OT Need some good problem finding minds!

alband

Senior Member
Hi there,

This is a bit of a cop out of a post really because I'm going to link to another thread on another forum where another guy discribes the problem more thoroughly. Brief description though:

A particular guitar effects pedal "Digitech Whammy 4" (WH4) is used to bend the pitch of the input note up and down (various settings give a range of +2 to -6 octaves each way!). It has a rocker foot pedal to variably adjust the note (using an LDR, patterned semi-transparent film, LED sandwich to detect the angle of the pedal out of interest) and a dial (rotary encoder) to select the particular mode, and a simple PTM switch to turn it on off and a few other things. It takes a 9v AC power supply.

I took it appart (hoping at least a few of you will understant the need for this?).

I didn't actually change anything, all I did was probe at voltages with a multimeter, but somehow it went duff. Now the thing powers up when I attach the PSU (standard unit from digitech) but after a tiny flash from all the lights, it dies, leaving just the sensor LED on for the tredal.

The first post in this thread has a ciruit diagram attach that I cannot pick fault with. My case seems to be most similar to the very last post on the 2nd page but with similar voltage levels at the regulators to the first post on the 1st page. None of my diodes seem to have the expected 0.6v drop on them though. Some are very hot (burn your finger hot, but not combust spontaniously hot), one of the caps is quite hot, as are a couple of the regulators and (worryingly) the main CPU chip.

The question
I'm not good enough at electronics to understand that power supply system. The furthest I get is that it appears to be a clever way of getting 3 voltage levels out of one power supply. Hopefully though, some of the more developed brains on this forum will be able to scan that diagram and think "ah yes, so it works like that". Does anyone of those minds have any idea what could be the problem that keeps cropping up with these WH4 pedals?

Any help would be appriciated SO much and I can give exact details of any components in my case, eg voltage levels temperature, marking numbers (I've managed to remove the heat sinks, after the problem deceloped), however this seems to be a more broader problem with only a few common symptoms.

Did I meantion, these pedals are >£100 USED on ebay :O Hence, not wanting to just get another.
Like I say, any help = FANTASTIC
Cheers,

David.

PS, as a side note, how easy would it be to get a -9v, +3.3v and +5v power supply running from a 9v DC supply? Even possible?
Cheers
 

westaust55

Moderator
PS, as a side note, how easy would it be to get a -9v, +3.3v and +5v power supply running from a 9v DC supply? Even possible?
Cheers
to achieve +5 V and +3.3 V from +9V is very do-able. It depends what is the current draw and how efficeint you wish the power supply to be (ie whether to use linear or switch-mode power supplies)

For the -9 V what current loading is required?
There are simple charge pump IC's but they are typically < 1mA capacity (eg TL77660 or MAX1044).

More information is needed before folks can provide clear/firm answers
 

srnet

Senior Member
Power supply setup looks standard enough, if a little crude.

I would isolate the problem by disconnecting the power board from the rest of the circuit, and then verify that the output voltages of the regulators are what they should be ......
 
Last edited:

alband

Senior Member
Isolating wise, presumably, just cutting through traces is good enough, then I can bridge them all back up later after testing? Attached a picture with propsed cutting "route", also extra info.

As for the current draw, I would have thought it would be more than 1mA, but I'll check the datasheets of all the big chips:

Running at 3.3v is a 24-Bit Audio Digital Signal Processor. This seems to be running at 10mA but for the life of me I never know which of those ratings is which (bottom of p27 seems to be where the table of maximum ratings starts).

The next biggest chip HAD a sticker on it saying "WHammy4 V1.30". Sticker off, it appears to be an 8-bit Microcontroller with 20K Bytes Flash. This thing seems to run at 5v and *I think* draws 15mA (p26 but again, I know not which numberis the one I want).

Next we have an Octal D-type flip-flop. This seems to be mostly to do with driving the 16 or so LED's used to indicate which mode is in use, or helping with the rotary encoder. Again, a 5v character, I think running a 70mA (needs checking).

Then CS4224. Runs at 5V, I think running at 10mA (p5).

Triple 2-channel analog multiplexer/demultiplexer. Runs at 5v or 3.3v, either at 20mA or 50mA (p5).

Quad 2-input AND gate. 3.3v or 5v, 50mA I think (p5).

