Based on ratios and stuff, adding, say, an additional 2x250 watt panels would get _something like_ an additional 532(!) Wh if they're the same percent efficient as the one panel I had up there already, which is unknown at the moment.

That'd get us to 1150 Wh (ooo, 1.15 KILO-watt hours) and like... instead of three hours in a four hour collection period, we'd get five and a half runtime hours in a four-hour collection period. We'd actually _charge_ battery for a while, not just slow the drain.

but that's all speculation until i have more panels

and NOW I need to stop

#solar #murknet

oh may as well add some numbers:

total actual collected (approx) using two panels: 618 Wh

upper balcony: 320 Wh

single north panel: 298 Wh

best possible add w/a second panel on north: +101 Wh

net of setting up for both panels north and no balcony: -219 Wh

This means we got a bit over three hours of server cluster+refrigerator electricity in a four hour time period, which one can look at either as a net loss of an hour's worth of battery time, or as three hours of server+refrigerator runtime out of four.

so if you have 48 hours of battery (three sixteen hour days), this buys you basically nine additional hours, which gets you the majority of a fourth day. With four 12 hour days, you get 12 hours - an entire 5th day.

#solar #murknet

Next up:

A: Design a quick roll-out of the north side panel set akin to the upper balcony roll-out...

A(2): ...which may and may not last but seems like a decent starting point. Expect to refine it.

B: Test north side again once leaves are down. (on hold 'til November)

D: Look for cheap but still decent quality kit for next steps. I've got mail out to someone on Craigslist but I don't expect a reply. This is not fast.

So yeah. Item A is actionable, the rest... kinda at a stopping point for now. But it's a good stopping point, we should have the ability to stay up a couple of _days_ into a storm. That's not too bad.

(3/3 fin)

#solar #murknet

3: The cable re-org I was talking about in a previous update makes the power-down changeover routine dramatically, and I do mean dramatically, more efficient. We took down three of four servers without notice and were back before anyone even noticed we were gone. CHONK CHONK CHONK _boot_ and we're back.

4: Pending more kit, I think I have a basic plan for any surprise outages this winter. Like all plans, some of it will probably not survive contact with the enemy, but it's basic enough that I think it's okay. And we're a reasonable version of set up for it now.

(2/3)

#solar #murknet

Follow-up: The mail server power supply issue did not repeat when tried again. After restart, power draw levels were basically the same as before, so I don't think we're burning excess power continually.

So that's fun. Intermittent issues are the best issues.

I'll call it an improper power supply shutdown and pay attention if it happens again

#murknet

MAIL SERVER needs checking. It appeared to continue drawing around 50W after shutdown and that absolutely should not be happening. Power supply issues? Errant BIOS settings keeping USB ports powered, thus keeping power supply active? Really need to figure that out. Question: is whatever this is also affecting total power draw while running? Unlikely given its low power draw when running, but worth examining.

Cable management is good UNTIL you get to the power floor and then WOW it's not. I know why and don't know if we can fix it, but we should try.

I think that's all. It was an educational night.

(2/2)

#solar #murknet