1
\$\begingroup\$

In an application, I use an industrial AC-DC power adapter, the meanwell GST160A15. The power adapter gives 15V@144W maximum power. I carefully read the datasheet, and it states that the power adapter has overvoltage protection, 105 ~ 135% of output voltage. My question is, in case of fault, can the power adapter give out an hazardous voltage? With hazardous I refer standard convention for wet locations, 70V DC or 33 V r.m.s and 46.7 V peak

\$\endgroup\$
8
  • 2
    \$\begingroup\$ What is 135% of 15V? That should be the worst-case output voltage. \$\endgroup\$ – Adam Lawrence Jan 16 '18 at 17:09
  • 1
    \$\begingroup\$ This is not the place to ask. Assume yes unless the manufacturer definitively states otherwise. If you really need to know, ask them if they are willing to certify it for medical applications ... and watch them back away carefully. \$\endgroup\$ – user_1818839 Jan 16 '18 at 17:10
  • 2
    \$\begingroup\$ @BrianDrummond Why it should not be the place to ask? I wonder if theoretically it is possible, for a CE and FCC certified power adapter to ever give out hazardous voltage. I obviously refer to the theory, according to what the manufacturer declared. \$\endgroup\$ – Francesco Jan 16 '18 at 17:12
  • \$\begingroup\$ @AdamLawrence The datasheet states that. Or it seems so, under the category over-voltage protection \$\endgroup\$ – Francesco Jan 16 '18 at 17:13
  • \$\begingroup\$ Depends : if your interest is idle curiosity then asking here is OK, but if you have specific safety requirements (which we don't know) then ... probably no. For example, immersion probably doesn't count as a single fault scenario. \$\endgroup\$ – user_1818839 Jan 16 '18 at 17:34
1
\$\begingroup\$

Under no-fault conditions that type of power supply can be assumed to produce 15 volts (plus or minus a small percentage due to build up of tolerances and temperature coefficients). If part of the the voltage regulation circuit failed it could produce a voltage significantly higher than 15 volts and this is where a lot of designs use an independent crow bar circuit to clamp the voltage to an avarage safe limit should a component in the main circuit fail.

I suspect that design has protection against single component fails and uses a crow-bar circuit (as do a lot of designs that are certified CE). Some designs don't but they rely on an extensive analysis of what can happen under single fault/failure conditions.

I expect that if you look in the product data sheet there will be some information about this. As to whether this power supply is suitable for the environment type you wish to use it, read the data sheet.

\$\endgroup\$
8
  • 1
    \$\begingroup\$ OK. But I expect the crowbar circuit to be integrated into the power adapter itself, right? Being CE certified, the power adapter should not, by specification, give out an hazardous voltage, under any condition, both in normal mode and in single fault condition. \$\endgroup\$ – Francesco Jan 16 '18 at 17:21
  • 1
    \$\begingroup\$ Well, upon further reading of the data sheet it says that over-voltage protection is "Hiccup mode, recovers automatically after fault condition is removed" and this sounds pretty much like the sort of protection offered by a crow-bar circuit and I don't see any problem in this method - if voltage rises above (say) 30 volts DC then an active clamp (crow bar) will activate. \$\endgroup\$ – Andy aka Jan 16 '18 at 17:26
  • 1
    \$\begingroup\$ @Francesco CE self-certified \$\endgroup\$ – Jeroen3 Jan 16 '18 at 18:49
  • \$\begingroup\$ Isn't voltage inherently limited by the transformer? \$\endgroup\$ – Fredled Jan 16 '18 at 20:12
  • \$\begingroup\$ A transformer used in a switch mode supply is not operated as one in a linear supply hence it can produce significantly out of bounds voltages when driven by circuits suffering a single component failure. @Fredled \$\endgroup\$ – Andy aka Jan 16 '18 at 22:12
1
\$\begingroup\$

Certainly it could, if the insulation in the transformer fails or the Y capacitor fails it's possible, particularly if the earth connection is missing (eg. plugged into the end of a modified 2-wire extension cord).

If the adapter is counterfeit the chances higher than if it is genuine.

It's also possible that if there was a failure in the feedback circuit the switching supply could output higher than rated voltage, possibly approaching theoretically hazardous levels, however this should be prevented by the OVP circuit so it would take two failures rather than just one.

\$\endgroup\$
3
  • \$\begingroup\$ I assume that everything is connected as expected, and the power adapter is used within the limits, and in the way intended by the manufacturer. Earth connection in place, genuine power adapter. \$\endgroup\$ – Francesco Jan 16 '18 at 17:31
  • 2
    \$\begingroup\$ Failures are still possible, just relatively unlikely. \$\endgroup\$ – Spehro Pefhany Jan 16 '18 at 17:59
  • \$\begingroup\$ Sure. But the basic concept is that the power adapter has already well rigged certification and testing. Moreover, the concept of single fault condition applies. The combination of faults from different sources, although certainly possible, is not considered a likely event for certification purposes. \$\endgroup\$ – Francesco Jan 16 '18 at 18:03
1
\$\begingroup\$

The 105...135 % protection limit is an included extra circuitry in the DC output which shorts too high DC output voltage, which can be caused by failures.

But this is already told by other members. I want to add something else:

Power supplies of this type have a common overvoltage mechanism.(see NOTE1) They have grounded 3 wire AC input, where the protective earth wire is connected

  • to internal frame ground for normal Class I protection
  • capacitively to AC line and neutral wires for rf filtering (=to keep the produced rf interference inside the unit, out of the AC mains)
  • to the negative DC output

If one connects this PSU to a non-grounded AC outlet, it has 50% of the mains AC voltage between the output and the surrounding ground. It comes through noise filter capacitors. When one connects this to a PC and then connects his PC to other equipment (printer, audio mixer), something very likely gets fried if there's a connection to ground. I have seen it to happen and also destroyed at least one printer and one audio amp.

Microphone eaters also have something to complain when the AC hits their lips.

People seem to insert plastic tape to AC plugs to break the protective earth connection to break the ground loops which cause intolerable hum and buzz in audio systems. After doing it connecting the AC first and then the audio wires stops the humming and buzzing totally and the result is permanent until one gets new equipment.

NOTE1: This is the same with normal PCs which use normal grounded AC cables.

\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.