# LiPo heater for cold weather conditions

Discussion in 'The Projects Forum' started by Saphira Dragon, Jan 28, 2016.

1. ### Saphira Dragon Thread Starter New Member

Mar 5, 2015
10
0
Hello out there,

I'm working on a mobile device which is to be used under cold weather conditions. For convenience I'm using LiPo batteries to power this device. However, LiPo batteries discharge much faster below 12-10 degree celcius. Since I expect temperatures down to -20 degree and the device not producing enough heat itself, I'd like to heat the batteries while in use.

I wonder if someone attempted something similar already and has some experience with heating LiPo battery packs and has some pointers as what to take care of.

I'd like to have used an ATMega406, as it has cell balancing FETs, I2C and PWM outputs for heating, unfortunately it's not working below 0 degree C. So I think a Max6680 for temperature control and a Max2775 for battery protection along with a small ATtiny for heater control might be a way to go.

What would you think?

2. ### Papabravo Expert

Feb 24, 2006
10,021
1,757
There is something I'm missing here. If you use part of the battery capacity to produce heat to slow the discharge, won't that decrease the time to discharge below the usable voltage threshold. Maybe I missed something and you have a plan to deal with this problem.

3. ### hp1729 Well-Known Member

Nov 23, 2015
1,846
206
Something powerless like a hand warmer (chemical) pack?
Otherwise a big power resistor on a heat sink powered by separate batteries?
How much heat do you need?

4. ### Saphira Dragon Thread Starter New Member

Mar 5, 2015
10
0
Powerless heating would be nice, however it's not what I would like. The device is closed and the battery is fixed.
@Papabravo: You are absolutely right - it will decrease the discharge time. However, the capacity used for heating is much less then the decrease of capacity due to temperatures below 10 C. Therefore it would still increase the usuable time.

@hp1729: I don't need much heat. Just enough to keep the pack above 12c in conditions around 12c to - 20c.

5. ### wayneh Expert

Sep 9, 2010
11,920
2,855
A peltier panel might make a convenient heater, simply due to the form factor. At low current, it will have little risk of getting too hot. I don't think you'd need feedback control as long as you can control the current to a level well below its rating.

That said, I'm not sure I believe that heating from the internal resistance of the battery alone would not do the job. Maybe it's a matter of limiting the load until the battery warms up?

6. ### Papabravo Expert

Feb 24, 2006
10,021
1,757
So it seems like you could fabricate a heater to meet your needs from some resistance (Nichrome) wire, a constant current source, and some insulation so that the heat you do produce does not try to heat up the surroundings. You adjust the current source to produce just enough heat to meet your requirements. There are tables you can use to design heating elements but they start at 400°F which is a bit too hot to handle safely. My thinking is that you want something that will be at 16°F when the surroundings are at -20°F so you may have to experiment with this range. Maybe by graphing the data from one of the tables you can see if the data may be extrapolated. It is always dangerous to assume that you can extrapolate because the behavior may not be linear.

http://cecs.wright.edu/balloon/images/2/22/Nichrome_Wire_Heating_Element_Design_Basics.pdf

BOTE Calculation
Ni CR A #20 AWG has a resistance of 0.6348 Ω/ft and takes 3.8 A to raise the temperature of a straight section to 400°F. Now
$(3.8 \text Amperes)^2\cdot 0.6348 \Omega = 9.17 \text Watts$
$400^\circ \text F-70^\circ \text F=330^\circ \text F$
$\frac{330^\circ \text F}{9.17 \text Watts} \approx 36\frac{^\circ \text F}{\text Watt}$

So let's say we want a constant current source to put one watt into a load of 0.6348 Ω.
$\frac{1 \text Watt}{0.6348 \text \Omega}=1.58 \text Amperes^2, so \sqrt{1.58}=1.257 \text Amperes$
to produce a temperature rise of 36°F. That seems like a large current draw for a battery to sustain for any length of time. So lets make it four feet and let it be looped back and forth like the roads up the side of a mountain. Now we still want 1 watt into a load of 2.54 Ω. So
$\frac{1 \text Watt}{2.54 \text \Omega}=394 \text millamperes^2, so \sqrt{0.394}=0.628 \text millamperes$

So maybe #20 AWG Ni Cr A is not going to work for us. You see where this is going. We need to find a piece of nichrome wire that will give us a safe temperature rise when connected to a constant current source, not drain the battery too rapidly, and not require too long a length. Do you have enough to proceed on your own?

