left on too long. A 37ohm resistor that is between the 3904 emitter and the positive output of the 9v output plugged up second will also get immediately and intensely hot, melting it's own solder joints (!!!).
So, I am quite befuddled, but interested on account of the high temperatures. I have never worked with a power supply before, so I am admittedly a little timid with the thing. Any ideas? I will get up a schem tomorrow.

Octavia: (http://fuzzcentral.ssguitar.com/octavia.php)
The 9v jack failed and shorted out. I was disappointed, because it had been salvaged from an arion distortion, then subsequently used in three separate DIY projects. The better half or the jack's plastic body was melted away. It smelled bad. But it was cool.

Anyways, it shorted out. New jack, but no octavia... sort of. It was sort of functioning, but the clipping sounded really funny a farty, etc. So thinking along the lines of diode, I tried bypassing one. But while dicking around on the board, something touched something it was not supposed to and POP, dead octavia. I just want to know what could have been fried on the thing? I don't know much about transformers, but could that have been damaged? Diode? I would just swap this stuff out but I don't stock any of it... small basement operation going on here. And I have confirmed that bypassing either or both diodes will not cause a signal to get through. There is also no visible damage, cracks in components etc.

[JKowalski]

For your power supply, it sound like you have a problem in your construction where the LM317 is somehow running far too much current through it (whether through a circuit build problem, or an internal failure). From your description of the heating it sounds like they are probably already fried by now. Troubleshoot, and don't try running it again until you figure it out, so as not to fry any other components.

Try replacing the transistors on the octavia.

[Gregory Kollins]

It's been a busy week. Sorry I did not get this up until now. I know you've been hurting to help my situation.

[tempus]

I'm not sure why you have that 3904 in there at all. Is that for current limiting? Also from your diagram, it appears you have the polarity reversed, i.e., the +12v should be going to the input (unless I'm misunderstanding the use of the 3904). Do you have the data sheet?

[Gregory Kollins]

Oh wow I drew the 12v input and 1000uf cap backwards.... 

Corrected:


Edit:
I just saw that you already pointed this out for me.

Edit 2:
There was another error... again corrected.

[tempus]

There was something else I was wondering about this (aside from the 3904 circuit) - are you sure you have 12v DC and not 12v AC? You have the DC voltage coming from a battery in your schem (which I'm assuming is not actually the case). Are you taking your voltage from a transformer, then a recitfier circuit of some kind?

[slacker]

What's the transistor supposed to be doing, I've never seen that arrangement before.

The other thing I don't understand is that your negative out isn't ground because of the 37ohm resistor, this will probably do strange things with one pedal plugged because the pedal will expect the negative connection to be ground. With more than one pedal plugged in things could get very weird.

There's probably no real need to use more than one LM317, with a suitable heatsink they can provide 1.5 amps which will be enough to power an enormous pedalboard.

[tempus]

There's probably no real need to use more than one LM317, with a suitable heatsink they can provide 1.5 amps which will be enough to power an enormous pedalboard.

That depends - there are 100mA and 500 mA versions of these in TO-92 packages.

And check out this that I stumbled across:

http://www.cpemma.co.uk/317calc.html

[slacker]

Good point, I tend to assume people are talking about the LM317T.  Even the 500ma one will power a decent sized pedal board though providing you haven't got loads of power hungry pedals on it.

[Gregory Kollins]

There was something else I was wondering about this (aside from the 3904 circuit) - are you sure you have 12v DC and not 12v AC? You have the DC voltage coming from a battery in your schem (which I'm assuming is not actually the case). Are you taking your voltage from a transformer, then a recitfier circuit of some kind?

I am using a transformer, labeled "Output:12v". It is not designated AC or DC, but when measured with my voltmeter on DC, it measures ~12.5v, while on AC it measure ~30v. I therefore assumed it was DC. It came with the power supply board.

What's the transistor supposed to be doing, I've never seen that arrangement before.

The other thing I don't understand is that your negative out isn't ground because of the 37ohm resistor, this will probably do strange things with one pedal plugged because the pedal will expect the negative connection to be ground. With more than one pedal plugged in things could get very weird.

There's probably no real need to use more than one LM317, with a suitable heatsink they can provide 1.5 amps which will be enough to power an enormous pedalboard.

I really have no clue what the transistor is doing. I just traced the thing out (and my knowledge of circuitry is... elementary). The resistor on the -OUT is odd, I'll try abridging it for the -OUT only and see what happens.

The 1 lm317 per output was also kind of throwing me off. I've never seen this sort of thing before (well, I've only ever SEEN the insides of two power supplies before and I built neither).

I may just hack apart the thing and build the GGG "ultra-clean" one. I can scrounge up a 25v transformer from a salvage store or something.

[Gregory Kollins]

I bypassed the one resistor, and everything seems to be in working order. I ran four pedals off of it for about 20 minutes, no flames or sparks, not even a noticeable temp increase anywhere on the board. All outputs are about 9 to 9.5 v. Thanks for the help! Still some confusing things in the circuitry though...