Good day! I am making a battery monitor device. The device has to monitor 2 batteries in series, monitor the charging of each battery separately and of the 2 batteries toghether.
Specification: charger on 1 end with 14.7V, 20A. When the charger is disconnected the batteries will be connected to a load. The idea is to have a battery pack of 2 batteries at all times, connected in series, which I will monitor and later send the information to an MCU. There should be a defence against opposite connection (+ and - switched).

I have made these circuits.
1. 12V monitor control - "Working.version.small.PCB.png".
What bothers me is: will the traces need to be too big in order for the fuse to burn first? Will 30A be enough for the diode "D1"?

2. 24V monitor control (2 batteries in series) - "Medium.battery.monitor.circuit.png".
What bothers me is the same as before?

3. Diode control circuit - "Diode.version.battery.monitor.png".
Why is it not good to use the diode version circuit?

 

Sorry for the double post. Between the two batteries there should be a switch, as they will be charged separately and used toghether. Recommendations for the switch? This is the block diagram:

 

What bothers me is: will the traces need to be too big in order for the fuse to burn first?

I wouldn't have any part of the PCB carrying 30A. The fuse can be an inline type and the diode connections should similarly not require any parts of the PCB traces.

Will 30A be enough for the diode "D1"?

I'd be reluctant to have the diode and fuse at the same amperage rating. I'm not sure of the details but in the case of a short - perhaps a 40A current - there is some finite amount of time it will take the fuse to blow. The diode might be destroyed during this instant. You might alleviate this fear if you carefully read the data sheets off both the fuse and the diode, but without reading those it makes me nervous.

Why is it not good to use the diode version circuit?

You tell me. Who says it's not good? I admit that zeners in series confuses me. Their combined voltage exceeds the supply, so I'm not sure current will ever flow?

 

First- *thank you* for posting schematics-- that is so appreciated.

However, I think you're solving a problem with components that is much more easily solved in software. You stated you intend to use an MCU. Why use it 'later'? Speaking for just the charge monitoring circuit, you don't need all the OpAmps, and LEDs, and diodes, and all that. You need 3 resistors and a 4N35 for each battery you want to monitor, and one LED. The 4N35 not only acts as an isolation between the battery and the circuit, but because it has virtually 100% transfer of energy it lets you draw just 1mA off the battery to read it's voltage via a ratiometric divider, and then input that into an ADC to evaluate. You can then use a single RGB LED, and color it any way you like. Add a temp sensor for each battery, and if a battery starts to go thermal, you could disconnect it, start blinking your LED RED, and turn on a piezzo buzzer alarm.

Single circuit board, small. Charging side can have another 4N35 to control transistor or solenoid or what have you, with a fuse for protection, to charge the battery. If you use a fuse, learn about fuses so you get the proper rating. Don't use a 30A fuse for a 20A circuit. You can build polarity protection into the charging side using FETs on the ground side, and an LED to indicate battery backwards.

One last note- if you really want to 'test' a battery voltage, you could modify the circuit to run the battery you're testing through a heavier load, and do a test, to see how it behaves.

 

Speaking for just the charge monitoring circuit, you don't need all the OpAmps, and LEDs, and diodes, and all that.

To be fair, it's a single comparator IC and is a pretty reasonable solution if you're not using a micro.

 

There should be a defence against opposite connection (+ and - switched).

I have made these circuits.
1. 12V monitor control - "Working.version.small.PCB.png".
What bothers me is: will the traces need to be too big in order for the fuse to burn first? Will 30A be enough for the diode "D1"?

You can use a P-MOSFET to protect against a reverse polarity connection without the voltage drop of a series diode, thus you wouldn't need the fuse.
You connect the drain to the battery connection, the source to your circuit, and the gate to ground.
A simulation of the circuit below shows the output voltage equals the input when the input is positive and is zero when the input goes negative.
It works because a MOSFET conducts equally well in both directions when on.
The forward voltage drop is just the circuit current draw times the MOSFET on-resistance (which is very low for typical MOSFETs).

3. Diode control circuit - "Diode.version.battery.monitor.png".
Why is it not good to use the diode version circuit?

It would be less accurate and the trip points are temperature sensitive.

Note you appear to have the PWR-FLAG connected to ground on your schematics.

 

To be fair, it's a single comparator IC and is a pretty reasonable solution if you're not using a micro.

I wasn't downgrading it; it just isn't necessary, given that he mentioned an MCU was going to be involved at some point. Why not take advantage of it now?

