Cut off voltage lithium battery?

[Rorut]

Hi,
Im working on a portable boombox powered by four panasonic NCR 18650B, 3400mah cells in series. I bought a PCB and I have a small 4s balance charger. The PCB will be secondary if anything fails and cuts voltage at 3v.

Thats to low for me during daily use. I want a primary cut off at 14v (3.5v *4). I was thinking maybe this can be done with a zener diode and a relay? Something like my image illustrates it. (Sorry for the swedish in the image)

Translation in image:
Öppen=open
Last=load
Batt=battery
Vipp=on/off

Dont know if its a good idea? Or is there any other simple way of doing this? I'm not very good at electronics but understand some basics and can solder. Appreciate any answer.

Thank you!

 

[MikeMl]

Read Post #5 in this old forum thread

That circuit would work for you if you change R8 to match your required cutoff voltage.

**broken link removed** that would work in that circuit.

 

[JimB]

Dont know if its a good idea?

Basically it is a bad idea.
The voltages at which a relay will open and close are not well defined, and there is a big difference between the opening and closing voltages.

Or is there any other simple way of doing this.

A good way will always be more complex than a relay and a zener diode.
However, the more complex way will work reliably.
My initial thoughts were to suggest a solution using a comparator to drive the relay, but having seen the suggestion by MikeML I recommend that you try that.

JimB

 

[Rorut]

Read Post #5 in this old forum thread

That circuit would work for you if you change R8 to match your required cutoff voltage.

**broken link removed** that would work in that circuit.

Thank you very much for your answer!
I dont know how to calculate for the R8?
Just want it to cut at 14v. Can you help me?
Load will be 2-3A ,max 5A

Can I choose any PFet from the list?

Appreciate your help a lot!

 

[MikeMl]

Here is a modified circuit. The green trace is with the pot at 90%; blue=centered, red=10%. I show the CutOut and CutIn voltage hysteresis (with the pot centered) at the little steps on the V(cc) vs time trace. Note that the steps are due to the modeled internal resistance of the voltage source V1. If you need more hysteresis, make R9 smaller.

I would choose an M1 that has a low Ron so that you do not have to mount it on a heat sink. Note that M1's Source pin is tied to V(cc)

 

[spec]

Hi,
Im working on a portable boombox powered by four 18650 panasonic NCR 3400 mah cells in series. I bought a PCB and I have a small 4s balance charger. The PCB will be secondary if anything fails and cuts voltage at 3v.

Thats to low for me during daily use. I want a primary cut off at 14v (3.5v *4). I was thinking maybe this can be done with a zener diode and a relay? Something like my image illustrates it. (Sorry for the swedish in the image)

View attachment 99245
Translation in image:
Öppen=open
Last=load
Batt=battery
Vipp=on/off

Dont know if its a good idea? Or is there any other simple way of doing this? I'm not very good at electronics but understand some basics and can solder. Appreciate any answer.

Thank you!

Hy Rorut,

Nice informative post.You can see the echos of English is your translation of Swedish.

You say that you have four Panasonic NCR3400 batteries, but I cant find that part number. Did you mean NCR18650B?

I ask, because one of the critical areas to get right in a cut off circuit (COC) is battery rebound, and this depends on the load (last) and the battery characteristics.

Basically, without the correct window (hysteresis), the COC will oscillate: on load the battery voltage drops to 3.5V and the COC activates and removes the load. This causes the battery voltage to rise (rebound) and the COC reconnects the load and so the cycle continues.

Rebound is not significant for very light loads, as in the real time clock back up battery COC I did recently on ETO, ( https://www.electro-tech-online.com/threads/help-with-backup-circuit-for-digital-clock.147727/ ) but in your case, rebound will probably be significant.

Is there any reason why you want a cut off of 3.5V per battery, rather than the already conservative cut off of 3V per battery that you already have?

spec

 

[Rorut]

Here is a modified circuit. The green trace is with the pot at 90%; blue=centered, red=10%. I show the CutOut and CutIn voltage hysteresis (with the pot centered) at the little steps on the V(cc) vs time trace. Note that the steps are due to the modeled internal resistance of the voltage source V1. If you need more hysteresis, make R9 smaller.

I would choose an M1 that has a low Ron so that you do not have to mount it on a heat sink. Note that M1's Source pin is tied to V(cc)

View attachment 99255

Thank you very much for the updated circuit, much appreciated!
What is R5 5 i the load "box"?

 

[Rorut]

Hy Rorut,

Nice informative post.You can see the echos of English is your translation of Swedish.

You say that you have four Panasonic NCR3400 batteries, but I cant find that part number. Did you mean NCR18650B?

I ask, because one of the critical areas to get right in a cut off circuit is battery rebound, and this depends on the load (last) and the battery characteristics.

Basically, without the correct window (hysteresis), the cutoff circuit will oscillate: on load the battery voltage drops to 3.5V and the cut off circuit activates and removes the load. This causes the battery voltage to rise and the cut off circuit reconnects the load and so the cycle continues.

