Diago PS01 pedal supply fault

[Mick Bailey]

A customer complained of getting shocks from his pedal board until it was plugged in to his amp.

It turns out that all of the metalwork is at 85v referenced to earth. The PSU output measures 9.1v, but both the positive and negative are floating at 85v. The sealed unit has a 2 core lead so there's some leakage and lack of isolation to the supply.

Would anyone with one of these check to see if this is more widespread?

Diago PS01
Model STD-409K-E

[Steve Conner]

This is unfortunately quite common in modern switch mode power supplies. In order to pass EMC they need a capacitor between the primary and secondary sides. As the unit is double insulated with no ground pin, they simply connect the capacitor to the live or neutral. It's the old "death cap" all over again.

However, electrical codes are stricter this time round. The capacitor should be a Class Y safety unit with a maximum of 2nF, and so the leakage current should only be a few microamps, but many people (myself included) can feel it as a distinct tingle. It is worse in the UK because we have just about the highest mains voltage in the world.

If in doubt just throw the power supply away and get another one of a reputable make. The 1-Spot is good because if it sucks, you can complain to RG.

[Mick Bailey]

Out of curiosity I checked through an assortment of ungrounded SMPS and at worst I'm seeing just less than 130mv, but these are decent quality supplies and probably better engineered.

It's not a good situation from a musician's perspective - anything that can give shocks, tingles or other-worldly sensations has to be suspect and not ignored. Accepting an effect from one piece of kit may lead to complacency with others, or confusion around where the sensation may be coming from (if it's from something else which may indicate a more serious condition).

[Steve Conner]

I agree, hence why I use a Sound Lab DC-9 for my pedalboard. It's an old fashioned linear supply in a sturdy metal box. Sound Lab DC-9 Power Bank | DiscoFreq's Effects Database

Out of the supplies you tested, did you try the mains plug both ways round? If the EMI capacitor ends up connected to neutral, you won't get any tingles. Some makers may take the extra care to make sure the cap ends up on the neutral pin in the UK version of their product, but I bet not all of them do, especially if the supply is sold worldwide.

Many countries have plugs that can be inserted either way round, so they probably decide that it's not worth the hassle to make the UK version special.

[Mick Bailey]

I didn't try the plug both ways - I'll see if this makes a difference. If it does I'll mark it up with polarity. I'd personally rather have linear supplies for pedalboards and I've built quite a few (36) recently. A lot of SMPS put out plenty of noise when used with pedals. I noticed a ring-mod effect with one and it was entirely unintentional. Must be sum and difference components between noise and audio getting through.

I got a reply from Diago on the subject;

"Depending on skin dampness, footwear, ground conditions etc, tingling from a 2 pin SMPS is actually very normal, although I understand that it can be a little disconcerting. Most of this kind of power supply that I've ever tested float around 80-90V AC until they're hooked up to something earthed.

Of course, voltage alone isn't a problem. Have you or your customer tested the following:
1) Isolation between the mains inputs and the DC outputs. Tests here show them to be well over the limit of our test equipment - >200MOhm.
2) Leakage current between DC outputs and earth. Test here show them to be around 0.1uA AC which is way below guidelines and any level of danger. (EN/IEC regs allow up to 1.5mA at 277V AC, 60Hz)

As you've tested, with the system not earthed, and the AC voltage from the 9V DC outputs to earth should each be 80-90V AC. This is caused by the input filter caps used to keep as much of the SMPS noise out of the AC mains.

If you're getting the figures stated above, then it's all working within spec and there's nothing to worry about. If you have any other similar power supplies around - laptop power supplies for example, you should find almost exactly the same situation."

[Steve Conner]

This is a bit disingenuous of the Diago guy. The 1.5mA limit is for Class I devices where the leakage current can be discharged harmlessly to ground. The limit is 0.25mA for Class II devices like 2-pin SMPS. Touch Current - Leakage Current - IEC 60950

The theoretical leakage through a 2nF capacitor at 240V AC, 50Hz, is 0.15mA, so I think his assertion of 0.1uA is a bit optimistic, unless it was tested with the capacitor connected to neutral.

[Mick Bailey]

I'm getting the unit back to test leakage current. Should be interesting.