UPS for FreeBSD

[Erich Dollansky]

Hi,

On Mon, 1 Dec 2014 22:46:58 -0500
Alejandro Imass <aimass at yabarana.com> wrote:

> On Mon, Dec 1, 2014 at 9:38 PM, Erich Dollansky
> <erichsfreebsdlist at alogt.com
> > wrote:
> 
> > On Mon, 1 Dec 2014 12:39:24 -0500
> > Alejandro Imass <aimass at yabarana.com> wrote:
> >
> > > On Sun, Nov 30, 2014 at 12:33 AM, John Johnstone
> > > <jjohnstone at tridentusa.com> wrote:
> > >
> > the average volage over a square is much higher than over a sine. If
> > the components are not designed for this, they will fail. You can
> > see this effect by connecting a light bulb to an UPS and switch
> > mains off an on again. The smaller the difference in brightness
> > gets, the better the UPS emulates a sine. Of course, the life time
> > of the light bulb gets reduced.
> >
> >
> Should not be a problem for any modern switching power supply. In
> fact the higher the RMS the more efficient it will operate. The RMS
> of the sine x/SQRT(2) of a square wave is just x, where x is the peak
> to peak voltage. The peak to peak of 120V is around 170V which is
> just fine for most switching power supplies which are rated for about
> 250V RMS. It _could_ be a bit of a problem is the UPS is for 240V but
> I have never personally seen one in a North American store. Besides
> any 240V UPS manufacturer would have to know this and honestly I
> don't really think that any UPS uses a square wave.
> 
the world is a bit bigger than the USA. 

> I will be sure to test this with an oscilloscope when I finally buy
> the UPS that started this thread - I promise I will buy the cheapest
> one I find and post pictures of the output waveform ;-)
> 
There are sites which do this already. I just do not have URLs at hand.
They normally do not do a simple squere but use i.e. 3 squares. The
higher the price gets, the smaller the squares will get.
> [...]
> 
> 
> > This is the question. Most power supplies will be able to handle the
> > square waveform. Escpecially modern power supplies being able to
> > handle any voltage from 100 to 250V.
> >
> >
> Actually I was referring to another comment that suggested that HF
> harmonics from saw-tooth emulation or square wave "ring" could somehow
> affect the switching PS of a typical computer. I think this is
> unlikely because it will get filtered by the first rectification
> circuitry of the switching power supply (I am no expert and could be
> wrong and maybe some switching PS would be sensitive to this, but I
> really doubt it). HF does affect some transformers and linear
> equipment and still is very unlikely.
> 
Yes, it should be like this.
> [..]
> 
> 
> > It is the volume. When the cheaper UPSes appeared, their price
> > advantage was pretty minor.
> >
> > I wonder meanwhile if it is not cheaper to use an inverter used in
> > solar panel installations. At least for larger inverters, they have
> > a clear price advantage but require batteries which are not
> > available at the next corner shop.
> >
> 
> The problem I think is the switch-over speed. It would be really cool
> to find UPS that use *any* external battery, and you could just use a
> car battery for example which is very reliable, durable and readily
> available.
> 
Oh, I did this once for a client at a very remote location. He only has
had a lousy generator but wanted to run a normal PC in a decent manner
without spending too much money. We took the a cheap car battery
charger, two car batteries and an off-shelve inverter to be used in a
truck. Nothing can beat this as the UPS is always on. So no switch over.

The clear disadvantage are the batteries. Car batteries are designed
for a high load for a short period of time. The result with a low load
over a longer period of time is bad. The best I have seen yet are
batteries used for solar panels. The quality of the batteries is really
outstanding. The price is even more outstanding.

Erich