9V vs. 18V power supplies?

[erics]

Relative noob here.  I have a DC Power Brick with both 9V and 18V power outputs.  I know some pedals can run on either 9V or 18V power (e.g., a Fulltone OCD), but I've also heard that some pedals will fry with 18V power.  How do you know if a pedal can handle 18V power?  Will a pedal only pull the amount of power it can handle, or can you fry your circuit?  I just built a DIY wah, and I'd like to try it with 18V power.  Safe?  Thanks for any input.

[R.G.]

I have a DC Power Brick with both 9V and 18V power outputs.  I know some pedals can run on either 9V or 18V power (e.g., a Fulltone OCD), but I've also heard that some pedals will fry with 18V power.  How do you know if a pedal can handle 18V power?

You don't unless (1) the manufacturer tells you it will or (2) you have the skills to look at the circuit and components and determine that everything is OK at 18V.

Well, OK, there is the option of just hooking it up and seeing if it fries. If it lives, it could, and if it dies, it couldn't. But from your note that's not the test you're wanting to run, right? 

Will a pedal only pull the amount of power it can handle, or can you fry your circuit? 

There are multiple answers. If the voltage is within the pedal's safe range, it will only pull as much current (and therefore power, because power = voltage across it times current through it) as it needs. If the voltage is too high, it will break down the parts and cause them to take too much current. This is usually destructive if the power supply does not limit the amount of current it supplies.

For instance, many pedals use 10Vdc or 16Vdc rated electrolytic caps. An 18V power supply which does not have internal limiting *will* kill a 10V rated cap, and *will eventually* kill a 16V cap. Some pedals use 25V caps, and these will not fry upon applying 18V.

I just built a DIY wah, and I'd like to try it with 18V power.  Safe?

It depends on what schematic you used and the rating of the actual parts you put into it. No way to tell without that information.

[erics]

Thanks for the response.  That's very helpful. 

As a followup, if I want to check my schematic, are the capacitors the things that can get fried (I suspect not)?  I've looked at some of the capacitors, and they say things like ".01uF/630V".  Does this mean this capacitor is rated for 630 volts, or am I misunderstanding something?  This seems high. 

Thanks.

[R.G.]

As a followup, if I want to check my schematic, are the capacitors the things that can get fried (I suspect not)?  I've looked at some of the capacitors, and they say things like ".01uF/630V".  Does this mean this capacitor is rated for 630 volts, or am I misunderstanding something?  This seems high. 

The issues are almost totally with capacitors, transistors, and integrated circuits, and of these, mostly capacitors.

There are almost never inductors in pedals any more. The ones that are there are almost always in wah pedals, and this application is not voltage sensitive. This is not necessarily true in tube amps, but don't go poking around in tube amps unless you know how to not get electrocuted in there.

Resistors do have voltage limits, but they're uniformly high, 200-400V for even the little ones we use in pedals.

It used to be that transistors might have a voltage breakdown specification of as little as 10-20V. That's in general not true any more. Generally you can't find a transistor which breaks at less than 30V. You have to go look. It's still possible to overheat a 30V or more transistor if the increase to 18V tells the circuit to conduct too much current, but just the voltage won't break over the transistor. A transistor at risk from overheating might be a transistor in a voltage regulator or LED dimming setup where the extra voltage makes the current it runs overheat the transistor. It's a low-probability event, but you'd have to know the circuit to know for sure it would or wouldn't.

Integrated circuits will simply die if you put too much voltage on them. It used to be that only linear opamps were used in pedals, and they were uniformly good to 24V up to 36V, depending on the device. However, there are now opamps which die at over 6V, and 12V. Most of the ones in pedals go above 24V, but you have to check the part number and look up the datasheet to be sure. Also, many more types of ICs are used in pedals now. There are lots of CMOS logic chips used, and those only go to 15V or 18V if you use the CD4000 family or 74Cxx family. The 74HCxx and other families only go to 6V before they die, and some logic families die at over 3.3 or even 2.5V. The low voltage families generally use voltage regulators (7805 or LM317 are examples) to keep even the 9V from killing them, but 18V instead of 9V can in some cases overheat the voltage regulators or make them shut down in protection mode. Any pedal with a microcontroller in it will have a voltage regulator because there are remarkably few controllers that work at over 5Vdc.

