Battery backup interference

I have my router (2wire 2700HG-B) on an APC battery backup. During a couple of power outages recently I have noticed an immediate and massive increase in CRC errors, rendering the connection pretty much useless, which obviously defeats the object of the exercise.

Is this a fault with the APC, or normal behaviour? I was sort of expecting an improved signal as all the surrounding kit (printers, dect phones, network drives, indeed all the household electrics, would be turned off)

BrianG

UPSs tend to be very noisy.

If you can, move the router as far from the UPS as possible.

What sort of voltage? If it's supplying 240v as a general UPS there might be a lot of RF coming from the invertor.

In reply to a post by john2007:

What sort of voltage? If it's supplying 240v as a general UPS there might be a lot of RF coming from the invertor.

yes 240v

In reply to a post by JonRennie:

UPSs tend to be very noisy.

If you can, move the router as far from the UPS as possible.

I'll see what I can do ! Not much manoeuvering room in the study, but point taken

Probably expected, I'm afraid. The vast majority of modern UPSs, especially the affordable ones, don't generate nice clean sine waves (which would have little harmonic content and hence little RF noise). Downmarket ones use what the vendors generously call a stepped sine wave which will splatter RF noise in all kinds of ways. Others generate pseudo sine waves using pulse width modulation, which may be almost as bad.

Random thought: how serious are you about this? Running direct from DC rather than UPS is perhaps an option to avoid the RFI if you are serious and technically capable (or have a budget ). It should be reasonably easy for the router, but running a PC and monitor won't be so easy. A laptop, on the other hand...

In my box of bits at home I have the parts for a 300W MOSFET-powered pure sine wave inverter (nb inverter not UPS), from a design in Elektor many years ago. Being sine wave it will be electrically cleaner, but being sine wave it will be inevitably inefficient, and the switched mode ones are both more efficient and cheaper (no expensive transformer in many cases). I haven't actually built it yet, but maybe there are similar designs still around. A 300W 12V toroidal transformer was hard to find then, it may be harder today.

Take JonRennie's advice first, keep things as far apart as possible. Get a few clip-on ferrites from Maplin and put them near to the UPS on cables attached to the UPS, and so on. Maybe filter the LAN cable from the PC? Just play around and see what helps.

If your UPS has a "surge filter" for the phone cable, ignore it. Direct from wall socket to router, as best you can.

thanks for techie explanation; most of it straight over my head, although one thing I do know from experiment is not to use the phone line surge filter.

All things considered I think my best course of action will be to ditch the UPS (and find a good book for the odd occasion when the lights go out )

A couple of suggestions to reduce teh RFI. Ferrites on all power leads in and out of the UPS - two or three different sizes too. Put a foil covered board next to the UPS on the side facing the router.

I had give thought to having my router running from a battery instead of mains,

my solution would be a 12v lead acid sealed battery and some electronics or adapt the guts from a spare power brick to make sure the output amperage(milliamps) to the router was the same as the mains power brick,

i would imagine a router would run for days even weeks before the battery would need charging up

For me it would beat buying a pure sine wave inverter type device to do essentially the same job

Edited by tommy45 (Wed 25-May-11 19:16:41)

Get wrapping.