[SOLVED] Lithium Ion safety

My project includes Lithium Ion batteries in a headband, which the user wears as s/he sleeps through the night. And until now, I wasn't concerned about safety. But now with the uproar over Samsung's #7 phone problems, I'm deeply worried.

I use Lithium Ion for the combination of great features you can't get elsewhere: Small, powerful, long-lasting, light-weight, steady output voltage until drained, rechargeable, etc. But of course it would be absolutely horrible if a battery blew up in someone's face as they slept through the night.

My question is simple, what steps do I need to take to make sure no one gets hurt?

Use batteries that have a built in circuitry protecting against overcharge and overdischarge.

Avoid wired connections to the headband -- transmit data wirelessly.

If you get sued are you really willing to testify that you know your design was OK because of information you got from the arduino forum?

jremington:
Use batteries that have a built in circuitry protecting against ... overdischarge.

Thanks jremington.

I'm using the ZS-040 Bluetooth module for Arduino, which radios data to a tablet, so the headband has no external wires to worry about.

This is the battery and charger I'm using, so I'm not concerned about overcharging. Charging is done at the standard 110v wall jack, far from the headband or user.

Also, the battery states completely draining it before recharge will make it last the longest, and it does rather quickly fall to 0.0 volts at the end of 10 hours. So I think there is no danger of being over-drained.

But a short-circuit in the headband would be another story, discharging it far too fast, and therefore the one real danger. Do you have some specific information about avoiding this final risk?

KeithRB:
If you get sued ...

That's a good point too, Keith. But my concern here is a genuine interest in the safety of others.

It seems to me this is likely the best place to get honest and knowledgeable information in this area, from people who also care.

My point is to make sure to get corroborating information from the mfg. Unfortunately if you contact them directly they will probably tell you not to use their batteries for wearable medical applications.

That's a tough one because anytime you've got lots of energy you've got the potential for lots of heat. I had a short with NiCd batteries once and it melted solder.

If Samsung can't get it right, hobbyists like us are taking an even bigger risk. I'm not "afraid" of LiPo batteries, but I'm not sure about putting them around someone else's head while they are sleeping... I'd have to think about it...

We do have a couple of advantages - We can test each unit-project for several days or a week, whereas Samsung or Apple can only test for a few minutes before shipping-out to the customer. And, if one out of 10,000 batteries starts a fire that's bad-bad news for Samsung but we are never going to build 10,000 projects so the odds are in our favor.

But now with the uproar over Samsung's #7 phone problems, I'm deeply worried.

Did you miss all of the exploding/burning hoverboards last Christmas? :smiley:

P.S.
Don't buy your battery or charger off eBay or Alibaba!

KeithRB:
My point is to make sure to get corroborating information from the mfg. Unfortunately ... they will probably tell you not to use their batteries....

Ha ha, yeah! I've learned from past experience mfg. info is pretty worthless. It's based on what will bring them the highest profits with the lowest legal responsibility. They are like children who have been taught, "Just say 'No!'"

DVDdoug:
Did you miss all of the exploding/burning hoverboards last Christmas? :smiley:

P.S.
Don't buy your battery or charger off eBay or Alibaba!

Ha ha ha! You made me laugh too. I remember the exploding hoverboards, but since those boards were racing 150-pound kids up steep hills -- while I"m only sending a tiny signal 10 feet -- I didn't see much of a connection.

Thanks for the advice about where not to buy; that's a point I need to take seriously.


To reduce the short-circuit/over-current risk, I'm thinking to add a 100ma fast-blow Littlefuse.

Any advice on this? Like, should I use a larger fuse value to reduce voltage-drop across the fuse? How much larger is safe? What's the best fuse (or equivalent)?

Well to be fair, if the Samsung batteries had ended up in a medical device and caught fire there would be hell to pay.

Sorry...why use Li?

Is there an alternative method to using them?

Can the bluetooth be turned off and only used to transmit data say every half hour rather than leaving it on?

It seems a lower power system...so why the need for high current batteries?

CosmickGold:
Thanks jremington.

Also, the battery states completely draining it before recharge will make it last the longest, and it does rather quickly fall to 0.0 volts at the end of 10 hours. So I think there is no danger of being over-drained.

Your design runs the battery until it fails?

It is my understanding that discharging a LIon below 3V/cell will cause damage to the cell. And a discharge below 2.5V should be grounds for removal from service. Here is a good site with simple discussions Battery University.

Chuck.

Discharging an Li cell below about 3v could do it some damage, and may give problems with overheating on the next charge cycle.

Incorporate a low voltage test, and turn the load off below 3v

regards

Allan

From your mini bio:

"Now I'm working at home to create my own brainwave/sleep/dream analysis system based on Arduino. "

Care to enlighten on the actual current draw of the system? I am still trying to weight the advantage and disadvantages of NiMH vs Li-Ion in terms of safety, required current, packsize and weight.

Johnny010:
[1] ... why use Li?
[2] ... Can the bluetooth be turned off and only used to transmit data say every half hour rather than leaving it on?
[3] ... why the need for high current batteries?