And now capacitors. Generally, film-type capacitors have breakover voltages of 50V or greater. No problem there. Often they are as high as several hundred volts. Those work fine in the lower voltage power supplies of pedals. Ceramic caps are likewise, usually 50V or higher, no problem with 18V. Electrolytic caps are the ones to watch. They are specialized for high capacitance at lower voltage, and may be rated as little as 6.3V. The common ratings are 6.3V, 10V, 16V, 25V (and then higher). 6.3v and 10v *will* usually die if you put 18V across them unless the current is externally limited by a big-value resistor. Even then it's not good for them. 16V may or may not live at 18V, and may die slowly over time. 25V will live their full working life at 18V. The big-name pedal makers like Boss and Ibanez have and continue to use voltages as low as 6.3V at places in their circuits. Will these live on 18V? Somewhere between probably not and maybe, depending on the rest of the circuit around them. For boutique pedals, it's all over the map, from "sure, all ours are always 25V or more" to "Hmmm... don't remember exactly. Hey, Jimmy, what voltage caps did we buy last month?"

[PRR]

> Does this mean this capacitor is rated for 630 volts

Yes.

Cap voltage is determined by insulator thickness. Thinner is more capacitance but a lower voltage breakdown.

But some insulators can't be made very thin. Or there is little point: thin stuff may cost as much as thicker stuff, and the thick hi-volt stuff will work in low-volt circuits.

Until recently, Mica was not shaved thinner than 500V. The ongoing micro-miniaturization fad has brought 50V Mica caps, but a 10V Mica is still silly.

You can get 0.01u in 50V, smaller and cheaper than 0.01u 630V. But the difference is small. I've used lots of 0.01u 250V caps in 15V work. If I did more tube-work I'd stock-up on 630V to cover anything from 400V down to zero.

And as R.G. says, "Electrolytic caps are the ones to watch." In these the insulator can be made ANY thickness. By making it hyper-thin we can get HUGE capacitance in a small cheap can, at LOW voltage. A 47u 16V electrolytic is probably no bigger than your 0.01u 630V, and somewhat cheaper, despite giving thousands of times more capacitance. (There's several reasons the designer did not use an electrolytic here: didn't need that much, or could not stand electrolytic leakage, or felt the choosen insulator had "magic mojo".)

A 16V cap with full 18V will "work", for a while, with maybe some leakage. In cool occasional use, it may "work forever": I have "35V" caps run at 36V for decades. But at some overvoltage there will be trouble. First excess leakage, then excess heat, and if enough power is available, a small (or large) POW! and soggy paper all over the inside of the box (best not to have your face in there). Or more often: the box just stops working right and it isn't clear why.

But in general..... it is the same as asking if a car engine can be over-revved. (Assume you defeat the rev-limiter many engines now have.) There's many ways it can go. Even if you know some stuff.

It is very much the same. Electric heat stress tends to go as Square of voltage (9V to 18V, FOUR times the heat). Engine crank and con-rod stress goes as Square of RPM: from 4000RPM to 8000RPM, piston-tug on rod and crank rises from 50,000 pounds to 200,000 pounds.

My Ford 351W engine was rated max-Power at 3600RPM. Typically an engine is not built to turn much faster than its max-power point.... you don't get any more work done, it just rubs itself to death faster.

However I knew the 1974 version was rated at 4400RPM, that my 1979 version had been strangulated to meet smog and economy goals. That the big breakable parts had not changed, this was a de-tuning not a holistic re-design. And in fact the original idea for the 351W was a low-cost HIGH-performance 4800RPM-5200RPM muscle-car, with ability to run 7500RPM (for 500 miles, after careful parts-testing).

So I never worried about over-reving it. Though it didn't like to rev, I "knew" it was safe.

Then one day I was passing a pokey BMW, may have touched 4000RPM, and "___" the engine failed. On autopsy, it was not any of the "stressed" parts. Casting and lubrication flaws in the distributor let an advance weight get loose, jammed in the weight-rotor, broke the teeth off the distributor gear. No turn, no sparks, had to get towed home.