3
$\begingroup$

I am building a robot where power density is critical, and looking into using lithium thionyl chloride (SOCl2) batteries. I will be drawing around 20A constantly and need between 12 and 17V. The batteries I have found so far are AA-sized and designed to deliver 100mA, 3.6v, 2.4Ah, and weigh 19g each. I could picture a pack with 4 blocks of these batteries in series, where each block has 20 batteries in parallel. That would mean 48Ah, which is way more than the ~10Ah that I need, and 1.52kg, which is more than I would like the robot be carrying.

So, the question is, is there a way to achieve 10Ah at 20A and 14.4V (i.e. for 5 hours) using SOCl2, carrying much less weight than 1.52kg?

$\endgroup$
4
  • $\begingroup$ You are wrong re 192000mAh. Note that 20 ea. 3.6V 2400mAh batteries in parallel give 48000 mAh, or 48 Ah, or 172.8 Wh. Four such blocks in series has four times as many watt-hours, ie 691.2 Wh, but still only 48 Ah. ¶ Also, specify how long your robot needs to run with 20A @14.4V, or specify total Wh requirement $\endgroup$ Jun 29, 2013 at 5:51
  • $\begingroup$ Oops. Fixed now. $\endgroup$ Jun 29, 2013 at 13:03
  • $\begingroup$ Do you really need to use the type of batteries you wrote, or do you have the possibility to use another type? For example, using 2x4 cells of these: hobbyking.com/hobbyking/store/… you will get to 10Ah and about 1kg of weight, plus it gives you really big discharge current and weighs around a kilogram. $\endgroup$
    – Plecharts
    Jun 29, 2013 at 19:37
  • $\begingroup$ So, I wanted to use SOCl2 because they reportedly have higher energy density than LiPo. But, I had a bit of a d'oh moment -- I have a 6.6Ah 4s lipo on my desk which weighs 800gr. So I guess when you take into account the casing of these SOCl2 cells, they are in fact NOT higher energy density. $\endgroup$ Jun 30, 2013 at 14:59

1 Answer 1

2
$\begingroup$

Lithium thionyl chloride / SOCl2 batteries have excellent energy density but are not aimed at high-current applications and are not rechargeable, so it seems likely that you can save weight and cost by using some other kind of batteries.

Quadcopter batteries (typically lithium ion polymer technology, rated for 5C to 15C discharge rates and rechargeable hundreds of times) might be a better choice. You could use half-a-dozen 2200 mAh packs, or a couple of 5500 mAh packs.

Previous question Quadcopter liPo battery weight/capacity trade off includes some charts comparing energy densities, etc., for several battery technologies.

Note, if your application is controlling several motors, with several motor controllers, you may be able to use higher voltages to the controllers, which use PWM to deliver desired amounts of energy to the motors. This technique is usable when you have a high-voltage, low-current power source. However (see below) that doesn't help in the present case. Likewise, a DC/DC buck converter (that converts high voltages to lower voltages, at ~ 95% efficiency) doesn't help.

The 20A, 10Ah parameters of your SOCl2 example imply 30 minutes of operation. The 20A, 14.4V parameters imply energy rate is 288 W. The current drain per cell in the example is 1A (which exceeds the rating of a typical AA SOCl2 cell). If we limit per-cell current to 1A, we get 3.6 W per cell. Then the number of cells required is 288/3.6 = 80 cells, regardless of the cells' series or parallel configuration, which is why a higher voltage to the motor controller, or to a buck converter, doesn't help.

$\endgroup$
1
  • 1
    $\begingroup$ So, I fly quadcopters and have a bunch of LiPo. I was hoping I could have a significantly longer flight with a custom built SOCl2 pack. I guess I'm satisfied with your answer which supports my research that essentially, no, I can't, at least not with the AA-packaged ones. $\endgroup$ Jun 30, 2013 at 15:03

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.