(1) Light weight. Lots of power through 10-hour night. Rechargeable 1200 times.
(2) The Bluetooth sends 4 bytes of data, once every second. Response needs to be in real-time.
(3) Current appears to continuously vary from about 35ma to 80ma through the 10-hour night.

chucktodd:
....It is my understanding that discharging a LIon below 3V/cell will cause damage to the cell....

Very good to know, Chuck. Thank you for the education.

allanhurst:
Discharging an Li cell below about 3v ... may give problems with overheating on the next charge cycle.

Yikes! ...also very good to know.

jremington:
Use batteries that have a built in circuitry protecting against overcharge and overdischarge....

I didn't know any battery came with protection circuitry, but now I know to look for it. Thank you.

Johnny010:
Care to enlighten on the actual .... I am still trying to weight the advantage and disadvantages....

My headband uses only 3D-motion detection and an infra-red source and sensor. It does not make any electrical connection to the subject at all; there are no "electrodes".

The best description I've found of a "protected" 9v lithium ion battery is here, where it states that even 500ma continuous discharge is no problem. And that it has automatic cut-off to prevent (1) overcharge (2) over discharge (3) short circuit.

I've been unable to find such a full description for the battery I'm using described here. But it is evidently the same, as I did find mention of it containing a "protection ic", being widely used in medical equipment, officially certified by CE and ROHS, and able to be fully discharged without a problem (evidently due to a discharge limiter, explaining why mine stay nearly level around 7 volts for 10 hours and then fall to nothing rather suddenly).

So I'm assuming I don't need to add a fuse or other protection circuitry, as it's already built in. (...although if such protection is already in place, I'm left wondering why the side of the battery says not to "short circuit".)

4 AAA eneloop NiMH (750mAh) weigh in at like 60g.

A 2s LiPo for RC cars etc. is about 45g.

Is the 15g really that noticeable with all the other bits and pieces?

I finally located definitive information about the above battery, which I now feel safe to use:

The battery's home page says virtually nothing about any protection circuitry, nor have I gotten an answer to my email to them, or my online form message, of phone call to their simple answering machine. (Very frustrating).

But a review on this Amazon page was extremely helpful. It states it uses "the battery protection IC (HY2120-CB)".

I then found a datasheet for that protection IC, which explains: "These ICs are suitable for protecting 2-cell rechargeable lithium-ion/lithium polymer battery packs against the problems of overcharge, overdischarge and overcurrent." EXACTLY what I needed to know!

The above mentioned review also has a great image of the battery, cut open, showing precisely what's inside:

CosmickGold:
I finally located definitive information about the above battery, which I now feel safe to use:

The battery's home page says virtually nothing about any protection circuitry, nor have I gotten an answer to my email to them, or my online form message, of phone call to their simple answering machine. (Very frustrating).

But a review on this Amazon page was extremely helpful. It states it uses "the battery protection IC (HY2120-CB)".

I then found a datasheet for that protection IC, which explains: "These ICs are suitable for protecting 2-cell rechargeable lithium-ion/lithium polymer battery packs against the problems of overcharge, overdischarge and overcurrent." EXACTLY what I needed to know!

The above mentioned review also has a great image of the battery, cut open, showing precisely what's inside:

The data sheet actually says;

"Overdischarge detection voltage V 2.00 to 3.20V Accuracy ±80mV"

So what is the battery pack set up for ?

If its at the low end of the low voltage detection (2.0V), that is not good at all.

At what voltage does your circuit turn off the power ?

If its at 3.3V, that is good for the battery and safe.

The so called 'battery protection' circuits are there to protect the user from fire etc, in the circumstance that the devices own low voltage detection and power off fails.

srnet:
The so called 'battery protection' circuits are there to protect the user from fire etc....

Exactly! It's protecting the user from possible fire and injury, that has been my total concern and focus here.

But of course, you're clear explanation of why external circuitry is needed to meet the needs of the battery not handled by its internal circuitry, is valuable information for anyone who wants to get it right.

CosmickGold:
Exactly! It's protecting the user from possible fire and injury, that has been my total concern and focus here.

But of course, you're clear explanation of why external circuitry is needed to meet the needs of the battery not handled by its internal circuitry, is valuable information for anyone who wants to get it right.

Most Lithium Ion batteries have these circuits built in, a common cut off voltage is 2.4V. That is very low and you certainly dont want to be allowing the battery to go that low repeadly, its not good for the battery and dangerous.

There is virtually no charge left in the battery below around 3.3V, so set your device to power off when the battery gets that low and do not be tempted to ommit from the device this seperate low votage cut off.

srnet:
2.4V ... is very low and you certainly dont want to be allowing the battery to go that low repeadly, its not good for the battery and dangerous

In other words, what's bad for the battery is also dangerous for the user! The two topics cannot be cleanly separated.

Thank you for one more valuable lesson.

So, if no one has a better suggestion, I'll order the parts for the schematic below, lifted from this page.