Repairing power-fried pedals

[R.G.]

We've had several of these come up here in recent weeks, so here's a guide to getting a pedal running when you suspect - or know - that the damage was caused by a wrong power supply plugged in.

Before we can repair the damage, we need to understand the damage. Power-fried pedals come in three varieties: over voltage, reverse voltage, and AC damaged.

Over voltage is simple. An adapter with too high a voltage was plugged in. No polarity protection will stop this damage. An on-board power regulator will, but no commercial effects and almost no DIY effects have that. Simply put, everything on the board that has a voltage rating is likely to have died. If you're working with surgical precision, you'll get on the internet and look up all the chip and transistor datasheets to find out their maximum power supply voltages and find out which ones were likely to have been exceeded. If you're doing a hunt-and-replace repair, you'll just replace things in the following order:
Most likely to have died - electrolytic caps. Some commercial effects use 10V caps. The standard voltage ratings for electros are 6.3V, 10V, 16V, 25V, 35V, 50V and so on. If the caps were 10V, even mild overvoltages may have killed every single one. If they were 16V, they will survive an 18V supply for a while. If you don't know what overvoltage the pedal got, replace them all.
Next most likely to die is opamps. Generic opamps usually will work from a +/- 18V power supply, so they're good up to 36V. There's a big "but" here, though. Later opamps have become ever lower voltage. Many modern opamps will withstand a total of 12V, 10V, or even only 5V. Either look up the datasheet or replace it.
Discrete transistors are pretty durable. There are few discrete devices in through-hole packages that will not withstand 30-40V, at least for a while.
Likewise, diodes are pretty durable. The 1N4001 is a 50V diode, the 1N4148 is a 100V diode. Generally these will survive. Zeners will survive if the power dissipation does not simply burn them up thermally.
Ordinary resistors will survive as long as they are not thermally burned out. Likewise non-polar caps.

Reverse voltage survival is different. If the pedal has no polarity protection, reverse polarity is a death sentence for every IC on the board. This happens in milliseconds to a few seconds. Longer reverses will kill all of the electrolytics. Often the diodes and transistors will survive this. Sometimes they will be damaged, often not. Ordinary R's and C's usually survive. If the pedal has polarity protection, everything depends on what kind and how good the protection is. The common reversed-diode protection is really only good for short periods of reverse connection to a battery or adapter. If you leave a reverse-diode protection pedal connected to a strong battery or AC adapter for a long time, it will eventually burn out the diode or other protection parts. Then you have no protection, and you're back to the first situation above, replacing ICs and electrolytics.

The worst thing you can do to a pedal short of melting it down in a pot of liquid iron or pounding it to bits with a sledge hammer is to connect up a power adapter that provides AC instead of DC. There are several commercial pedals which come with their own adapters that supply 9VAC, not DC. The adapters have the same power adapter plug as other DC pedals, so it's easy to plug them into a DC-only pedal. What this does is to first destroy any reverse-diode protection circuits. Then it kills the ICs, then the electro caps. After that, it may kill transistors and other parts.

If you don't know what killed the pedal, then first replace any obviously fried parts. Replace the reverse-protection diode, and any electrolytics with bulges or leaks. Power it up on a battery and read the DC voltages on the ICs. If the voltages are not within expected tolerances, especially if they are all almost 0, start replacing all the electro caps. Still not working? Start replacing ICs. Still not working? Start replacing transistors and diodes.

[BDuguay]

Thanks for this post R.G.
In my repair experience, this is the most common occurrence. To have it all explained the way you have is extremely helpful.
B.

[petemoore]

  Thanks RG, exellent reference post !
  ...the next best thing to having a world that uses non-compatible connectors for non-compatible power supplies. dohh...More reliable / Less expensive, I figured that one out way long ago before I even knew anything...lol,  as brilliant as some of these designs are, the boat was missed on this one.
  AFAIK, about everything else 'electronic for consumer' is PS dummy proofed.

[marlin]

The worst thing you can do to a pedal short of melting it down in a pot of liquid iron or pounding it to bits with a sledge hammer is to connect up a power adapter that provides AC instead of DC. There are several commercial pedals which come with their own adapters that supply 9VAC, not DC. The adapters have the same power adapter plug as other DC pedals, so it's easy to plug them into a DC-only pedal. What this does is to first destroy any reverse-diode protection circuits. Then it kills the ICs, then the electro caps. After that, it may kill transistors and other parts.

I did that to my Orange Squeezer   
20V AC into it and i had to change the capacitors and IC. If i remember correctly i didn't have to change the diode though.

[analogmike]

There are several commercial pedals which come with their own adapters that supply 9VAC, not DC. The adapters have the same power adapter plug as other DC pedals, so it's easy to plug them into a DC-only pedal. What this does is to first destroy any reverse-diode protection circuits. Then it kills the ICs, then the electro caps. After that, it may kill transistors and other parts.