 

Traces at 30A:
I agree about the PCB not carrying 30A. I was thinking of only mounting the elements on the PCB and connecting them with wires (5mm should be enough) to a connector (a picture would be appreciated here, I learn visually better).
D1 can be 60A. Datasheet http://www.st.com/en/diodes-and-rectifiers/stps12045.html

Zeners:
The zener circuit is only for reference, I will give a better one. What bothers me is shouldnt there be some resistor for the 13V zener, how will that resistor increase/decrease the voltage as there will be a change in the battery voltage and in the current forwarded by the zener. If anyone can give a good zener circuit with measuring at least 4 voltages (10.5, 12, 13, 14) its good.
As for zeners in series, the idea is the upper zener to be 8V, the lower 3.3V, some resistor also that I dont know how much should be, a LED
connected in paralel to the 3.3.V zener to be light from 8V to 11.3V, when the 3.3V zener also breaks at 11.3V the LED to be off.

Hardware and software problems:
Which problems should be solved by hardware and which by software? I will have only 1 analog pin. I dont have the time to make it with an MCU, I will add an MCU with a wifi module (Amica node mcu rev. 1.0 not 0.9 a.k.a rev2) later.

How will the P MOSFET be cooled? Arent 20A - 30A a lot for it?

KiCad and a footprint for a pot:
I am new to KiCad so why shouldnt I have a power flag connected to ground (I saw it on some tutorial)? I also have a problelm with finding the right footprint for this pot:
https://www.tme.eu/en/details/1028f-5k/single-turn-tht-trimmers/sr-passives/

Latest circuit:
This is the latest circuit, what will happen if the switch is off but the load is connected (12V will flow through the load and the battery will get discharged twice as fast?). Please pay attention to the diodes D14 and D6 at the bottom right of both circuits for both batteries. Sorry if the picture quality is bad, I only have it in ".XPS" version and it can not be uploded.

Compponents specification:
The switch should be rated 30A for a 25A load?
The P MOS FET should be 30A, 50V for a 24V, 25A load? How should it be mounted to the board, again only screwed to the board and connected with wires?
Everything up to the load connector should be rated 30A (I am guessing there is no voltage rating) and connected through wires 5mm or 6.5mm for both the wires and connectors?

 

The P MOS FET should be 30A, 50V for a 24V, 25A load? How should it be mounted to the board, again only screwed to the board and connected with wires?

What P-MOSFET?

The MOSFET's current rating is not the most important characteristic. You want the MOSFET to have a low enough on-resistance so it doesn't require a heatsink.
For that you want to keep the power loss below a watt which means the on-resistance should be <1.6mΩ for 25A.

 

It also has to be some package that will only be mounted on the board, but connected with wires?

 

It also has to be some package that will only be mounted on the board, but connected with wires?

If it has 25A going through it, yes.

 

I will make some changes to the circuit and post it if I can.

Can anyone answer if a 5mm multi thread wire is enough for 30A, I think it should be?

Also what bothers me is that on the upper circuit, the comparator's and diodes are connected to ground, maybe they should be connected to the negative pole of the upper battery instead with no ground?

 

This MOS FET is good, but its an NPN one.

 

OK, the load will be an AGV, should I count that as resistive or inductive load? Because if I put a clamping diode and the battery is connected in reverse, the diode will conduct. If I put a current limiting resistor with the clamping diode, I don't know what the value should be, so the inductive load can be discharged, while the battery if connected in reverse, does not get discharged much.

 

I was thinking of doing something like this:

Also I belief you mixed the source and drain in post 6. My biggest problem is if 1 battery only is reversed, that's why the N MOSFET would be better? But I should drive it with a zener resistor voltage divider and a bipolar BC547 or 2n3054 if 10.5V (discharged battery) are not enough for it.

 

I guess I overstaied my welcome . Thanks to everyone for the help!

 

Also I belief you mixed the source and drain in post 6.

No.
I explained in that post why the source and drain are connected as they are.
The P-MOSFET conducts equally well in both directions when ON, but blocks when the drain goes negative (applied reverse polarity).
So the normal current flow through the P-MOSFET is from drain (+) to source, the same direction as forward-bias for the source-drain body.

 

No.
I explained in that post why the source and drain are connected as they are.
The P-MOSFET conducts equally well in both directions when ON, but blocks when the drain goes negative (applied reverse polarity).
So the normal current flow through the P-MOSFET is from drain (+) to source, the same direction as forward-bias for the source-drain body.

Ok. Thanks for the explanation.

What will happen if both batteries are connected plus to plus and towards the load is minus and minus? Meaning only 1 battery is reversed from the 2?

Edit: I think nothing will happen as current will not flow.