| General > General Technical Chat |
| Accidentally grounded live wire from UPS line |
| << < (3/4) > >> |
| ajb:
--- Quote from: thm_w on August 04, 2021, 08:20:46 pm ---The RCD is detecting an imbalance between L and N, so it doesn't really matter as long as the supply is coming from L and N. Some discussion here: https://www.iceinspace.com.au/forum/showthread.php?t=82649 --- End quote --- The benefit of an RCD/GFI on the output depends on the construction/topology of the UPS. If the output of the inverter is isolated from the equipment ground (except maybe via some suppression caps) then there is no real path for current to escape from the output, because there's no way to complete a circuit from the line output except via the neutral output. If the UPS is a type that has a direct connection between input and output when not in battery mode then an output RCD will still be effective in that mode. If the UPS is an 'online' type that always runs the load from the inverter then the output RCD will never be effective. This is all regardless of any ground connection AT ALL to or through the UPS. If the UPS somehow has its output referenced to ground then an RCD could be effective on the output--I don't know for sure, but I doubt that's a common configuration among cheap consumer UPS systems because doing so would imply certain assumptions about the way the supply or load are wired that are probably too risky for household situations. I assume that building-scale UPS/backup generator systems integrated at the service or panel level use a similar neutral/earth bonding system as utility power does. Those schemes vary with locality, but such UPS systems would be exactly the same as utility power in terms of RCD protection. It's really important to note that if you plug an RCD into the output of a UPS, just because it trips when you press the test button DOES NOT mean that it's providing any protection! The test button just connects a resistor from the output side of the differential current transformer to the input side, which causes the sort of current imbalance the RCD is designed to detect and protect against. It's done this way, rather than connecting the trip test load from line to ground, because it allows the test button to work even if the circuit has no equipment ground* or the line and neutral conductors are swapped, and it ensures that the test button won't trip an upstream RCD instead of the one it's built in to. The only way to test that an output RCD is doing anything useful is to use an external plug-in tester, which DOES route the test trip current to ground, since that's the only way it can cause a current imbalance in the line and neutral conductors, and to test it while the ups is running from the battery. * At least in the US, installing an RCD is the only legal way to fit grounding receptacles on a circuit that doesn't have an equipment grounding conductor (so long as those receptacles are labeled "no equipment ground" as well), so this is an expected configuration. |
| NiHaoMike:
I read on Facebook a week or two back about someone who had a UPS that would give him a shock if he unplugged it to test it and touched the prongs in the process. Turns out that UPS didn't isolate the neutral and the load had a ground fault H-G, which means that there was a path for voltage to N-G on the input. |
| cmosfxx:
--- Quote from: tom66 on August 04, 2021, 08:59:11 pm --- 3. There is no short circuit protection on APC UPS? Shouldn't the UPS switch off to protect its self? If it has electronic short circuit or overload protection, it likely applies to the battery output only. If it's anything like older UPSes I used to repair, it was purely based on the battery voltage dropping below a low threshold, as the batteries were usually too weak to support anything more than say 800W load. 4. Is the UPS ok now or there is a hidden damage somewhere? It's probably fine if you can't find any obvious defects. --- End quote --- I can't seem to find any infos about overload/short protection. If it's like an older model I have, the fuse on the back is just an overload (thermal?) Fuse on input hot side only. Now there is a big problem. The 230v versions, like the one I've got, are missing a critical feature. The reversed polarity warning. Guess what? All this time I had reversed polarity on the input of the UPS. So the input neutral was actually the hot. This is not a problem untill you really need the fuse to play its role. And that's why the onboard fuse never tripped. The UPS is working fine so I'm thinking maybe the relay handled all the power throught it for that time. |
| NiHaoMike:
--- Quote from: cmosfxx on August 08, 2021, 01:18:24 pm ---Now there is a big problem. The 230v versions, like the one I've got, are missing a critical feature. The reversed polarity warning. --- End quote --- US 208V and 240V are balanced meaning there is no reverse polarity. Therefore, there was originally no need for a reverse polarity indicator and they didn't want to redesign it just to sell overseas. |
| cmosfxx:
--- Quote from: NiHaoMike on August 08, 2021, 01:47:22 pm --- --- Quote from: cmosfxx on August 08, 2021, 01:18:24 pm ---Now there is a big problem. The 230v versions, like the one I've got, are missing a critical feature. The reversed polarity warning. --- End quote --- US 208V and 240V are balanced meaning there is no reverse polarity. Therefore, there was originally no need for a reverse polarity indicator and they didn't want to redesign it just to sell overseas. --- End quote --- Ooh that makes sense now, thanks for the info. |
| Navigation |
| Message Index |
| Next page |
| Previous page |