Yes this is getting to be an epidemic lately... we get so many pedals back with blown up chips and melted protection diodes. A few of them even admit it was their fault. Be careful!

[petemoore]

A few of them even admit it was their fault.
  'Fault...as in, it could happen to anyone, including me'.
  I have a DC + pin supply connector, use battery for pos gnd., and the rest are - pin DC connectors, except the AC supply for the Echo Park, sound confusing...lol.
  White surgical tape on the + pin plug, the EP's is small [I should mark it better I guess...I'd sure like it better for sure if it we're fault proofed instead]..pretty dern easy to make a miscon and fry stuff when everything else is going on..dark and rushed, about to be dead on the spot.
  and of course.. 'the dominoe daisy chain effect'.

[Mark Hammer]

There are several commercial pedals which come with their own adapters that supply 9VAC, not DC. The adapters have the same power adapter plug as other DC pedals, so it's easy to plug them into a DC-only pedal. What this does is to first destroy any reverse-diode protection circuits. Then it kills the ICs, then the electro caps. After that, it may kill transistors and other parts.

Yes this is getting to be an epidemic lately... we get so many pedals back with blown up chips and melted protection diodes. A few of them even admit it was their fault. Be careful!

Here's the thing.  Pedal functioning assumes that outside-positive is best (you need it to switch between battery and external power), but human safety assumes that outside-ground is best (you can touch something grounded and the outside of the plug with impunity...more or less..under those conditions).  Since not everything operates from batteries OR wallwart, the need to have outside-positive is not everyone's need.  Particularly given the human safety consideration.  Laid over top of this is the pedal+human safety consideration that makes barrel plugs better than phone  plugs.

So we have this "situation" whereby there is every reason to have multiple types of adaptors all using the same plug/jack type, but no universal reason for everyone to use the same polarity.  Said it before, and I'll say it again:  Manufacturers would be more than delighted to NOT have to produce and sell adaptors but they have so little assurance that end-users will plug the right type in, that they have no choice but to either make their own brand so as to assure the right polarity and current source, or to have a higher retail price and generous return policy for people who keep blowing them up because they don't understand pedal-powering well enough.  Obviously the one choice is far more pragmatic than the other.

BUT, life would be SO much easier if outside-positive adaptors had a different plug or even different coloured plug than outside-negative ones, and if AC wallwarts had a different-sized or coloured plug than DC ones.  Steve D and I discussed this several years ago when he was whipping up the Small Wart project.  We both liked the idea of different coloured plugs for power patch cables serving different functions or carrying different voltages.

[calculating_infinity]

Great info here guys!  I like the idea of color coding the plugs, too bad they couldnt standardize a color coding system for all wall warts.  Thanks for sharing!

-Jonathan

[Processaurus]

hmm, about the AC death...  does that kill shunt protection diodes, or series protection diodes?  Wouldn't a series diode just half wave rectify the power?

[Meanderthal]

 Ya know, funny thing... I've been using/messing around with with pedals for... a long time. Decades. I've made some really stupid mistakes over the years with my gear in... many ways. For some reason, I've been very lucky with this- it never happened to me. Yet. A couple close calls("Why ain't it workin'? Ohhhh... DOH), but nothing fried this way...

But, if it does...

Thanks R.G., this is great!

[sfr]

What would happen if one plugged a DC adaptor into an AC pedal?

[Paul Perry (Frostwave)]

What would happen if one plugged a DC adaptor into an AC pedal?

Usually, it's OK. It won't work, but ther eis no smoke, either.
I had a d'oh! moment once, I had a MIDI to CV converter that ran on 10V AC, one side of the AC to ground, the other to two half wave rectifiers for the + and -. Inclufding a +5 reg for the digital part of the circuit...
yeah, plugged a +9 DC supply in...... and the polarity was the "right' way around..... and so the digital side worked just fine. Took a while to work out why the analog side wasn't doing too good

[petemoore]

color coding the plugs,
  Still room for misconnects besides the dark room or the drunk guy helping.
  If all you have is a blue thing, and a red thing that match physically, chances are greatly improved that the blue thing will become connected somehow to the red thing, making the blue thing can not possibly be plugged into the red thing because of a physical incompatibility is the best bet I think, a larger diameter, perhaps squares/triangles.
  For SB's there are quite a few supply types..
  3vdc
  1.5vdc
  9vdc
  18vdc
  Reverse polarity types for some of those
  Then add the AC types...
  Now I want to recant...
  I like standard stuff.
  I had some trouble with 2.5mm vs 2.1mm incompatibility.
  Having 20 different types of jacks and plugs to choose from when a jack and plug that match are needed brings it full circle...user requirements, standard or specialized connectors, color coding.
 

