Battery protection circuit - i had to try it

[angelioto007]

Hello guys, i´m new on eevblog, in fact, this is my first post and first circuit designed completely by me  . I have tried it, but it didn´t work, but i´ve made some changes, but i don´t have time to test the circuit. So i was thinking, if some of you guys can help me out.
here´s the schematics: http://www.mediafire.com/download/exgvrabwfoxfe1m/battery+protection+circuit.png i have forgetten on R4 = 150R
i have tried to do the circuit as best as i could, but i need someone to judge it. keep in mind it´s my first circuit designed entirely by me and to be honest, i expect catastrophe, but i just have to try it.
p.s. please excuse that the circuit is in png file, its still better than a raw eagle file...
thx guys 4 support 

[FireBird]

You don’t have to change the „functionality is NOT guaranteed“ with „Guaranteed to blow up your house“  but there are some issues. In case I didn’t miss something, here are my findings:

*The maximum charge voltage, the battery sees, is about V+ Charger minus 0,7V (D1). How should IC1C switch, when V Battery will never get higher than V Charger ... except if the charger is disconnected?

* I’m not completely sure about your intentions but I guess that the Zener diodes won’t do what you want them to do. If you want to compare the battery voltage to a fixed 6,3V and 5,6V, you need to put D2 in the place of R2 and D3 in place of R5. As you’ve connected them, they would subtract the Zener voltage but only if current is flowing through them. Because the comparator’s inputs have practically infinite resistance, no current will flow and the non-inverting input will see the voltage of the voltage divider R1/R2 and R6/R5.

* The LM339 has open collector outputs. This means that it can sink current (pull the output low) while –In > +In, but whenever +In > -In, the output is floating and you need additional pull-up resistors to drive Q2 and T1 and to make sure Q1 shuts down.

* VOUT2 is connected to a FET and a transistor at the same time. The base-emitter-voltage of T1 is about 0,7V, the gate-source-voltage of Q2 should be several volts and therefore Q2 would never turn on in this setup. Either find a good combination of pull-up and T1 base resistor, or to make it simpler for you, use the fourth LM339 with inverting and non-inverting inputs of IC1B and IC1D in parallel and one output driving the transistor, the other the FET. Still don’t forget the appropriate pull-ups.

* The dead battery indicator seems to be an logic AND of VOUT1  and VOUT2 (LED is lit when both VOUT1 and VOUT2 are high). Why? If the battery has dropped below 5,6V, it has definitely already dropped below 6,3V. Connect R4 to VBattery and forget VOUT1.

* Maybe there is some use for VOut1. Whenever Q2 disconnects the load due to a dead battery, the battery may recover a bit and the voltage may rise above 5,6V. Q2 will reconnect again, the voltage drops and you have a nice oscillator. Just not sure if the device will love it.  Add a Flip Flop, which will be set if the battery voltage is above 6,3V and reset if it is below 5,6V. This way it won’t oscillate.

There might be one or another point but don't stop to experiment. The start is always hard.

[angelioto007]

thx 4 your help, i´m still not an electrical engineer  , i´m gonna continue experimenting. As the title says, i´ve had to try it  ...
this schematics was a second version, the first version used one resistor to drop the voltage for the zener diodes, but when i´ve tested on breadbord, it kinda worked, so i thought that it might work, but i have thought it can´t be so easy to construct a battery protection circuit. 
I should say, that i´m making the circuit for my NiMH battery pack, 10 AA batteries in series, 800mAh each and i want it for my small pringels boombox i made, it uses TDA1517 and it has built in vu meter  , so in total it draws around 300mA.
Anyway thanks again  for your help and i think that if the batteries are charged, the chrger might raise the voltage due the disconnected load.
i´ll continue making it better don´t worry, but i had in plan to buy an Arduino over xmas and learn programing.

[angelioto007]

maybe i shouldn´t start with something hard like this 

[poorchava]

Most battery protection circuits use back-to-back N-mosfets on one of the rails. This ensures that current doesn't flow in any direction, and the circuit becomes permanently latched until appropriate voltage is applied between contacts. You need to remember that every mosfet has a parasitic diode, which is not clearly depicted on your schematic (IMO drawing MOSFETs without parasitic diode shown is just wrong)

[angelioto007]

You don’t have to change the „functionality is NOT guaranteed“ with „Guaranteed to blow up your house“  but there are some issues. In case I didn’t miss something, here are my findings:

*The maximum charge voltage, the battery sees, is about V+ Charger minus 0,7V (D1). How should IC1C switch, when V Battery will never get higher than V Charger ... except if the charger is disconnected?

