PCB's not working PLEASE HELP!!

[LukeB]

https://easyeda.com/editor#id=|8d81327d6dc844d18b79c904581fdee8|5d46fcd8dacf4a44b908ff023d83213a|d86dbc9f9e114e7c8ef41c07b81c5d83

This PCB is not working can anyone help. It is a battery charge, protect, and boost circuit. The charge and boost part work ok but the prtect part doesnt stop voltage getting to the output. Dont know how it can do this but if someone knows better I would appreciate it.

PS If you need more details or tests done I have a multimeter and oscilloscope handy and ready to do what you say.

[JS]

In what conditions does that? How do you know it's not working or how did you measured this behaviour?

JS

[LukeB]

When I turn the voltage on my power supply down below 3.00 Volts (the threshold mentioned in the datasheet) I would expect the voltage on the output to go to 0.00v however it still gets voltage. In this case it is around 2.5 volts. I have measured this using an oscilloscope and multimeter and have even changed the power supply for a real battery and nothing works

[JS]

There might be some residual voltage on the output, what are you loading the circuit with?

JS

[.rpv]

Hi, you didn't mention if you're getting the voltage expected on the boost output, so on your circuit the low voltage protect IC will cut the battery from it's B- side, so if your measurement it's from B- to vin (on the boost IC)  your always will get your B+ B- (battery) voltage, your measurement should be from B+ to GND, another way to do the measurement it's from the boost output, of course.

If you're still getting this behavior I'll check for a possible bridge between B- to GND, and then maybe a failed mosfet or the protect IC.

[LukeB]

I have used a 2ohm resistor as a load and a few other higher ohm loads too.

In regards to .rpv, i am measuring the voltage across the 5v out pin and ground and I am in fact getting 5v out all the time. This is even when it should be off due to the protect feature.

[LukeB]

Just a quick update, I have also measured the continutity between gnd and the battery negative and found that it has continuity when at the correct voltage and no continuity when below the threshold voltage. How can there be voltage at the output if the battery negative has no continuity??

[chrisl]

1. Your U2 symbol has two OD pins; 1&3. I believe the pin 3 of U2 is OC.
2. The MOSFET (Q1) gates connections seem swapped.
3. Don't forget to add a 0.1uF between the pin 5 and 6.

[LukeB]

Yeah I noticed that. That is a mistake on the part of easyeda. For my use the name doesnt matter as they do the same thing.

[LukeB]

What difference does it make swapping the mosfet pins around??

[LukeB]

Ok so I have done a very very brief test of switching those two wires around like you said. (the ones on the mosfets) It seems to have worked. I will do more tests soon but looks promosing.

What I really want to know now is WHY? Why did that make the difference. Isnt the mosfet symetrical? Is it something to do with the internals of the BMS circuit?

thanks for all the help so far

[chrisl]

Hello Luke,

OC and OD are not the same. OC turns off when over charged and OD turns off when over dis-charged.  This is important since you want to turn off the correct section of the MOSTFETs.
Since the MOSFET is mirrored, not symetrical,   the protection diode between DRAIN and SOURCE can provide the leakage path if the wrong section is turned off by the wrong pin.
Also adding a 0.1uf CAP between the pin 5 and pin 6 of U2 is important to stabilize U2  since the part is connected to the battery via the 200 ohm resistor. U2 may go unstable at some battery voltage level and temperature.

Cheers,

Chris

[LukeB]

Oh ok, yeah I never would have figured that out by myself. I thought that since a mosfet was basically a switch it wouldn't matter but as you have pointed out i was very wrong.

Thanks for the help, I really appreciate it. This issue had me really stumped.

I will do some further tests to be certain nothing else is wrong too and then I will reorder the PCBs.

Thanks heaps
Luke

[LukeB]

One more question, the gate to source leakage is only 0.1 micro amps. How can the incorrectly connected mosfets be the problem if they only let less than 0.1 micro amps pass?

[chrisl]

It is not the leakage through the gates.  The FET on the left is turned on since it is connected to the pin 'OC' in your original design.
The FET on the right is turned off ('OD' in your original design) but the diode provides the current path.

The FET on the left should be turned off and the FET on the right should be on in order to work correctly.....

[kjr18]

I see that your intention is to charge battery with around 1A current. It would be good to connect U1's thermal pad to ground, this way you will have better heat dissipation. You could also use some vias to connect this thermal pad to ground plane on other side of pcb.

[LukeB]

Oh ok I will keep that in mind for future versions etc