How high will working-order power supplies read on their own?

[Mark Hammer]

Many of us have had the experience of measuring the output of a wallwart with our meter, when it's not plugged into anything, perhaps to verify that it's working or confirm plug polarity, or whatnot.  And it is not uncommon for a PS that isn't powering anything at the moment to read higher than the actual voltage indicated on the chassis.  So, it might say 9V @300ma on the sealed black wart, but you're measuring maybe 12V at the plug.

I've gotten accustomed to such overvoltage readings over the years.  But what I don't know is how high they can go, and when a noticeably higher voltage is an indication of something having gone wrong inside that little sealed black box, like an electro that is past its best-before date, or perhaps bare minimum rectification.

Is there any rule of thumb that can be applied in discerning whether a supply will likely deliver exactly what the chassis says when plugged into something, versus delivering the "wrong" voltage?

I'll note that a friend's kid came over the other day with his Boss RC-1 looper.  It ran fine off a 9V battery but would not run off his external supply (one of those "universal" things with a set of interchangeable plugs, and a switch to select a variety of voltages from 3 to 12V).  We tried out a couple of supplies I had lying around, and they all worked fine, powering up his pedal without difficulty.  It looked like the internal diameter of the tip connection on his supply may have been a little large-ish, so I figured it probably wasn't making a good connection with the jack and grafted another plug onto the cable.  But that didn't help.   The switch on the unit was set to 9V, but it read 12.6V on my meter.  I figured it was the normal sort of overvoltage.  Just for kicks, I moved the switch down a notch to the 7.5V setting, which did read a tad over 9V at the plug.  And when I plugged into his looper it fired up right away with nary a glitch.  I don't know if there was an issue with the switch, or with the transformer inside, or what, but when we disregarded what the unit indicated, everything  was ironed out.

So, again, what sort of overvoltage on an unloaded supply should be treated as conspicuous?

[Rob Strand]

Stats for *unregulated supplies*  *unregulated* wall-warts

Min unloaded V:     110%
Max unloaded V:    150%
Typ unloaded V:     130%
Typical output voltage at full rated current:    90% of rated V
Load current when output voltage  = rated voltage occurs at 66% of rated current
Filter cap typ 1000uF.

These are all average voltages and don't consider ripple min/max peaks.

What's clear is in order to reduce the unloaded to rated voltage ratio it is typical (for manufacturers of wall warts) to allow the voltage to drop 10% below the rated voltage at full current.

So, again, what sort of overvoltage on an unloaded supply should be treated as conspicuous?

As far as a product working it's pot luck.

[Rob Strand]

FWIW, the RC-1 is spec'd for the newer PSA power supplies.  I believe they are regulated to around 9V.  So technically you shouldn't plug the older unregulated ACA power supplies into it.

Most of the Boss stuff I've seen spec'd for PSA chucks out the diode + resistor + zener on the power rails and puts in a one-transistor active filter.  Some have one-transistor zener regulators.   The latest "digital era" of pedals have switch-mode devices in there, so perhaps they have overvoltage protection.

[PRR]

Some of the very small transformers have 40% regulation: the no-load is 40% higher than the full-load voltage.

I don't doubt somewhere sometimes someone saved another penny by designing to 50% over.

Welding transformers typically had no-load voltage much higher than loaded voltage, even 3X. However this involved added complication to take full load without stupid excess heat.

[antonis]

A brute way is to connect a proper wattage resistor of a value such as to get about 3/4 (or slightly less) maximun rated current and measure actual voltage..
(e.g. for 9V@300mA, a resistor of 39R/2W should be OK for a short period of testing..)

Of course, that's only usefull for proper voltage verification concering the specific load..
A more appropriate way should be to use a resistor of a value: wallwart nominal voltage / particular effect estimated current and measure voltage actual voltage drop across the resistor..
(e.g. for the above mentioned wallwart and an effect current of 23mA say, a resistor of 390R/200mW(*) should be OK..)

(*) Better use a 400 - 600 mW, in case of higher than nominal voltage output..

[amz-fx]

Measuring an unregulated supply without some amount of load on it really doesn't tell you much.

I use a small load consisting of an LED and 560 ohm resistor that is enough to pull down the output to give a better indication of what you have. A 9v wall adapter that reads 12v when open will drop to 9.6v or so under even a light load such as the LED. You may be surprised how much the output drops once it has a load on it.

If you have a high amperage unregulated adapter, you might have to go with a heavier load as Antonis suggested.

Best regards, Jack

[Mark Hammer]

Thanks so much, guys.  This is a very useful, and maybe even important thread, deserving of sticky status.  Granted, more and more players are using a power brick these days, but a great many, especially beginners, are not.  Knowing more about wallwart behaviour is useful for them.

I've referred people on another forum to this thread.  Much appreciated. 

[bool]

Calculate the req. load; extrapolate to resistance in ohms, and stock up on some wirewound resistors; you can also make a "load box" with a switch to select a few loads for the common scenarios; with power socket and voltmeter output...

[Rob Strand]

FWIW, the stats I did were black-box tests.  I didn't assume anything I just measured external behaviours.   

Based on those results there seems to be a recipe used by wall-wart manufacturers.   

It doesn't look like they just stick a 12V ac transformer in a box and call it a 12V DC supply.  The behaviours don't show the excessively high un-loaded voltages you would see with a small 12V ac transformer with poor regulation.   These things seemed to be designed to keep within *reasonable* bounds.    I'm sure not all are like this, especially in this era of selling junk and fakes, but a significant percentage are.  I guess that stops things blowing up. If a 12V DC wall-wart put out 20V it's going to cause trouble.

[antonis]

If a 12V DC wall-wart put out 20V it's going to cause trouble.

They usually drop significantly their "open air" voltage even with a couple of mA load..

But yes, for an almost zero current demand, they ARE potentially troublesome..!

[bool]

...
They usually drop significantly their "open air" voltage even with a couple of mA load..
...

When I was younger I once opened some jap. wallwart for repairs for my then-gf's father and noticed they put a small resistor across the 'lytic. Now I suppose this wasn't just to zap the DC when the wallwart was plugged-out but also to make it behave better;  a too-high DC on a lytic could easily zap a device if you plug the DC jack into it after the wallwart was plugged-in and charged..

[antonis]

they put a small resistor across the 'lytic.

To be honest, I can't tell if it's a "bleeder" resistor 'cause if it indeed is then output voltage should be (almost) of nominal value..

[Rob Strand]

To be honest, I can't tell if it's a "bleeder" resistor 'cause if it indeed is then output voltage should be (almost) of nominal value..

I know the problem you mentioned.   You can get crazy high values without the bleeder.
I can't remember if I had a bleeder or not when I measured the open circuit voltages (it's a long time ago).