Then there's 4 "4580 jrc" which seem to be op amps, from google. But I cannot find a datasheet, which is annoying because and op amp is likely to want a negative supply, right? They're in little 8pin soic packages (like our SMT 08M's) and their markings are: "4580 784G JRC".

There's also a High Speed Optocoupler, 100 kBd which is for the MIDI comunication to the board. Being an optocouple it has two distinct circuits within? Neither uses a negative supply but both draw around 50mA *I think* (p2).

Given that, the only things that could be needing the negative supply are the op amps? Anyone know of a datasheet for them?

I'll get back to you on the attached picture. Need to get it off the camera along with 119 other photos.

Thanks for the replies!
 

alband

Senior Member
Ok, I've cut all the output lines and connected it to the power supply.
Btw, I've traced the traces from the -5v regulator and it leads to all 4 op amps (unsurprising) and 3 ceramic capacitors (one next to the "Triple 2-channel..." all grounded). So that's the need for the -5v. If we had a datasheet for them, we'd know the current draw.

For U1 there should be a 3.3v output. I'm reading 11.74 in and 3.27 out. So that's fine.
U2 is our good old friend the 7805. I'm reading 14.23 in and 4.99 out. Fine.
U3 (I got this wrong earlier) is a -5V regulator, not 9. It's a 7905. Reading -14.39 in and -4.98 out. Fine.

So the power supply section of the board is working no problem! This is BAD though, because that's the eaasy bit to fix. Now what should I do? Bridge gaps until sometihng doesn't work and see what it was that did it?
 

vttom

Senior Member
The next thing I would try would be to re-connect each supply with an ammeter in-line to see how much current each one draws. That might help you narrow in on which voltage domain has a problem.

Also, have you examined any electrolytic capacitors on the board? Do they look puffy? If so, that might be your problem. There's been a lot of problems in the last decade with manufacturers using sub-standard capacitors that fail when the electrolyte chemical evaporates and the capacitor plates short out.
 

alband

Senior Member
I bridged the gaps I'd made one by one. started with the 7805, then checked the outputs, everything was normal. When I plugged it in, the sensor LED stayed on and the other "mode LED's" flashed, as is symptomatic with the initial problem.

Next, the 7905. Checked the pins, all was as before; same GOOD voltages at each pin, same light show.

Next the LD33v. :/ :( That's the one. Connected it up, got the same light show and voltages from the other regulators. However, at the LD33v, there was only about 5V going in and something like 1.9V going out. And the regulator was quite hot, diodes were starting to heat up, and the main big chip (first one listed in post 3) was hot.
The 5V going INTO the regulator seemed odd to me, given that it was a result of a bridge being made AFTER the regulator. Could it be due to very high current draw? Or does it just sound like the big audio processor chip is fried? :(

I've attached pictures that hopefully can help illustrate the set-up.

Cheers.DSC04654.JPGDSC04653.JPG

Edit: can do much bigger and different pictures on request but if bigger, I'd have to PM them
 

alband

Senior Member
I'll have a hunt, the ammeter idea sounds good and I can try that too. My bet is a large current draw on the 3.3v regulator output! I'll have a look for the puffy caps but I don't think there are any. I'll get back to you in a while. Need to walk the dog.
Thanks for the fast replies :)
 

srnet

Senior Member
The 5V going INTO the regulator seemed odd to me, given that it was a result of a bridge being made AFTER the regulator
You only have 1/2 wave rectification on the input, so it could well go as low as that under a large load .........

The load is probably not expected to be large, in which case the single diode and capacitor may keep the input volts up at the 7V or so the 5V reg needs, but only up to a certain load .....
 

1968neil

Senior Member
from experience it would almost certainly be the main DSP I/C if its getting hot then it has an internal short and the only way to prove the point is to lift the power supply pins on the chip and this is next to impossible.

I have repaired many "stomp Boxes" and this one is always the hardest as parts are non existant, sending the unit back to the makers proves expensive and sometimes they just scrap it.

The DSP I/C is VERY sensitive to stray voltages/static etc and can easily be damaged.
Whilst an excellent pedal its design is a little fragile electrically.
Sorry to be the bearer of bad news.

Regards
Neil
 

alband

Senior Member
Alas, that does seems too possible. I think it's give up time, and after such a short time too. Oh well, of to the bay to find the next pre-defunct cheapy.

Thanks everyone!
 
Top