Abusing brick on the rope

Hi all, I just got done fixing a piece of apparatus that was returned to us. The thing was trashed, (all sorts of blown components) and I've been trying to figure out what happened. Perhaps what I found weirdest is a power monitor LED. This sits across the +/-15 V rails, with a 12V zener and 5 k ohm resistor in series. (the zener stops it from turning on if one of the rails fails.) It just barley put out any light.... trashed. It can take 30 mA forward so >~150 V for too much current. And in reverse it doesn't break down till -80 V. (I just tested it... it shows a photo response too, but don't distract me.)

So I was in contact with the school that was using this.. checking return address, and I asked if they wanted a new brick on the rope. (Because they hadn't returned theirs with the unit.) It was just then that I learned that the brick on the rope had failed!! And they had purchased a new one from the vendor (astrodyne)

So my current theory is that the brick on the rope failed (somehow) and blew 120 Vac through my circuit.. with devastating results. Any idea how I can test this theory.. how do I blow up a brick on a rope? (Astrodyne Mod num. SPU45-210) +15 V @ +2A / -15V @ -1A.

Thanks George H. (maybe I should contact astrodyne.)

Reply to
George Herold
Loading thread data ...

why would you want to blow it up when you can bypass it?

NT

Reply to
tabbypurr

I have seen a LED become dim after being exposed to episodes of reverse polarity. In the case I saw it was spikes of 18-30V (so well above the

5V usually seen for Vr on LED datasheets). The first few zaps seemed to have no effect but damage must have built up.

If that is a SMPS then 120V 60Hz appearing on the output seems unlikely but if an output rectifier shorts then spikes of reverse HV at the switching frequency could explain what you saw?

piglet

Reply to
piglet

Bypass how? I've got ~100 of these out in the field (class rooms) if there is a way to fry the power supply that then fried the electronics I'd like to know.

George H.

Reply to
George Herold

Right, last night I was thinking just a reverse voltage would do it. (I was picturing some burly student cramming in the power plug the wrong way.) So I tested the LED here... it takes 80 V of Vrev. (And then seems to turn into another dang spad.) It's happy sitting at -40V for several hours

Spikes at the switching frequency would explain some of the damage, but there is one section that has an LC filter on the power lines and that should have taken care of the ~100 kHz switching freq. There is the same LED as indicator on that section of circuit. (Oh... checks schematic) The led is before filter.. So your theory stands.

I wish I'd heard about the failed SMPS before today... I'll ask more questions... maybe they still have the failed unit and I could do an autopsy.

George H.

Reply to
George Herold

Power line lightning strike? Was any other equipment in the school damaged or destroyed at the same time?

Astrodyne claims 3KV isolation between line and output...

John :-#(#

Reply to
John Robertson

tirsdag den 2. oktober 2018 kl. 21.56.03 UTC+2 skrev George Herold:

no GFCI? else I think you need two short to happen at the same time to get full mains on the output which I don't think is very likely

Reply to
Lasse Langwadt Christensen

There must be some miscommunication, as asking how to put 120v into a low voltage connector seems like a crazy question.

NT

Reply to
tabbypurr

Right sorry, I'm guessing some type of HV (>50V) out of the SMPS. I'm not sure what. So my first idea was to pound on the supply, somehow? just a dead short? There are no 'big' inductors in this apparatus, loads are mostly resistive. The good news is all of the circuits had cap multipliers on the PS inputs, and mostly, whatever it was, just blew the unprotected input filter transistors.... next board spin I can add screw terminal blocks and call them input filter/ fuses. :^) to-93 pacs

2N4401/3.

The same apparatus is already using ~$2 opamps as amp/ fuses.

George H.

Reply to
George Herold

Oh, and I need to add a bit of resistance in the power supply leads, where it was missing before. Should every low power, supply lead have 10 ohms in series? (low power, less than 100 mW)

George H.

Reply to
George Herold

The value that suits you depends on several things, only you can decide. How much R the thing tolerates staying within spec, how much P you want to diss, how much efficiency matters etc. A safety R can double as a fuse.

NT

Reply to
tabbypurr

what if they substituted a lab powersupply for the broken brick, but fouled that up somehow?

--
     ?
Reply to
Jasen Betts

transorb and polyfuse? over voltage disconnect with some high voltage transistors?

Reply to
Lasse Langwadt Christensen

George Herold wrote in news: snipped-for-privacy@googlegroups.com:

It sounds like the result of a severe transient (lightning strike) on the input line to the brick which then passed it through in a rare failure mode.

Instead of trying to recreate the failure, diagnose the failed unit you already have and see if the isolated AC input side found a path through the PS to the output side.

What device are you running that requires a dual output supply? That itself is an odd instance.

Reply to
DLUNU

** I hear alarm bell going off.....

** There are any number of ways - one I had to deal with involved a badly damaged analog synthesiser ( keyboard instrument ) that had vaporised ground tracks and 30 odd blown small components, bead tantalums, diodes CMOSs ICs etc.

I later learned it had been briefly powered from a wrongly wired extension lead with ground & active swapped and connected to a normally earthed valve amp. Bang.

In your case, the device and the "brick" would likely both be seriously damaged, assuming earth is carried trough to the outlet.

.... Phil

Reply to
Phil Allison

** I hear alarm bell going off.....

** There are any number of ways - one I had to deal with involved a badly damaged analog synthesiser ( keyboard instrument ) that had vaporised ground tracks and 30 odd blown small components, bead tantalums, diodes CMOSs ICs etc.

I later learned it had been briefly powered from a wrongly wired extension lead with ground & active swapped and connected to a normally earthed valve amp. Bang.

In your case, the device and the "brick" would likely both be seriously damaged, assuming earth is carried trough to the outlet.

.... Phil

Reply to
Phil Allison

George Herold wrote

It being a school, it would not surprise me (has your 'thing' any output? if 'they' (accidently??) put some ...mains into your output and that way in reverse killed your unit and that power supply (a more common word) too. He 'whodiddit' will remain anonymous and life goes on. In that worst case they will do it again with the new replacement. From a legal POV you can use an expert witness to show your thing could not be the cause. Shit situation.

Not enough data

Reply to
<698839253X6D445TD

How about this:

  1. Ungrounded power supply with filter caps
  2. Hot connection of power plug so that GND connects after supply rails
  3. Apparatus grounded through some other route

This could discharge the mains filter cap through the power rail. Depending on what's on the rail, the positive/negative discharge could go through the LED.

--
mikko
Reply to
Mikko OH2HVJ

Microwave oven?

Reply to
Ingvald44

Right, I just don't know. The professor seems adamant that the students did something wrong. I can't diagnose the failed supply, 'cause I ain't got it.

Dual supply +/-15V is for opamps. I know they are rare these days (split supplies), but at the time it was easier than splitting some single supply.

George H.

Reply to
George Herold

ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.