home | list info | list archive | date index | thread index

Re: [OCLUG-Tech] Re: power usage

Stephen Gregory wrote:
I have retested the Tripplite UPS and an APC UPS. As far as I can tell the Tripplite does consume 10+ watts at all times. Oddly when I connect a low load, less than 1A, into the Tripplite the ammeter reads a little lower then when nothing is connected. It could just be that the Tripplite has a crummy power factor. I would need better equipment to figure it out.

The APC UPS behaves as expected. It draws close to 2w. It does not show the odd spikes of current draw either. The APC is an older basic 200VA model. I used to seem them under cash registers all the time at the grocery stores.

--sg


I would surmise from this that the Tripplite uses a diode bridge rectifier running from the transformer on the input side to create the DC to charge the battery. This is far less efficient than using a switch mode power supply to do the same task. The APC uses a switcher IIRC. Tripplite has been around a lot longer, therefore it is probably they are using the older technology (which was cheaper back then). You can't turn off a transformer like you can a switcher.

Tripplite (Again IIRC) had a real affection for ferroresonant transformers - very robust, but not extremely efficient. If run in saturation, "ferros" add quite a bit of brownout holdover, greatly helping the switchover delay while turning on the inverter in the event of a mains failure. If your hear a persistent light buzz when the Tripplite is running, it is probably stack vibration in the transformer. If the buzz goes away when the system is running off the inverter, that is another piece of circumstantial evidence.

Conclusions: The higher efficiency of the APC will have much lower long term operating costs, but a predictable problematic potential waveform "blip" during switchover to battery driver inverter. The Tripplite (if I'm right) probably has a more consistent switchover without the potential for wave collapse, but does so at a high efficiency penalty. The way to determine this conclusively is to capture and observe AC output with a storage scope, triggered on the zero crossing detect circuit.

For general PC usage, the APC will be fine. For sensitive equipment, medical applications etc., the Tripplite would probably be better (but not as good as a true UPS that always runs off the inverter). True on-line UPS cost much more in both efficiency penalties and purchase cost as the batteries and inverter circuit have to be stressed to handle full time operation.
--
Bill

replies

references

message navigation