Great testing! The results are somewhat discouraging on a couple of levels, but it is much better to know rather than to hope that it works without conducting a fair assessment.
I kind of expected the sort of results we are getting from the unmodified device (although perhaps slightly better performance from the battery). I pulled apart the second charger, and it is fairly obvious that the batteries they are using are reworked factory rejects, (or more probably recovered from recycling electronic scrap). This is not necessarily a bad thing, given the price and the fact that otherwise they would be landfill.
As you can see in this picture, the original battery protector strip has been removed, (that long, off white, sticky strip on the battery beneath the solder tabs looks like it once had the small circuit board with the DW01 plus and dual mosfets attached to it) and the capacity label has been defaced, either deliberately, as it said something other than 2600mAh, or it was damaged when removing it from whatever it was originally fitted to.
The LI42B 900mAh battery on the right is for size comparison. Based on size alone, I would expect much better performance from the larger battery in the solar charger. It looks about the right size for a 2600mAh battery. It also takes about 8 Hrs to fully charge the solar charger, which, given the standard DW01 Plus based charge circuitry employed, does suggest the battery is a bit higher capacity than the LI24B (which takes about 2 1/2 Hrs to charge).
With regards to the solar / buck regulator test: do you think the poorer performance was related to the battery quality or the buck regulator? You recorded a 4.16 voltage, and that's still enough to power the camera. What's the dropout voltage on that unit? Was it not letting the battery get drained beyond that?
I have yet to determine that, but I suspect it is down to the battery. So far as I understand its operating characteristics, the buck regulator will try to maintain its output until the input dies. It may not be particularly efficient however as the input voltage is pretty close to the output voltage, and this is not the most efficient way to run a buck regulator. Lets see if we get better results with wires straight from the 3.7V internal connections. Bear in mind we have two regulator stages here, we convert 3.7V to 5V then back to 3.7V and this is bound to use a fair amount of power.
With regards to the real Canon battery: you get what you pay for it seems. Much better return and the battery was drained down to similar levels that I saw when I tested the Ultimate Intervalometer script a couple of weeks ago.
Couple the Canon battery with a decent solar charger and then you're cooking! To wit: I've finished up my version 1.0 time lapse camera rig last night. It's out for it's first field test starting today until.... I've got solar charging a lipo and an off-the-shelf buck / boost converting mediating the power supply. I'm waiting on some 25K trimmer potentiometers so I can finish the low dropout regulator. I'm running the Ultimate Intervalometer v1.7 for as long as possible. I've made it through 8 hours of bright, direct sun, which is much much longer than the rig runs without the buck/boost or an LDO. Fingers crossed! I've attached a couple of pics below. The enclosure is made out of hardware store grade PVC. It's a 4" cap, coupler and pipe, and a bunch of electronic guts that I will describe in another post.
I look forward to seeing the results.
EDIT: In case you were wondering how I was going to produce a suitable board to use the Mosfets I mentioned previously...
Are you physically able to surface mount those MOSFETs? You have the hands of a brain surgeon. I spent 10 minutes undoing some lousy through-hole joints I soldered up last night.
Brain surgeon, I don't think so, don't ask me to open up your skull, your next of kin would be disappointed by the results. There is a trick to this (and I bought a bunch of mosfets for a reason).
Drag soldering. The trick... use quality flux, drag soldering and a large magnifier, (and probably several crude Anglo Saxon phrases), followed by isopropanol and an old toothbrush to remove the excess flux. I expect at least one failure on the way.
Reworking the Solar charger may actually be trickier than soldering the mosfets, as there is not a lot of room there, but the technique is the same, flux (and lots of it), a magnifier, and most important of all, a freshly cleaned tip on your soldering iron, use a dollop of squeezed wet tissue paper, quickly wipe the tip on the damp tissue and, no you wont electrocute yourself with the wet tissue paper.
Most soldering stations have a special small sponge for keeping the tip clean. Don't use plastic sponge, it will melt and make a terrible smell.
Keep everything taped down to the workbench to stop it moving around, since your shaking hands will be doing enough moving of their own
.
Youtube has lots of good videos about soldering technique, the most important trick is to keep the soldering iron tip clean. Keep cleaning it obsessively with damp paper (and when it gets very grubby, a little flux, but this tends to eat away the tip) as you go along.
EDIT: Here are the test results for the "5000mAh" battery pack.
shot:767 12:46 4.26VSo it is better than the 1120 mAh Canon LiPo, but not the "four times as good" I might have hoped for. I forgot to bring home the Gold solar charger with the scruffy battery, pictured above, so you will need to wait till probably tomorrow night before I put that on test. Meanwhile the black solar charger is back on the test rig so we can average the three results.
EDIT: Original solar charger, third run
shot:480 07:59 4.17V - Gold coloured charger has just gone on test. I should have the first result in the morning (around 08:00 GMT)
