That's about 2 watts out of the peltier device, which seems reasonable, and 77% efficiency for the lvboost, which is also reasonable.

Are you measuring voltage and current at the same time?

What are you using as an electrical load on the device while you are measuring the voltage and current? Do you know what its resistance is?

When you measured 1.7 volts with your multimeter, it sounds like that was the open circuit voltage. (Voltmeters have very high resistance.)

If you measured current the same way, meter directly connected to power source with nothing else in the circuit, that is the short circuit current. (Ammeters have very low resistance.)

No device can supply full short circuit current while maintaining open circuit voltage.

Any results?

What I don't understand is why a PET. It's one of the most difficult things to get heated/cooled. Also, you'll nuke it if it's not vented enough, the solder joints will likely fail is you just toss it in a fire.

If you don't want to be dependent on the sun, other sources like heat, etc, why not use a manual charger, you know, the kind with a handle crank (My English fails me. Dinamo?). IIRC, there are some out there that put out quite a bunch over 5 volts, so simply soldering a LM7805 in there will give you stable five volts.

There are devices out there that do this, such as a crank 12-volt car lighter socket you can just use a cable with.

Dismantling a handle lantern is also a good option, some actually have a capacitor inside (or a small battery) that can sustain the output a few seconds while you rest your hand. It's light, reliable and works just as well in the dark, cold, mechanical vibration, etc.

But then again so does a spare battery.

Yeah I was measuring short curcuit current. I guess I should use a battery pack that I can charge with solar or teg. If I used 4 1.2v AAs with 2 in series the lvboost should be able to output 500ma at 5v. Anyone know if the N900 draws 500ma from usb or will it try to draw 1200ma like a wall charger puts out?

Maybe I should put all batteries in series and get 4.8, volts then I would be ok as long as the N900 doesn't draw over 1200ma.

based belong to feeling in my arse:

you would need something like constant 1,5V 4A from your thermoelectric device

Amps aren't "drawn". The 500 mA limit is at the source (as is the case everywhere), regardless. However, the charging strategy should change to adjust - First, there's the multi-stage charging of the battery (Li-Pos are mighty temperamental). Second, there's the charge minimum - if you don't provide as needed the battery will discharge instead of charging. But the phone thinks it's charging so there's a real possibility to over-discharge it.

And finally, to charge from an USB at 500 mA you have to negotiate the draw with the USB hub. That is, you only get about 100 or so unless you identify yourself and report you need more. Letting N900 try and charge at 100 (via bad cable, for example) will NOT charge, because draw is larger than charge, destroying the battery (eventually).

The wall wart (dumb charger) has shorted wires, signaling the N900 that it is a wall wart. As a result, the N900 starts a charge is it is satisfied with the conditions. That is, if you use a battery pack that can't deliver it will cut the charge short and give you the finger (as reported several times). The fact that it distinguishes them means it has different strategies (will only charge if shorted and satisfied, or if has a hub on the other side even if it can't deliver).

Additionally, N900 is a very sensitive device, as it is quite complex. Not delivering, poor delivery, noise, etc can all do bad, bad things to it, starting with hanging and end of charge condition, ending with root FS corruption and nuking the battery. Your solution to adapt varying power to it directly is likely to have repercussions.

IMO, you need at the very least two stabilization stages to pull it off, or a self-correcting DC-to-DC, both of which require higher voltages and some assembly. In fact, without a way to guarantee that the charging system will have the needed input voltage and plenty of power, any smart Li-based charger will give up mid-charge, at the very least eating up charge cycles.

Appreciate the ingenuity, like the idea, not a good real life solution. My 2 cents.

As a stopgap, use batteries but if you carry all those with you it's easier to get a travel charger - it's lighter and more compact.

APC sells 4-AAA pocket chargers with USB. They will NOT charge the N900. Will charge other Nokias and older iPhones.