Which UPS?

[Scott Allen]

On 13 June 2012 10:19, James Knott <james.knott-bJEeYj9oJeDQT0dZR+AlfA at public.gmane.org> wrote:
> As I mentioned in another note, the high voltage DC in power supplies
> approaches 170V.  Why should a power supply that runs on AC and be more
> efficient than one that runs on a similar DC voltage?  What is it about that
> referenced DC supply that makes it so inefficient?

It's not a similar DC voltage. The AC input supply rectifies 120VAC to
170VDC. The DC input supply (that you're questioning my reasoning
about) requires 48VDC.

> What is it about that referenced DC supply
> that makes it so inefficient?

I don't know why the supply that Lennart referenced is so inefficient.
That's why I bumped it up to 85% for my calculations. And again, I
used only 80% for the 120VAC input supply. So, my calculations assume
a DC input supply that is more efficient than an AC input supply.

>> UPS needs to convert 120VAC to 48VDC at, let's say, 85% efficiency.
>
> What is the conversion efficiency of the rectification part of the AC
> supply?  Why should a separate rectifier be significantly different

We're talking about 120VAC to 48VDC. It's not just rectification. We
have to get the voltage down to 48VDC. I Assumed that this would be
done by a switching supply in the UPS, since the voltage would have to
be regulated to charge the battery, hence my 85% efficiency estimate.

If the PC supply accepted a fairly large voltage swing, you could get
away with just a 120VAC to 48VAC transformer then rectify and filter
it (and then just regulate the battery charger circuit). I don't know
how efficient the AC transformer would be but we're now less efficient
than just strait rectification to 170VDC. Also, a transformer that
could provide the 500 or more watts required for a typical UPS would
be big, heavy and possibly expensive.

>> Note that I was talking about the specific case where the user is
>> limited to a 120VAC source, which is on line, and either a 120VAC
>> output UPS is used, or a 48VDC output UPS is used.
>
> An AC output UPS will have an additional stage to convert DC to AC.  The
> computer will then have an additional stage to convert that AC back to DC.

Read my statement that you quoted above. I said "which is on line".
The mains are up. For the AC output UPS the AC is just fed strait
through (I presumed via a mechanical relay). There's no conversions
taking place in the UPS and therefore no losses in the UPS.

>  Even if UPS is not used, why should producing DC in a separate box be less
> efficient than when in the same power supply.

Again, in the UPS (or other separate box) we have to bring the voltage
down to 48VDC. Inside the AC input PC power supply we just rectify to
170VDC.

-- 
Scott
--
The Toronto Linux Users Group.      Meetings: http://gtalug.org/
TLUG requests: Linux topics, No HTML, wrap text below 80 columns
How to UNSUBSCRIBE: http://gtalug.org/wiki/Mailing_lists