Author Topic: Single cell lipo protection circuit with replaceable battery design help  (Read 2055 times)

0 Members and 1 Guest are viewing this topic.

Offline vis321Topic starter

  • Contributor
  • Posts: 14
  • Country: ca
Hello,

I am currently working on a portable project which used a single cell lipo.

Without giving too much away, it's simple device that controls a few high power leds controlled though a esp32(via mosfets of course).

I need to implement battery protection circuit from over discharge and potential short circuit.

My battery is a tiny 3.8v LiHV (300 mah 45c), tested works fine.

I am boosting it to 5v via a TI tps61022rwur, that's all good, my device pulls MAX 3 ish amps for no more then 1-3 seconds (flash led).

I had a few issues when the battery gets low(empty), in software that I wrote for the esp, its measure the battery voltage and is supposed to shutdown the MCU when it gets low, still had issues when it was low battery. Instead I want to also implement a dedicated IC.

I came across a Ti BQ29700EVM-610 Eval board which uses a BQ29700 chip, the eval board work perfect for my application short circuit and UV protection work perfect. Was just going to copy the dev board schematic with same parts.

MAIN POINT:
Issues I ran into which was only mentioned on one page on the TI form and only on page 16, that unless that chip sees a charger first or V- is shorted to VSS temporarily it wont power on when I swap out a battery.... (for the first time when the chip gets powered).

Datasheet, http://www.ti.com/lit/ds/symlink/bq2970.pdf
Eval board datasheet, http://www.ti.com/lit/ug/sluuaz3/sluuaz3.pdf

That's no good for my application when switching batteries out.

Is there another chip someone can recommend for my power requirements? another idea I thought was to implement a one shot timer with a 555 that would briefly toggle the protection chip on (but it would need to work down to 3v...), am I going the wrong way about this?

If  V- is shorted to VSS the circuit works but then the protection does not kick in (must be in always reset mode)....
 

Offline Marco

  • Super Contributor
  • ***
  • Posts: 7081
  • Country: nl
Can't you just put a 100K resistor across the protection FETs?
 

Offline amyk

  • Super Contributor
  • ***
  • Posts: 8533
A standard DW01 circuit may suit your application.
 

Offline vis321Topic starter

  • Contributor
  • Posts: 14
  • Country: ca
amyk, I would need it work work upto 4.35v lipo cells as we are using LiHV which can be charger to higher to 4.35v instead of regular 4.2v lipos.

Marco. do you mean across the gate of Q1 mosfet? This is the schematic of the dev board I am using, I did try a 100k ohm from TP4/Q1 gate/DOUT to BATT+ which of course did enable to output/ but that would disable all the protection from the output.
950556-0

What if I was to implement a LMC555 as a one shot timer to trigger the protection on and the LMC555 would be low current enough not to kill the lipo as it would work with lower voltages ? thoughts?
« Last Edit: March 16, 2020, 07:55:30 pm by vis321 »
 

Offline amyk

  • Super Contributor
  • ***
  • Posts: 8533
 

Offline vis321Topic starter

  • Contributor
  • Posts: 14
  • Country: ca
amyk, Yes I do see that in the datasheet, but unfortunately I don't it available on digikey or mouser, or even lcsc, trying to stay to more common parts available from mouse or digikey at the moment.
 

Offline SuzyC

  • Frequent Contributor
  • **
  • Posts: 816
You should not  have to add another ic to detect a low-battery condition. You just have a software problem.

Keep it simple.

You've not mentioned what problem you had using the mcu to detect low-battery voltage.
 

Offline vis321Topic starter

  • Contributor
  • Posts: 14
  • Country: ca
SuzyC,

I have implemented a software detection as well, but I don't want to rely on that. I want to implement a hardware solution that would disconnect the battery.

Any other readily available Lipo protection chips you can recommend ?
 

Offline kjr18

  • Regular Contributor
  • *
  • Posts: 208
  • Country: pl
You could look at AP9101C, it probably worked for user in this thread
 

Offline vis321Topic starter

  • Contributor
  • Posts: 14
  • Country: ca
amyk,

I actually did find some small boards that I order a while ago that used the dw01b. Now the boards that I have actually are rated for 2.5A which was not enough when I tested it, so I put two boards in parallel. and it worked perfect, short circuit protection and UVP. so awesome. Now my question is. I took a look on banggood where I originally bought that board, and I see 4A rated one. I am trying to read the datasheet. I don't quite understand how the CS resistor is calculated? because I'm sure I can just use a single chip just with bigger mosfest?

I am use that is all I need to do is change the resistor value for the higher current and bigger mosfets ?

The two boards I put in parallel for testing: https://bit.ly/3biHmMH

The 4A version of the same chip: https://bit.ly/33yiiPa
 

Offline amyk

  • Super Contributor
  • ***
  • Posts: 8533
See section 12 of the datasheet. The Rds(on) of the MOSFETs sets the current limit.
 
The following users thanked this post: vis321

Offline vis321Topic starter

  • Contributor
  • Posts: 14
  • Country: ca
Thanks for your help,

So ok, after reading into more how to IC detects and samples for over current discharge and all that. As well as much better availability. I am going to try and fit in the bq297002 still. which has a UVP of 2.8v and a OCD of 0.160 V. Given that my boost converter takes upto almost 10 amps on when a new cell is inserted. I figure a 10-12 or so amp cut off limit could be safe? in the datasheet on page 20, In the example,  for 5a, they give an upper limit of 7a, they use the bq29700 which has a 0.100v cut off. So 100 mV/7A = 14.3 Ohm (would need for RDS on?).

Also few clip snap shots of the scope when a new cell is inserted, I used 3 1.5 ohm resistor in parallel for a quick test,(0.333 Ohm), and after a little ohms law I got almost 10A (9.369A), for roughly 50 uS, ODC protection kicks in, OCD delay is 16 ms...

So For my application lets take the bq297002, OCD of 0.160v, I calculated about 16 mOhm for my application. (at 10 amps, 160/10).

So I should be on the scope out for a low RDS on mosfet of around 16 mOhm? with a low of like 3v or lower(2.5v? lowest). And most likely will need to double up on the mofest (2 output and 2 input) to get my 10 amps of current? Any suggestions on mosfets?

For example I was looking into CSD16301Q2, might be a bit over kill:
http://www.ti.com/general/docs/suppproductinfo.tsp?distId=10&gotoUrl=http%3A%2F%2Fwww.ti.com%2Flit%2Fgpn%2Fcsd16301q2

Few other FET's I was looking at as well.

Alpha and Omega:
AON2406
http://aosmd.com/res/data_sheets/AON2406.pdf

Rhom RQ6E080AJTCR
https://www.rohm.com/datasheet/RQ6E080AJ/rq6e080aj-e

RF6E065BNTCR
https://www.rohm.com/datasheet/RF6E065BN/rf6e065bntcr-e

RF4E100AJTCR
https://www.rohm.com/datasheet/RF4E100AJ
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf