Best car batteries for server backup? / General Science and Electronics / Forums

Forums

4hv.org :: Forums :: General Science and Electronics

« Previous topic | Next topic »

Best car batteries for server backup?

1 2 3 4 next

Move Thread

LAN_403

Backyard Skunkworks

Sun Jul 17 2011, 02:35AM

Registered Member #1262 Joined: Fri Jan 25 2008, 05:22AM
Location: Maryland, USA
Posts: 451

Looking to put a couple of the servers at my house on extended battery backup after some extended power outages.

I've got a good source of car batteries, a U-Pull-It yard selling all for $25/each including core fee.

Obviously I'll do a quick open circuit voltage and drain test on each one at the yard, and not buy any manufactured too long ago.

Tried googling a bit but couldn't find any nice charts of amp hours by group size.

I know amp-hours can be highly varied by discharge rate, but I'd expect each battery to see 5A and go down to 11V before the inverter kicks out.

So what automotive battery group code would be the best capacity-wise? I doubt I'll find any L16s lying around...

Also can a server PSU be driven directly from 10 car batteries in series with no inverter? Or would DC stress two of the full bridge diodes too much?

Austin the Ozone

Sun Jul 17 2011, 04:32AM

Registered Member #3989 Joined: Thu Jul 07 2011, 05:10PM
Location: In a van down by the river.
Posts: 52

hey thats an interesting idea about the 10 batteries in series, obviously test it on an old computer.. You could maybe charge them directly from current limited and rectified line current.

Patrick

Sun Jul 17 2011, 04:35AM

Registered Member #2431 Joined: Tue Oct 13 2009, 09:47PM
Location: Chico, CA. USA
Posts: 5639

The traditional path:

[12v battery] boosted to [120Vac (Inverter)] then SMPS'ed down to the usuall [12V, 5V, 3.3V, -5V, -12V] isnt very effcient.

If i were you and had the time and money, i would make a [12Vdc battery] to the final PC voltages of
[12V, 5V, 3.3V, -5V, -12V]power supply. I think this would be worth persuing since you want "extended" power back up.

Just a thought, maybe im crazy.

Austin the Ozone

Sun Jul 17 2011, 04:50AM

Registered Member #3989 Joined: Thu Jul 07 2011, 05:10PM
Location: In a van down by the river.
Posts: 52

Patrick wrote ...

The traditional path:

[12v battery] boosted to [120Vac (Inverter)] then SMPS'ed down to the usuall [12V, 5V, 3.3V, -5V, -12V] isnt very effcient.

If i were you and had the time and money, i would make a [12Vdc battery] to the final PC voltages of
[12V, 5V, 3.3V, -5V, -12V]power supply. I think this would be worth persuing since you want "extended" power back up.

Just a thought maybe im crazy.

I think you would need two 12 volt batteries? The second one for the negative -12 and -5

Patrick

Sun Jul 17 2011, 05:07AM

Registered Member #2431 Joined: Tue Oct 13 2009, 09:47PM
Location: Chico, CA. USA
Posts: 5639

Austin the Ozone wrote ...

I think you would need two 12 volt batteries? The second one for the negative -12 and -5

Im not a big believer in linear regulation, id prefer a single 12v or 24v input to a SMPS (85% eff. is quite easy) then you nget much more duration, from much fewer cells, thus the optimized solution. *(normal polarity, then inverting polarity SMPS's can be made form single 12v input with 10 car batteries in parallel.)

(since the most cost comes from securing batteries, you want the most power for the fewest USD's $$$)

PS, i have been drinking the booze.

Bjørn

Sun Jul 17 2011, 05:11AM

Registered Member #27 Joined: Fri Feb 03 2006, 02:20AM
Location: Hyperborea
Posts: 2058

Better check that the PSU does not have a 115V/230V switch that enables a voltage doubler or something at the lowest setting. Then 10 batteries will not work.

Also it is possible that some PSUs will fail to work properly on DC because they have a standby PSU or something that relies on AC.

Patrick

Sun Jul 17 2011, 05:15AM

Registered Member #2431 Joined: Tue Oct 13 2009, 09:47PM
Location: Chico, CA. USA
Posts: 5639

BjÃ¸rn wrote ...

Also it is possible that some PSUs will fail to work properly on DC because they have a standby PSU or something that relies on AC.

i meant to say this too, most ATX type PS's seem to rectify to 300Vdc or 330Vdc (across main caps), so thats pretty high for car batteries in series. anything less and they may UVLO.

Sulaiman

Sun Jul 17 2011, 09:38AM

Registered Member #162 Joined: Mon Feb 13 2006, 10:25AM
Location: United Kingdom
Posts: 3140

Car batteries are not sealed and are not suitable for indoor use due to the possibility of spillage, corrosive fumes and explosive gas release.
Use the Yuasa type of VRLA batteries for safety.
OR put the batteries in an enclosure with good ventilation to the outside to reduce risk.

Ash Small

Sun Jul 17 2011, 11:02AM

Registered Member #3414 Joined: Sun Nov 14 2010, 05:05PM
Location: UK
Posts: 4245

Probably a stupid idea, but it 'should' be more efficient that the inverter/psu method.

four 12 volt (actually 12.6 V) batteries in series with centre tap gives +/- 25.2 V,

A string of eight electrolytics across 25.2 V gives 3.15 V across each electrolytic, use diodes to parallel.

A string of five electrolytics gives 5.04 V, again use diodes to parallel,

diodes from two batteries in series gives 12.6 V.

Current drain from batteries is equal.

Is 12.6 V, 5.04 V and 3.15 V close enough to 12, 5 and 3.3?

EDIT: I've not taken diode voltage drop, etc. into consideration here.

mikeselectricstuff

Sun Jul 17 2011, 11:06AM

Registered Member #311 Joined: Sun Mar 12 2006, 08:28PM
Location:
Posts: 253

Bear in mind that fusing and switching are harder at DC due to non self-extinguishing arcs.
PSUs with active PFC may get confused by a DC source

1 2 3 4 next

Moderator(s): Chris Russell, Noelle, Alex, Tesladownunder, Dave Marshall, Dave Billington, Bjørn, Steve Conner, Wolfram, Kizmo, Mads Barnkob