[Power] 3 portable solar choices for XO-1? (and other XO laptops, if poss!)

[Richard Smith]

On 01/14/2014 12:03 AM, James Cameron wrote:

> On Mon, Jan 13, 2014 at 11:17:31PM -0500, Adam Holt wrote:
>> 1) Our community deployments are re-ordering from Richard Yu @ Gold
>> Peak (GPI): “Portable Solar Module, Item #GPSP151400, 15V 14W solar

14W? That should be 15W.

>> panel with 18V limiter” currently used for the expanding XO-1
>> deployment in Lesotho [... 29°28′S ...] can anyone remind/clarify
>> whether the above works across all XO laptops (XO-1, XO-1.5,
>> XO-1.75, XO-4), or is the above unit only really for XO-1s?

Any GP panel 10W,15W or even a 20W with the 18V limiter will work for 
all generations of XO as long as its setup for "12V" operation.  In 
solar speak 12V means a typical output of 14V with Voc up in to the low 
20's.  Which is why you need the limiter for XO-1.

> I don't think I have one of these, so I can't say from experience.

I believe thats correct.  I sent you all unregulated panels so you could 
test.  I don't think I even have any of them either.  I have a sample of 
the 15W panel but I'm not sure if it has the limiter in it.

>> 2) We'd obviously like a suitcase-ready solar panel that works in as
>> many planetary situations as possible, aware this is unrealistic.
>> Still, Lesotho's high-altitude/cold/bright XO-1 deployment will test
>> a couple of AuriaSolar.com's rigid panels in adverse conditions
>> (cold/bright/high-altitude) in both Idaho/Lesotho in coming
>> weeks/months, just in case these can actually work with XO-1s, so we
>> all understand better.  Very inspiring they're putting their neck
>> out for all of us, if others have advice/warnings!  That Auria Solar
>> model being:
>>
>>     "15-watt-solar-panel for XO-1.5, 1.75 and XO-4"  "XO-1 with minor
>> constraints"

Here's where I need to insert a link back to where I've discussed this 
before but I don't feel like digging though the archives.  ON XO-1 there 
are 2 possible failure modes.

The first typically happens when you first connect the panel and the 
output voltage is above the point where the XO-1 overvoltage trips. 
This one can typically be worked around by letting the panel warm up or 
by flipping the panel over, connecting it to the XO and then turn it 
over and expose it to the sun slowly.  The XO will load the panel and 
the voltage will stay low.

The 2nd failure doesn't have a work around.  The charging profile for 
the battery is a curve that decreases.  In the latter stages of charging 
the load will only be few watts.  If the battery only draws 3 watts yet 
the panel has enough sunlight the produce more than 3 watts you will get 
a voltage increase up to the panels output voltage for the current 
operating temp.  If that resulting voltage level exceeds the overvoltage 
level then the XO-1 stops charging before the battery is 100% full.  The 
actual point at which this happens is very variable.

With a 15W (or higher) panel it much more likely to happen sooner than 
it did with the 10W panel.  So in theory it could result with an XO-1 
that was only 75% charged or maybe even 50%.

The 15W aSi panels are also more prone to this because the temperature 
de-rate for aSi is much lower than panels from monoSi or polySi.

Its a big math game with a bunch of variables to know if you will hit 
this.  But it I have reports from 3 different places that match this 
scenario.  One of which I was able to verify with olpc-pwr-log.


> You might also search for more recent modules.

On this note I've found a reasonable source for cheap polySi and monoSi 
panels that are on a metal backing. Very similar to the GP panel.  They 
suffer the same problem though with XO-1 in that without some sort of 
limiter they may trip the over-voltage.

I'll follow up with a post on these panels.  Been meaning to do that for 
a while anyway.

-- 
Richard A. Smith