Rebound is not significant for very light loads, as in the real time clock back up battery cut off circuit I did recently on ETO, ( https://www.electro-tech-online.com/threads/help-with-backup-circuit-for-digital-clock.147727/ ) but in your case, rebound will be significant.

spec

Thank you, yes you are correct it is NCR18650B. I changed my post. I understand what you mean but don't know how to investigate it? Load will be between 3-5amps.
Maybe 3.5v is to high?
(Maybe less, hard to say exactly. Think it depends on volume,bass but it is a 40w class d amp and a bluetooth module)
Thank you all for the quick response in this great forum!

 

[spec]

Thank you, yes you are correct it is NCR18650B. I changed my post. I understand what you mean but don't know how to investigate it. Load will be between 3-5amps. Thank you all for the quick response in this great forum!

And thank you for your concise and quick response; that helps us a lot. You don't need to worry about the design. So far, Mike has suggested a good solution and I have been working on a cut off circuit since your post on the old thread.

I recently added another question to my #6 post above about the 3.5V cut off. I would be interested to know your reasoning.

While I'm in a question mode: is the cut off circuit (COC) current consumption important? When the battery terminal voltage drops to 3.5V do you need the battery drain to be minimum. Normally COCs are designed to have a low current drain, at the most equal to the steady state battery self discharge.

spec

 

[spec]

I have just been having a look at the NCR18650B datasheet. I don't like to be a kill-joy, but it appears that you would be lucky to get 500 mA/H from your batteries with a cutoff of 3.5V and that would be at a temperature of 25 deg C.

Also, I suspect that having a cut off voltage just 100mV below the battery nominal of 3.6V is going to cause all sorts of problems. For example when you consider the drop in terminal voltage due to the load current and the battery internal resistance, especially when switching on in a cold environment.

Not only that, but the terminal voltage drops with a decrease in temperature, so on a cold day the circuit may never turn on. I believe it gets a bit cold in Sweden from time to time. Battery rebound also makes the situation more critical. To further compound the problems, the datasheet applies to brand new batteries: with use, battery characteristics deteriorate.

spec

NCR1865B datasheet
**broken link removed**

 

[spec]

Another observation is that it may be possible to alter the cut off voltage of you present bought cut off circuit.

spec

 

[Rorut]

Thank you very very much for all the research!
The reason to cut off at 3.5v is because I have been told that it is not good for the batteries going lower than that. But if it is ok to go lower like 3.1-3.3v or something Im more than happy. A second protection if the PCB fails seems to be good as well.

Temperature is not a problem. (Well it depends how you see it only having summer a day or two every year ). Im only going to use it during summer days or inside.

I dont know if cut off circuit current consumption is imortant? Just thought it would be a good idea to cut everthing at a certain voltage.

Approximately At the same time I joined here I found this circuit that should be doing the same thing like mikes. I dont know if they are equal or wich one is better?
https://3.bp.blogspot.com/-Y1MicUfY...PSF4z1Mg/s1600/low+voltage+cutoff+circuit.png

Did not expect this project to grow so much but it is very fun and Im learning a lot from it

 

[MikeMl]

Thank you very much for the updated circuit, much appreciated!
What is R5 5 i the load "box"?

Your boombox, etc.... I=E/R = 14V/5Ω = 2.8A is the load for the test...

 

[MikeMl]

Do not some Lithium batteries have a cut-off built-in to the battery pack? If so, wouldn't the first (of four) to reach its cutoff point trigger the secondary cutoff circuit I posted...

 

[Rorut]

Do not some Lithium batteries have a cut-off built-in to the battery pack? If so, wouldn't the first (of four) to reach its cutoff point trigger the secondary cutoff circuit I posted...

Some cells have a built in protection but it often limits how much you can discharge the battery. I think this batteries can give between 4-5amps and often those PCBs limits to about 2A. But I'm not sure only what I read in forums. My cells are unprotected

Thank you!

 

[spec]

The maximum discharge current for the NCR1865B is not stated as such on the datasheet but the graphs show a maximum discharge rate of 2C, so I would imagine that is a safe limit.

spec

 

[MikeMl]

...My cells are unprotected...

Then, do you not have to have four cutoff detectors, one for each battery in the string rather than a single 14V cutoff?

 

[Rorut]

Then, do you not have to have four cutoff detectors, one for each battery in the string rather than a single 14V cutoff?

I think my PCB designed for four cells is checking each cell. And Im using a balance charger. Thought that would be enough but like you say, check each battery would probably be more correct.
Like thoose RC buzzers connected to the balance cable alarming at a specific cell voltage. To bad they dont cut off the power. Probably not very good for flying

 

[Rorut]

The maximum discharge current for the NCR1865B is not stated as such on the datasheet but the graphs show a maximum discharge rate of 2C, so I would imagine that is a safe limit.

spec

Is that the same as 6.8A?
2*3.4

 

[audioguru]

I made a boom-box for the beach about 40 years ago and it sounded much better than any other sound system that was there. It used Ni-Cad battery cells. When they went dead (the voltage dropped suddenly) then it was time to go home.