[Nut-upsuser] Cyberpower Drive in 2.2.0 Fedora 7

[Arjen de Korte]

> Well, I can help testing the connector on a separate system and seeing
> how that CyberPower supplied driver plays with it, since I do have quite
> a few systems in that network to work on, aside from my NUT master
> server. I am more than happy to test with it all, and get it working as
> best as I can.

Thanks for the offer, but this probably isn't worth the effort. In order
to determine the scaling factors (most likely some exponential function),
we need to be able to force the UPS to send different values for these
parameters. You might be able to tweak the input voltage a little, but
this is harder (maybe impossible) for the input frequency, temperature and
battery charge. So what we need here, is the 'reverse' driver I wrote
about earlier, in order to have full control over what it sends and what
the bundled driver reports.

> I haven't had the opportunity to get the Offline/plug
> pulled portion tested (I am across town right now, and got the driver
> and UPS software running remotely) so I will get that checked as well
> when I get home.

That would be nice, since this will allow us to determine what the device
is actually reporting (input or output voltage).

> I CAN test the AVR side too (My apartment has some
> circuits that are the suck in regards to wiring and power, older
> apartment for the most part) as soon as that is done, I will post the
> results and get the 'test' system up and running, and compare results
> between the live 1500 and the emulator.

Again, thanks for the offer, but this won't provide anything we don't
already know.

> I can provide as much as I can with that, but a skew of 5 degree's C and
> a load difference of about 6% isn't that bad of a skew, since it allows
> you to error on the side of caution.

This shouldn't stop us from shipping the thing. As long as the values are
not 'obviously' wrong, it's OK for now. The 'measurement' circuits in
UPS'es are notoriously unreliable and basically only useable to follow
trends. So if your UPS has always reported a temperature of 36°C and
suddenly rises to 45°C, this is an indication that something has changed
and need to look at it. Acting on absolute values is nonsense, since you
have no idea about the accuracy (although the resolution and repeatabilty
are usually fine) and most users don't know where to set a sensible limit
anyway.

Best regards, Arjen