Can you help me fix my SD1?

[wsal]

Hi there,
I've got a Boss SD1 with some Monte Allums modding (the GT kit from memory) that has recently stopped working after being fine for many years. It's displaying the following symptoms:
1. No overdrive - the drive knob has no effect on the sound, regardless of its position - it just stays clean
2. Tone knob has no effect
3. Volume control still works, but oddly, it works in both BYPASSED and ON positions. Seems to cut the sound out completely for the first bit of rotation, and then clean sound appears for the rest of its travel

My knowledge of circuitry is too rudimentary to know how to diagnose this but I'm guessing that as the volume control is working when both bypassed and on, perhaps something's gone wrong in the switching circuitry?

Here's a schematic: http://www.hobby-hour.com/electronics/s/schematics/boss-sd1-super-overdrive.gif
or this one laid out differently: http://www.freeinfosociety.com/electronics/schematics/audio/pictures/bosssd1.gif

Any ideas?
Cheers
Will

[GibsonGM]

Hi Will, welcome!  Please go here and read this, then try to post as much info as you can into your topic here:  http://www.diystompboxes.com/smfforum/index.php?topic=29816.0

There isn't enough information without the pedal in front of us to know what's going on without more details     Your symptoms suggest that something is messed up in the switching, yes.   I'm thinking that something may have gone bad in the 'mods', and any number of things could have happened.     Try to post that info, we'll have a look!

[wsal]

Cheers Gibson,
Alright I've had a go at the transistors and diodes to start off with - don't have time or patience to start removing leads of resistors yet!
The mods to the pedal included replacing some caps with presumably higher grade ones, swapping D6 for a 2N3440 Transistor with a leg removed (?), making R6 a switchable value, and D4 switchable in/out.

Here goes (I hope I've labelled my transistor pins correctly!)

Q1.
S/1 = 0
G/2 = 22.2mV when OFF, Bouncing all over the place when ON
D/3 = o

Q2.
S/1 = ~0.8mV
G/2 = 46mV when ON, Bouncing all over the place when OFF
D/3 = ~0.8mV

Q3.
E/1 = 0
C/2 = 55mV
B/3 = 8.5mV when OFF, 0.646V when ON

Q4.
E/1 = 0
C/2 = 23.7mV
B/3 = 0.619V when OFF, 19.7mV when ON

Q5.
E/1 = 0
C/2 = 8.61V
B/3 = 11.6mV

Q6.
E/1 = 0
C/2 = 8.65V
B/3 = 11.7mV

IC.
1 = 8.03V
2 = 8.06V
3 = 8.07V
4 = 0
5 = 28mV
6 = 8.1V
7 = 8.03V
8 = 8.65V

D1
A = 0
K = 8.63V

D2 (Zener)
A = 4.23V
K = 8.68V

D3 - just a jumper lead

D4 (Zener)
A = 8.07V
K = 8.16V

D5 - Part of the mod allows this diode to be switched out, when switched in it reads:
A = 8.09V
K = 7.81V

D6 - Another part of the mod means that this has been replaced by a 2N3440 Transistor with one of the legs cut off (?)
A = 7.77V
K = 8.20V

D7 - Zener
A = Bouncing around (like the transistors) when ON, 11-13mV when OFF
K = 5.27V when ON, 23.6mV when OFF

D8 - Zener
A = Bouncing around (like transistors) when OFF, 19mV when ON
K = 5.26V when OFF, 54mV when ON

[slacker]

If you look at the second schematic you linked to you'll see a load of points marked +4.5V, these should all be about 4.5 Volts. On your measurements they are all much lower than that, Ie: pin 5 of the IC and the bases of Q5 and Q6.
The problem could be around the voltage divider that creates the 4.5 Volts, this is R18 and R19 on the first schematic, measure the voltage where R18 and R19 join it should be around 4.5 Volts, if it isn't check the connections and solder joints to those resistors and to C11.

