As the proud owner of a Synology DS415+ NAS (network attached storage) device, I thought I had better protect the large amount of data accumulating on it by also becoming the proud owner of a suitable UPS (uninterruptible power supply). That way, when the power fluctuates or goes off unexpectedly (as is quite common in rural Australia, where I live) the NAS is protected and will have time to shut itself down in an orderly fashion. So I purchased a Schneider (APC) BR900GI UPS. This article is about how I set things up.
Get the right cables!
Most UPSes (except sometimes the very small ones) don’t have standard Australian/NZ-style sockets. Instead they have IEC-C13 sockets, sometimes called “kettle sockets”.
You will need one or more suitable adapter cables, with IEC-C14 at one end (“kettle plugs”) and AUS/NZ sockets at the other.
Don’t use power boards
It’s generally a bad idea to plug a power board into your UPS. It’s really easy to put too much on the power board. If you use a power board, make sure you don’t exceed the single-socket rating on the UPS. Also, it would be a good idea to label the power board so that you don’t forget that it is powered by the UPS.
Decide what goes where
You need to decide which of your devices you want to attach to which kind of socket. Things that should keep running in case of a power failure should be on battery-backed sockets. Things that don’t need to keep running can be on direct-power sockets. Things that depend on the master device should be attached to slave sockets; independent devices can be on non-slave sockets.
Wait – what are all these different types of socket?
Different socket features
Most UPSes have sockets offering various features. The BR900GI offers four battery-backed sockets and four direct-power sockets. The direct-power sockets deliver power as long as grid power is available, but when/if grid power fails, they stop supplying power – they are not backed up by the UPS’ battery. The battery-backed sockets, on the other hand, will switch instantly to providing power from the batteries if the mains power fails. The battery-backed sockets also smooth out minor voltage fluctuations.
One of the battery-backed sockets is a “master” socket. When the draw on this socket drops below a certain level, the socket will switch itself off entirely – and will also switch off one other battery-backed socket and two of the direct-power sockets – the so-called “slave” sockets. This is a power-saving feature. You might, for example, have your PC attached to the master socket, with an external USB hard drive attached to the battery-backed slave and a printer attached to one of the direct-power slave sockets. In case of a power failure, the computer and the USB drive will be kept alive by battery power. If you switch your PC off (or it shuts down in response to a UPS alert), the UPS will switch off the master socket, and also power down the USB drive and the printer. You can configure the power draw below which the master and slaves will shut off. Or if you don’t want the behaviour you can disable the feature entirely.
I have my NAS attached to the master socket, and the two external USB drives I use for backup attached to a power-board plugged into the battery-backed slave socket. Yes, I’ve labelled the power board 🙂 and I’ve also put child protector plugs into the two spare sockets so that I can’t accidentally plug something else in there. The NAS is configured to shut itself down if the UPS signals power loss; when it does, the slave socket will be switched off, shutting down the connected USB drives.
A word on surge protection
All the sockets on the BR900GI are surge-protected. This means that major voltage spikes will be clamped down to a less harmful value. That’s not as useful as it sounds, for two reasons. The first is that a really serious spike – like a lightning strike nearby – will overwhelm all but the most heavy-duty purpose-built device. A trillion volts will find a way, and a little component in a device the size of a UPS won’t really pose much of a barrier. So surge protection is only for (relatively) small surges.
The second reason that surge protection in the UPS is not so useful is that you really don’t want your UPS doing that. Surge protection works using a sacrificial metal oxide varistor – a component that dies a little each time it absorbs a surge. Eventually it is completely dead, can’t absorb any more surges, and your connected devices are no longer protected. The UPS will still deliver power, it just won’t be protecting your devices against surges any more – and will itself be unprotected.
A better idea is to use a surge-protecting power board, available from any hardware store, and put it between the mains power supply and your UPS. By doing this, you reduce the impact of surges on the UPS itself. You may have to replace a $20 power board occasionally; that’s better than having to replace a $200 UPS! Make sure you get a power board that has a little LED or neon indicator lamp that shows whether the surge protection is still available. Replace the power board whenthe indicator shows the surge protection has failed.
By the way – make sure the indicator lamp on the power board you buy is showing the state of the surge protection, not just whether or not the board is powered!