[R.G.]

hmm, about the AC death...  does that kill shunt protection diodes, or series protection diodes?  Wouldn't a series diode just half wave rectify the power?

Yes. I should have been clearer. AC supplies kill reverse-shunt protection diodes. They have no effect on series protection diodes, other than the diodes half wave rectifying the AC. But I can mentally count only about a half dozen pedals I've ever seen with a series diode.

That's the thing that makes series protection inherently better - it doesn't care that the impinging voltage is wrong AC or wrong DC. It just opens up and nothing happens.

There's actually a variant of the MOSFET protection that makes an active bridge out of MOSFETs. This was suggested to me a ways back, and it seems like a good idea. You can hook up a "rectifier bridge" out of two N and two P channel MOSFETs. This can be arranged so that both polarities of applied input voltage are rectified to the correct one. What makes this good is that instead of losing two series diode drops as you would with ordinary diodes, this arrangement could have a voltage loss of under 0.1V, maybe less. It was an expensive device back when TO-92 MOSFETs were $1.00, but now that they're nosing around $0.30, it's getting more practical.

[petemoore]

  You can hook up a "rectifier bridge" out of two N and two P channel MOSFETs. This can be arranged so that both polarities of applied input voltage are rectified to the correct one. What makes this good is that instead of losing two series diode drops as you would with ordinary diodes, this arrangement could have a voltage loss of under 0.1V, maybe less.
  maybe...'maybe less' also means 'maybe more'.

[Processaurus]

Thanks for clearing that up RG, I've been sold on series diode protection since reading the article at Geo, and generally using power supplies, not needing the more refined methods that drop less voltage.  The power schottkey diodes from small bear have been working out well for that.  The active bridge rectifier sounds like it would be a very good way for a higher end pedal maker to go, though, even if it were more parts than their actual design, because it helps reliability (and sellability on the reliability)  the same way heavy boxes and overspecced pots and jacks do.  You could even have a couple different power jacks on it so it could just whatever were around.

  Still room for misconnects besides the dark room or the drunk guy helping.
 

Equipment makers should absolutely have helpful non electronics oriented drunks in mind when designing their gear.

[petemoore]

  And how sturdy would the n and p mos bridge rectifier power protection be?

[Paul Perry (Frostwave)]

  I had some trouble with 2.5mm vs 2.1mm incompatibility.
 

Absolutely! the worst thing is, a 2.1 and 2.5 plug have the same outer diameter, it's the diameter of the inner conductor that the measurement refers to. So you can put a 2.5 plug in a 2.1 socket, and the inner pin in the socket can intermittently  make contact - or not at all.

[R.G.]

maybe...'maybe less' also means 'maybe more'.

I've taught you well about datasheets, Grasshopper. But in this case it really only means "or less". My experiments with a single MOSFET showed under 50mV per MOSFET at typical pedal currents for the BS170 and BS250P. The only currents will be the pedal current usage, which is usually small - for normal FX values of pedal current usage.

And how sturdy would the n and p mos bridge rectifier power protection be?

Pretty sturdy. It depends on how big the MOSFETs you buy are. The breakdown voltage which the bridge can't handle is something like twice the Vdsmax rating of the MOSFETs, and the current rating is the larger of the body diode current rating and the channel current rating. The cheapo BS170/BS250P devices are 40V devices, so you get voltage imperviousness up to nearly 80V.

I say "nearly" because the Grasshopper will sense the indeterminacy of that. 

The actual voltage rating would be 2x the drain voltage rating if the reverse leakage of the devices was exactly equal. That's not likely, so the real  number will be something less. It can't be less than the voltage rating of one device, so between 40 and 80, depending on N and P matching. It's likely to be up at 70-80V.

Notice that the voltage rating is what matters here, as the forward currents are just the pedal forward currents, well under the half- to one-amp rating of TO-92 devices. If you REALLY want voltage protection, both Supertex and Zetex make TO-92 package devices with Vdsmax ratings of over 500V. These cost about $1+ each. But they would rectify an AC line voltage happily up to their current and dissipation ratings.

Which brings up the issue - Polarity protection is not equal to over voltage protection. Both single and bridge MOSFET polarity protectors happily let through correct-polarity power that will fry the pedal.

[bwanasonic]

Great info here guys!  I like the idea of color coding the plugs, too bad they couldnt standardize a color coding system for all wall warts.

This isn't foolproof. I had my adapter plugs color coded, but still managed to have a lapse while hooking up my pedalboard, and plugged my Echo Park adapter into the daisy chain for my board. The only thing that got fried was my Nobels ODR1.

Kerry M