[Power] $30 resin panel in my hands from eBay

[Richard A. Smith]

On 06/20/2013 04:51 PM, Mike Lee wrote:


> Here are some initial photos:
>
> http://www.flickr.com/photos/curiouslee/sets/72157634232528701/

Probably won't matter for you for testing but the wire contact points 
and a short bit of the wire should be coated with something like hot 
glue or some other thick coating. If there isn't some sort of strain 
relief I expect those connections will fatigue and break with a moderate 
amount of use.

Electrically there isn't a whole lot of concern but if you happen to set 
it down on some sort of surface thats conductive you may lose some power 
that way.  I can image wet mud in a coastal area might have enough salt 
or other ions it it to be conductive enough to steal power.

>     through you swap the panels so that each XO gets run from each panel.
>
> Never would have thought to do crossover.

>     the XO into the lowest possible power mode that you can still log data
>     from.


> Yes, olpc-panelpwr-log. It took a couple extra pokes in Google to find
> that zip package on GitHub.

Github? eh?  I don't really know what you are talking about.  I've never 
put anything on GitHub (That I know of, perhaps theres a mirror?)

  Sorry. I though you knew where things were located already.

Upstream git repo is here:

http://dev.laptop.org/git/users/rsmith/olpc-pwrlogs/

Copies of the scripts are here:

http://dev.laptop.org/~rsmith/pwr_scripts/

Sometimes whats in my pub html dir isn't quite the latest but I try to 
keep know working versions there.


>     until they charge up or your sun dies.   A voltmeter on the panel
>     output while connected would be nice too especially if you have one
>     that can log readings.
>
> The meter I currently have is pretty basic. But I will have my intern
> record some readings off the two solder points of the panel using some
> alligator clip jumper wires.

Sounds good.  Logging every 5 minutes or so should give an approximate 
idea of whats going on.   2 readings you want to get that are important 
for XO-1 are Voc (thats open circuit, ie nothing connected) when the 
panel is both cold and after its heated up.  So I propose you keep the 
panel covered with a towel or something connect up your voltage 
measurement setup and get it all ready.  Then expose it to the sun and 
quickly take the reading.  That will be Voc cold.  If you are curious 
you can watch it. Voc should fall as the panel heats up.  The 2nd Voc 
you want is after its been in the sun for a while and is good and hot.

The alignment to the sun and amount of haze in your atmosphere will make 
a difference in your readings so you really want to do to this multiple 
times.  Say at the beginning and end of each test cycle.

> We'll make it standard practice to place the panel face down before
> plugging it in. Then flipping it over to start the test.

I think a better test is without doing that and see if your XO ever 
trips the overvoltages.  You can tell that pretty easy with the 
voltmeter connected.  You should see a large drop in voltage when the XO 
starts to draw power from the panel.  If you don't see that then the 
overvoltage protector kicked in.

Its a test for a negative though so even if you never see it that 
doesn't mean it won't happen for others but if you do see it then you 
know it will be a problem for XO-1 users.

Voc on solar panel is maximum when the panel is cold. So if you want to 
see how high it can go put it in the freezer and get it really cold then 
quickly put it in the sun.

I never saw the OV problem on at first connection with the non-limited 
GP panels until I was testing in the bright sun on a cold day in winter 
(IIRC around 0 deg C).

I've never seen the other end of OV problem because I don't think I've 
manged to test in strong enough sun.  Its was reported by several 
tropical users though.  Once the battery charging reached the constant 
voltage stage the power into the battery falls off pretty quickly.  Over 
85% or so of the battery charge will have completed by that time but its 
bothersome to never see it reach full charge.

-- 
Richard A. Smith  <richard at laptop.org>
One Laptop per Child