* I’m not completely sure about your intentions but I guess that the Zener diodes won’t do what you want them to do. If you want to compare the battery voltage to a fixed 6,3V and 5,6V, you need to put D2 in the place of R2 and D3 in place of R5. As you’ve connected them, they would subtract the Zener voltage but only if current is flowing through them. Because the comparator’s inputs have practically infinite resistance, no current will flow and the non-inverting input will see the voltage of the voltage divider R1/R2 and R6/R5.

* The LM339 has open collector outputs. This means that it can sink current (pull the output low) while –In > +In, but whenever +In > -In, the output is floating and you need additional pull-up resistors to drive Q2 and T1 and to make sure Q1 shuts down.

* VOUT2 is connected to a FET and a transistor at the same time. The base-emitter-voltage of T1 is about 0,7V, the gate-source-voltage of Q2 should be several volts and therefore Q2 would never turn on in this setup. Either find a good combination of pull-up and T1 base resistor, or to make it simpler for you, use the fourth LM339 with inverting and non-inverting inputs of IC1B and IC1D in parallel and one output driving the transistor, the other the FET. Still don’t forget the appropriate pull-ups.

* The dead battery indicator seems to be an logic AND of VOUT1  and VOUT2 (LED is lit when both VOUT1 and VOUT2 are high). Why? If the battery has dropped below 5,6V, it has definitely already dropped below 6,3V. Connect R4 to VBattery and forget VOUT1.

* Maybe there is some use for VOut1. Whenever Q2 disconnects the load due to a dead battery, the battery may recover a bit and the voltage may rise above 5,6V. Q2 will reconnect again, the voltage drops and you have a nice oscillator. Just not sure if the device will love it.  Add a Flip Flop, which will be set if the battery voltage is above 6,3V and reset if it is below 5,6V. This way it won’t oscillate.

There might be one or another point but don't stop to experiment. The start is always hard.

I have made the adjustments to the circuit , i can´t wait till i´ll test it on th breadboard . continue judging guys, you´re making so far good support and i think i´m doing good so far . Here´s the schematics (in png file again, dunno how to export it in pdf): http://www.mediafire.com/view/bhdmcmhb5vh2jy0/battery_protection_circuit_revC.png
I hope it won´t be as that big catastrophe as it was last time.
Yet again, thx guys 4 your support...

[angelioto007]

You don’t have to change the „functionality is NOT guaranteed“ with „Guaranteed to blow up your house“  but there are some issues. In case I didn’t miss something, here are my findings:

*The maximum charge voltage, the battery sees, is about V+ Charger minus 0,7V (D1). How should IC1C switch, when V Battery will never get higher than V Charger ... except if the charger is disconnected?

* I’m not completely sure about your intentions but I guess that the Zener diodes won’t do what you want them to do. If you want to compare the battery voltage to a fixed 6,3V and 5,6V, you need to put D2 in the place of R2 and D3 in place of R5. As you’ve connected them, they would subtract the Zener voltage but only if current is flowing through them. Because the comparator’s inputs have practically infinite resistance, no current will flow and the non-inverting input will see the voltage of the voltage divider R1/R2 and R6/R5.

* The LM339 has open collector outputs. This means that it can sink current (pull the output low) while –In > +In, but whenever +In > -In, the output is floating and you need additional pull-up resistors to drive Q2 and T1 and to make sure Q1 shuts down.

* VOUT2 is connected to a FET and a transistor at the same time. The base-emitter-voltage of T1 is about 0,7V, the gate-source-voltage of Q2 should be several volts and therefore Q2 would never turn on in this setup. Either find a good combination of pull-up and T1 base resistor, or to make it simpler for you, use the fourth LM339 with inverting and non-inverting inputs of IC1B and IC1D in parallel and one output driving the transistor, the other the FET. Still don’t forget the appropriate pull-ups.

* The dead battery indicator seems to be an logic AND of VOUT1  and VOUT2 (LED is lit when both VOUT1 and VOUT2 are high). Why? If the battery has dropped below 5,6V, it has definitely already dropped below 6,3V. Connect R4 to VBattery and forget VOUT1.

* Maybe there is some use for VOut1. Whenever Q2 disconnects the load due to a dead battery, the battery may recover a bit and the voltage may rise above 5,6V. Q2 will reconnect again, the voltage drops and you have a nice oscillator. Just not sure if the device will love it.  Add a Flip Flop, which will be set if the battery voltage is above 6,3V and reset if it is below 5,6V. This way it won’t oscillate.

There might be one or another point but don't stop to experiment. The start is always hard.

I have made the adjustments to the circuit , i can´t wait till i´ll test it on th breadboard . continue judging guys, you´re making so far good support and i think i´m doing good so far . Here´s the schematics (in png file again, dunno how to export it in pdf): http://www.mediafire.com/view/bhdmcmhb5vh2jy0/battery_protection_circuit_revC.png
I hope it won´t be as that big catastrophe as it was last time.
Yet again, thx guys 4 your support...

and forgot on R4 = 150R