Last edited: Jan 28, 2016
7. ### Saphira Dragon Thread Starter New Member

Mar 5, 2015
10
0
Wow! Many thanks for the formulas. They help a lot. Yes, I think I can continue on my own. Oh the lenght is not much of a problem, as the wire will be wrappend around / folded between the packs anyway.

8. ### Papabravo Expert

Feb 24, 2006
10,021
1,757
Whatever you do, safety should be your number one concern. If the resistance wire gets too hot there could be serious adverse consequences not only from the wire but from the batteries as well. I'd really hate for you to win a Darwin Award.

GopherT likes this.
9. ### Dr.killjoy Well-Known Member

Apr 28, 2013
1,164
150
How long do you plan on using the unit in the cold ???
I was just think about using a insulated pack with just cheap hand warmer to keep it warm or maybe those reusable hand warmers..

10. ### Saphira Dragon Thread Starter New Member

Mar 5, 2015
10
0
Of course. Safety is the first and foremost concern. It would be rather bad, if it spontanously combusted in an enclosed space.
@Dr.killjoy The device will be used for a few hours in cold, I expect between 2 to 4 hours.

Finally most of the parts have arrived and we can start designing the PCB. The batteries were hardest to find. Does anyone know a good and inexpensive supplier for LiPo batteries?

11. ### GetDeviceInfo Senior Member

Jun 7, 2009
1,571
230
What rating / quantity?

12. ### Bernard AAC Fanatic!

Aug 7, 2008
4,140
393
Sounds like my fish bowl warmer. User a matrix of 220 ohm resistors to give about 3 W with temperature control provided with thermistor, comparator & FET. Schematic & B's long gone.

13. ### Saphira Dragon Thread Starter New Member

Mar 5, 2015
10
0
That would depend on the project. For now I got 3.7v, 3000mAh flatpack cells from Olimex.

14. ### GetDeviceInfo Senior Member

Jun 7, 2009
1,571
230
I have a large lot of Sierra wireless w-2, 3.7 @ 3600 mAh. Border shipping might make it uneconomical.

15. ### wayneh Expert

Sep 9, 2010
11,920
2,855
If you want maximum warming for your precious battery capacity, you'll abandon resistive heating in favor of the peltier.

16. ### GopherT AAC Fanatic!

Nov 23, 2012
5,836
3,432
How is Peltier better. Resistive heating is 100% efficient at converting power to heat (as long as no light is produced).

17. ### wayneh Expert

Sep 9, 2010
11,920
2,855
The peltier also converts all supplied power to heat with 100% efficiency. But it can also move ambient heat from the cold side to the hot side, like a heat pump. This is reduced at cold temps, but even a little "free" heat will extend battery life compared to pure resistive heating.

It's not an option if delta T exceeds what the peltier can do.

[update] I just re-read the TS's needs and I don't think a peltier can help with -20° on one side and >12° on the other. It could still be a handy form factor for a resistance heater, but it won't be able to move heat from hot to cold with that large a ∆T.

Last edited: Feb 24, 2016
GopherT likes this.
18. ### Sensacell Well-Known Member

Jun 19, 2012
1,102
255
Hmmm?

If the battery is REALLY well thermally insulated, it seems that a simple on-off setpoint temperature controller that did not kick in until it got below 12C would work. You could make a dirt cheap simple heater out of a piece of 1/32" circuit board with a bunch of surface mount resistors mounted on it.

Choose the resistor values to generate the minimum amount of heat required to keep up with the thermal leakage of the insulation.

If the circuit had a very low standby current, you might even get somewhere with this idea.

Bernard likes this.
19. ### Dr.killjoy Well-Known Member

Apr 28, 2013
1,164
150
Hobbyking has a nice selections on lipo's and for a cheap price ..