Anyone explain this voltage anomaly

[Steven Campbell]

I have a Netgear Switch (GS108) that has been running 24/7 for around 8 years.

The other week it stopped working with all the lights lit up but no activity going through it.

On searching the web for any clues I came across 1 or 2 posts referring to "dodgy" power supplies. So before binning the switch, I swapped the power supply for another Netgear one. However the original power supply was 12V and the one I was trying was only

5V. To my astonishment it worked.

So I ordered a new 12V exact replacement for the original power supply. It arrived today but the switch is exactly the same. The lights all light up but no activity.

I can't understand what the problem is? The switch is marked 12V, the original power supply and replacement power supply are both 12V but they don't operate it. Yet the 5V power supply appears to operate it just fine.

Will I be damaging either the Switch or the 5V power supply by using them together? Anyone know why the 12V power supply isn't able to work it correctly even although it is certainly able to power it?!?!

Cheers.

[Tim Watts]

I have one of those.

Not sure. But I did have something like your problem happen - got it swapped under warranty, so I never investigated it.

I *would* have thought most of the logic is expecting 5V or less. Perhaps the internal regulator has gone short circuit - in which case, I'm quite surprised it hasn't fried. Unless the regulator has gone to some intermediate state of dodginess?...

If it's a bin it job, but a 5V supply seems to work, I'd go with that - but monitor it closely for a day and see if anything smells hot.

[Steve S (another one)]

I expect that the switch has a 5V regulator inside that has gone short.

Steve S

[Steven Campbell]

Cheers. I've been running it for the last week with the 5V power supply and all seems fine. I just didn't want to fry the power supply as its for my Netgear Access point that I'm not needing just now.

[Steven Campbell]

Ah right, that makes sense if I've got this right?

12V is being supplied but it gets changed to 5V inside the Switch? The regulator isn't working correctly so hence why the 12V power supplies don't operate it. Makes you wonder why they needed a 5V regulator inside and they couldn't just supply a 5V power supply.

I wonder if that is why all the recent GS108's come with a 7V power supply!

[John MacLeod]

$s6i$ [e-mail]...

They probably had a good deal on 12v supplies -- the 5v regulator inside would cost only a few pence, whereas they could save a lot by ordering tens of thousands of 12v supplies rather than a more limited number of 12v and 5v and who knows what else in between. Makes sense to standardise on the highest voltage required by any of the line of items which might require a smallish power supply.

[Dave Liquorice]

Is it 12v AC or DC? Some wallwarts are nothing more than a transformer and they feed a nominal voltage of AC to the equipment that then makes the DC rail(s) it wants.

Economies of scale combined with international sales. If you are buying 100,000 wallwarts and 12v ones are 10p cheaper than 5v ones... Also to deliver the same power you'll need a heavier connecting cable with the lower voltage supply to avoid to much loss in the cable. 10W at 12v is 0.8A, 10W at 5v is 2A.

Thats does sort of point a finger at 12v stressing the internal regulation of the unit rather more than was expected.

[pete]

I had something similar happen many years ago, though with a different piece of kit. It turned out that the power supply was just feeding 12 AC into the main unit where it was recitifed and smoothed before being regulated down to the supply for the electronics. The electrolytic capacitor that did the smoothing had either dried out or just died, so the regulator was getting full-wave rectified 12 Volts, which by it's nature dropped to 0 Volts 100 times a second. That in turn continually reset the electronics. Replacing the smoothing capacitor had the desired effect.

If your 5 Volt supply is providing already smoothed DC, that's probably enough to run your router.

[Grimly Curmudgeon]

We were somewhere around Barstow, on the edge of the desert, when the drugs began to take hold. I remember John MacLeod saying something like:

That happened with a PC radio receiver I bought years ago. The supplied wall wart put out regulated 18V, but it turned out the receiver was much happier with a regulated 12V (hardly surprising, as it was designed to run off a car supply for portable use with a lappy). On 12V it's much cooler, as the internal regulator isn't doing much work.

[Phil L]

Not sure but I have a 12v router that I can't find the power lead for. In my shed was a 9v charger from something that had long been got rid of, which had the same shaped jack, so i plugged it in and hey-presto, it works! As an experiment, I unplugged my cable modem (also 12v) and swapped that into the router and used the 9v charger on the modem and that too worked just fine. On a related note, I bought an old radio (for work) from a carboot sale a few years ago, no power lead, battery only. - the battery was an old fashioned 9v jobby, not the pp3 type but the big square chappies, pp7's....when it eventually ran out of steam, I didn't bother buying a new one as it would have cost me more than the radio, so I used a 7.2v nokia charger and wired it into the battery terminals, that was 3 years ago and it still works perfect

[Andrew Gabriel]

Folks who design the switch generally can't relay on a specific power supply, as it will be supplied with a different one in every country. In the US, they usually come with a very small SMPSU built in the plug. In the UK, they come with a sodding great unregulated wall-wart. If you want reliable operation in the face of such variables, not to mention people also changing the supplied PSU, you need to design your own PSU. Point-of-use SMPSUs are often the norm throughout larger commercial PC's (such as 4 and 8 socket systems), where the main PSU generates 48V, and tiny SMPSUs transform this down to levels from 1 to 5V for specific components, including the CPU. This resulted in high volumes, lowering cost, size, and excellent reliability, with the introduction of megahertz switcher moduless. You'll find these inside standard PC's now too, to generate the low CPU and memory voltages.

The other common failure with small netgear switches which get moved around is that the power connector often breaks free of the circuit board.

When I first bought the small 5 and 8 port switches (and WiFi access points), many of them ran from 5V or 12V, and I would power them from a spare disk drive connector. It's got quite hard to find 5 or 12V powered ones nowadays.

[Steven Campbell]

Thanks all for the valuable information. It certainly helps explain what is wrong. I'll try and get myself a 7V supply.

cheers