Building a battery

I wanted to take some pictures yesterday, so I dug out our Canon EOS Rebel T7i camera, which hadn't been used in a year or so, and found that the battery was completely dead - like, so dead that it wouldn't take a charge.  The Canon batteries have a protection circuit - once the voltage drops below the "safe charging" threshold, the protection circuitry cuts off the battery from the electrodes, so it can't power anything and also can't be charged any more.

Looking on Amazon, I found a pair of no-name batteries which claimed to be replacements for the LP-E17 Canon battery, with a charger, for about thirty bucks.  "But wait!" I thought.  "I have all the parts to BUILD a battery that should work!  I bet I can design a case, 3D print it, and assemble a working prototype in a few hours... and turn IDLENESS and SLOTH into ELECTRICAL POWER!"

Sure enough, I had the most crucial part - two Li-Ion cells which would both fit within the form factor of the Canon LP-E17 battery, and provide the 7.2V (nominal) which the camera needs. 

OK, the guts were taken care of.  Now, about that case...  I've been 3d-printing stuff just long enough to have some facility with Fusion 360.  How hard could it be?!?  Turns out, between making measurements and the fact that I really am just a rank amateur at 3D design, it took about four hours.(including some "proof-of-concept" printing to make sure I could ACTUALLY make it fit in my camera before spending time getting it all "just right".)

After a little printing, soldering, and electrical tape (for safety!), I managed to put together something that actually fit!  I always get a little nervous playing with lithium ion cells, because they'll quite happily dump several amps worth of current into that short circuit you didn't realize you were making. 

 I managed to get everything assembled without shorting anything out, and a quick check with the multimeter showed 8.25V in the correct polarity, so... let's try it out!

It fits!  I can even remove it!  Even better, IT WORKS!

Now... do I recommend doing any of this?  Absolutely not!  This took far longer than I expected at the outset.  It also runs the risk of fire and/or damage to an expensive camera... all to save $30.  But it does feel pretty good to have set myself a task and completed it.  Now... what was it I wanted to photograph, again?

I've had a camera with CHDK firmware (a replacement for many Canon point-and-shoot cameras) on and off for a couple of years, but I hadn't really started to play with the scripting capabilities until recently.  I decided to start with time-lapse photography, as it's one of the easiest for people to understand.  There are a bunch of scripts written by other people readily available - and that's usually a good place to start.  (One hopes that they're already tested to a certain degree, which should help with figuring out how moving parts work.)

My first test was with the so-called Ultra Intervalometer, a UBASIC script which seemed to be a recommended way to get started.  It was relatively straightforward to get working, and I played around with it for a bit, but the first time I tried to get a time lapse of more than a few minutes, I discovered a flaw: the power in the camera's battery runs out fairly quickly.  With this very basic script, the camera's display remains on the entire time, and lots and lots or refocusing happens.

So, I did a little more searching, and came up with a Lua script called lapse.lua which not only supports turning off the display, but has pre-focus support as well.  I was able to get several hours out of the battery, and the results are below.



This is the moon setting in San Francisco, with nothing particularly special done.  One frame every 30 seconds, and then a quick editing pass with Final Cut Pro (though something lighter-weight could certainly have been used).

I'll continue to play with this script, and the capabilities of CHDK and my PowerShot SD780IS that it points out to me.  More later!