Has anyone done any work on powering an Electron with Li-SOCl2 primary-cell (i.e. non-rechargeable) batteries? I saw one discussion that was talking about AA batteries, but leaving the LiPo battery in place. That won’t work for my application, thus the Li-SOCl2, which provides 3.6V of power.
The trick is that unlike LiPo, it is not “1C” (meaning that the battery’s total current supply is significantly less than its capacity). One possible solution to the very power-hungry cellular radio (which would, of course, be off most of the time) is to add a relatively large electrolytic capacitor in parallel with the battery (which could be a “supercap”, but may not be necessary for my intended duty cycles). In any case, I’ve not personally tried this design approach, so it would be new territory.
I’d be interested to know if anyone else has done this or has come up with another solution.
Just power the Electron from that battery and monitor the voltage drop to see how much voltage sag you get.
How many mAh is the battery your planning on using? What are the specs on it?
They do make voltage regulators that are designed to buck or boost the voltage out of the battery so the Electron sees a stable voltage regardless of the battery voltage. This also allows you to suck all the power from the battery possible before it’s completely empty.
I definitely plan to use a buck-boost converter like the TPS63031 or LM3281, but I don’t believe that this alone will solve the problem of peak power demand during 3G transmission. According to my calculations, the battery by itself will simply not deliver the current needed by the Electron’s radio.
This isn’t a problem for LiPo batteries because they are “1C”, meaning they can supply as much current as their capacity rating (i.e. a 2000mAh LiPo battery can supply 2000mA of current–continuously). But the Li-SOCl2 can not. The one I linked to has a 1200mAh capacity, but is only capable of supplying 30mA continuously and 100mA peak, which according to the Electron specs, is not enough to power the radio during transmission periods. There are special Li-SOCl2 batteries that also include a “supercap” to handle applications that have relatively low continuous current demand and periodic peaks (like devices with radios), but I don’t see any reason to use such a component, since I can easily add the cap myself.
I need a non-rechargeable battery with a very low self-discharge rate that has no transport restrictions, so that leaves LiPo out. I’ll look specifically at Li AA batteries. Do you have any suggestions on those?
No, it’s a sensor that only needs to wake up and send a very short message once per day under normal operation. It needs to last at least 3 years without anyone having to touch it. From a total capacity vs total demand, this is possible with the original battery I mentioned, except for it not being able to provide the peak current required for the radio.
The “Ultimate Lithium” would certainly do the job on that front, but so far it looks like I’d need two of them in order to use my existing converter, and I only have room for one. I’m looking now for a 3.3V output buck-boost converter that will work down to a ~1.3V input.
Yep, me neither (not at the current output needed–TI did have some options for single-cell Lithium, but at a max of 300mA). I simply don’t have room for two AA cells (barely room for one). Ugh.
I doubt a single supercap would help you either. They can certainly provide larger currents, but the response time of these isn't that great and might be too slow for the rapid cellular demands (at least for 2G, 3G would be on the border) on connect, so you may need some extra, smaller electrolytic caps to increase response time and all that might again counteract your demands for least self discharge of the whole setup.
BTW, where are you delivering the product to and how?
AFAIK, single LiPos as inbuilt parts of devices get lower risk rating on transport than isolated batteries and/or multiple cells.
You should be able to send the device with a LiPo in it with just with a Li-Ion sticker on it just like Laptops and Cellphones are shipped.
It’s based on the size the battery in watt hours and your project should be well below those ratings.
Remember though that LiPo cells self-discharge at a rate of 3% per month and even quicker at the temp increases. You can offset this by adding a small solar panel and ultra low voltage boost converter to keep small amounts of charge into the battery. There are many low power sensor designs using this type of setup to last for years and years running on very low power.