Maker Pro
Maker Pro

Protection circuit for Li-ion batteries

sweben

Jul 3, 2015
18
Joined
Jul 3, 2015
Messages
18
Hello,

I've got a bunch of 18650 3.7V li-ion batteries lying around, that I would like to use for connected sensors.

These batteries are fully charged at 4.2V, and I have a nice charger for them. The issue I have is that I need to make sure they do not discharge further than 3V. For that I designed a simple comparator that would cut down the power when the batteries reach 3V.

Here is the circuit I came up with
7pj6aH3.png


It is working great, the voltage divider divides the battery voltage by a factor of 10, then I can setup the threshold voltage with the potentiometer.

My problem is that when I put a charge on the connector, the current is not sufficient and my sensor does not start. (voltage drops to ~1.2V).

I'm quite unsure of what my issue could be, would you help me out?

Cheers,
Benoit
 

Arouse1973

Adam
Dec 18, 2013
5,178
Joined
Dec 18, 2013
Messages
5,178
The label VBAT, is this from your battery? If it is then the regulator will stop working when the input voltage reaches it's lower limit, what ever that is, you don't say what your using. How much current does your sensor draw?
Adam
 

Alec_t

Jul 7, 2015
3,591
Joined
Jul 7, 2015
Messages
3,591
If Vbat drops to 3V, what will the regulator output do? Is it a low drop-out type? The supply for the LM358 and the pot will, of course, drop too.
 

Arouse1973

Adam
Dec 18, 2013
5,178
Joined
Dec 18, 2013
Messages
5,178
If Vbat drops to 3V, what will the regulator output do? Is it a low drop-out type? The supply for the LM358 and the pot will, of course, drop too.
Its going to have to be really low drop to output 3.3 Volts at 3 Volts in :) My guess is it being a pass device, below regulation voltage vin is approx vout.
Adam
 

sweben

Jul 3, 2015
18
Joined
Jul 3, 2015
Messages
18
Right, I forgot to mention this. There are two batteries in serial, so VBAT never goes below 6V.
Concerning the consumption, the sensor (without WiFi) seems to be using ~40mA.
But here's something interesting, I replaced the 2n2222 with a larger BD135, and here it seems to work like a charm. Also with the 2n2222 it seems to light up for a couple of ms and then turn off.

So, what do you think? Could it be some kind of protection circuit in the transistor? too much current when the module starts the WiFi component?

Benoit
 

sweben

Jul 3, 2015
18
Joined
Jul 3, 2015
Messages
18
The gain is the ratio between the current flowing into the base and the current "allowed" to flow between the collector and emitter, correct?

So, if I take a much smaller resistor next to my transistor it should work? it seems like the transistor gain varies a lot between transistor, even of the same serie,
 

Arouse1973

Adam
Dec 18, 2013
5,178
Joined
Dec 18, 2013
Messages
5,178
The gain is the ratio between the current flowing into the base and the current "allowed" to flow between the collector and emitter, correct?

So, if I take a much smaller resistor next to my transistor it should work? it seems like the transistor gain varies a lot between transistor, even of the same serie,

Yes try reducing the value of the base resistor, put a 1K across the existing base resistor and try that. How much current does the circuit draw when fully on?
Adam
 

Alec_t

Jul 7, 2015
3,591
Joined
Jul 7, 2015
Messages
3,591
There are two batteries in serial, so VBAT never goes below 6V.
In that case you will need a separate voltage monitor for each battery (cell), because no two batteries are identical and the one that drops to 3V first will die if you only monitor the combined '6V' voltage.
 

sweben

Jul 3, 2015
18
Joined
Jul 3, 2015
Messages
18
In that case you will need a separate voltage monitor for each battery (cell), because no two batteries are identical and the one that drops to 3V first will die if you only monitor the combined '6V' voltage.
Right, unfortunately I would have to modify the battery holder quite a bit to measure each cell, do you think raising the shut down voltage to something like 7v (3.5V each) is safer?

Yes try reducing the value of the base resistor, put a 1K across the existing base resistor and try that. How much current does the circuit draw when fully on?
Adam
Alright, With a 1K resistor it seems to hold on a little bit more, but not enough, check out the output I get:
qLaKv5s.jpg


The drawn current is roughly 90mA.
 

Arouse1973

Adam
Dec 18, 2013
5,178
Joined
Dec 18, 2013
Messages
5,178
It looks like he has added a buffer 0.5 Volts per cell. To achieve the maximum capacity of the battery each cell can be discharged down to 2.5 volts.
Adam
 

sweben

Jul 3, 2015
18
Joined
Jul 3, 2015
Messages
18
It was around the charge - but I realized that it was just noise. The real output just goes to ~1.5v

taking an even smaller resistor for the base I manage to get a higher output voltage (1.7V for 220ohms) but then of course the transistor starts heating up quite a bit.

So, just using the bd139 sounds like a good solution, what do you think?
 

Arouse1973

Adam
Dec 18, 2013
5,178
Joined
Dec 18, 2013
Messages
5,178
What do you mean around the charge? -1.5 Volts? you don't have a negative supply, how are you getting this? Where are you measuring? With a suitable heat sink the current transistor can provide 800 mA with the correct base current.
Adam
 

sweben

Jul 3, 2015
18
Joined
Jul 3, 2015
Messages
18
What do you mean around the charge? -1.5 Volts? you don't have a negative supply, how are you getting this? Where are you measuring? With a suitable heat sink the current transistor can provide 800 mA with the correct base current.
Adam
- around the charge: on the connector from the schema above.
- it's not -1.5, it's approx. 1.5v

It's kinda strange that going as low as 220ohm is not sufficient, isn't it? could it be because the WiFi module (esp8266) have a very large current peak on startup?

seems like the WiFi module takes up to 170mA, and the DHT11 sensor up to 2.5mA
http://bbs.espressif.com/viewtopic.php?f=6&t=133
http://www.robotshop.com/media/files/pdf/dht11.pdf
 

Arouse1973

Adam
Dec 18, 2013
5,178
Joined
Dec 18, 2013
Messages
5,178
Exactly, I was just going so mention that. You may need to add some large capacitors to support the supply rail. Do you have any? Another option would be to use a logic level MOSFET.
Adam
 

sweben

Jul 3, 2015
18
Joined
Jul 3, 2015
Messages
18
hahaha! that worked :)

I get 3.17V on the connector now, I wonder if it's safe to run it at a lower voltage.

Edit: well, it's probably not enough, the module turns on but doesn't connect to the WiFi
 
Top