[MrStab]

is it possible Q2 is permanently jammed On for some reason?

edit: didn't realise Q2 was lacking the bias voltage on Drain & Source

[PRR]

> check the connections and solder joints to those resistors and to C11.

+1.

I'd also consider replacing C11 on general suspicion. The "bounce" makes me wonder if it is not suppressing sneak-through among the several stages.

[wsal]

Thanks for all the quick suggestions!
So I replaced C11, and redid the solder joints connecting to R18 & R19, and they're still not showing 4.5V at the junction. The voltage does go up and down with the level control though - from about 0.4mV with level all the way down, to 1.57V with it all the way up.

Does this help?

[armdnrdy]

Post pictures of the component and trace sides of the board.

Look for solder bridges.

[MrStab]

The voltage does go up and down with the level control though - from about 0.4mV with level all the way down, to 1.57V with it all the way up.

maybe still no fixed-resistor bias to the JFETs, and the volume pot output is the only seepage of bias voltage getting there

[Groovenut]

Just an idea... Disconnect R2 and check the bias voltage, if the bias voltage is normal then the input buffer is pulling the bias down. If the bias voltage is still low, disconnect R4 and check again. Sounds to me like somethings pulling down the bias voltage causing the switching (& everything else) to misbehave.

[wsal]

Ok, I tried removing R2 and R4 and neither produced the desired bias voltage.
Here's the photos:
Please note the mods involved some replaced capacitors, C6 removed, D5 replaced with a switch to add/remove it from the circuit, and R6 replaced with a switch with 2 different resistor options (2.2k, 4.7k)

[MrStab]

have you by any chance used an AC adapter or over-volted the pedal or something? because it's only just ocurred to me (i'm useless) that all but the ground pin on your IC are near, but not quite at, the supply voltage. so not shorted to it, but i've had dead opamps show signs like that. Q5's emitter is only reading a few mV so doesn't seem to be from that.

it's either one unified problem or widespread damage. i'm having difficulty finding a common issue between the switching failure and the possible opamp failure.

[Groovenut]

I suppose the opamp could be dead from overvoltage/polarity and is semi-shorted, pulling the bias rail down?

[MrStab]

^ just seems to be one of the common signs of opamp death for me, either pins hanging at just under the supply voltage (ruling out shorts to V+ outside the chip) or at 1.x volts. whether or not this is the cause of the problem, i can't tell. but there's a fair chance the chip is fried anyway. if you've exhausted every other possibility, Will, i'd consider removing the 4558 chip and checking the bias voltages again (before replacing it).

when debugging i inferred from the OP that it just stopped for no apparent reason, though. was there any event that might've caused it, Will? have your power supply or other pedals been acting weird at all? new pedals? spillage? sauna? sketchy ninja friend?

[wsal]

Alrighty, opamp removal is next!

I can't recall any event to have caused the problem - it has always been on my pedalboard, powered by a One Spot daisy-chain arrangement. I leave it in theatre pits, and often the techs will switch off the mains power overnight and back on again - I don't suppose that could have cooked it? None of my others have suffered the same fate though...

I think I've got some spare 4558D's around - is there any way to test that an opamp is good, before I install?

I really appreciate everyone's efforts so far!

[Groovenut]

check and post voltages with the opamp removed. That'll help determine if thats really the issue

[duck_arse]

there's no chance that the giant metal case of the leaning tower of mod is shorting to one of the nearby connections or leads, is there. is there?

[MrStab]

I think I've got some spare 4558D's around - is there any way to test that an opamp is good, before I install?

use an IC socket. unless your spare 4558s have been salvaged from another device, they're most-likely good to go, but a socket would be invaluable in the off-chance they're not. the only way i can think of testing em is in a known-good circuit (with a socket).

[wsal]

Fixed! And it looks like duck_arse was on the money - pulling it all completely out of the chassis got the bias voltage back to where it belonged. Lots of extra heatshrink and insulating tape later and it's all working again - with a bonus IC socket so I can experiment with different opamps!
Thanks for all the help, and sorry the solution was so mundane!

[duck_arse]

well done, and an answer to your original question.