Smart charger for deep cycle battery (12V 145AH) and microcontroller

[djdaga]

I have a transformer (40A in the secondary) and a rectification stage (18V DC). Everything's fine. I have a circuit with an operational and a mosfet, as shown in the photo. The operational with 5V input, and a 14.6V-14.9V output, to control the mosfet. The image is for reference only, the values in the resistors and voltage have been changed, for work at 14.6V-14.9V

Now I want to control everything with a microcontroller. I have the voltage reading on the battery. So far so good.

I understand that some chargers work by limiting the current, and others work based on the voltage in the battery, and that the battery determines what current is flowing.

I'm putting a resistor in the gate to limit a certain current.

My question is, what voltage should be applied when connecting a battery, or else, which I should consider in my algorithm, for time, for the initial voltage... For example, if the battery has 10.8V, if it has 11.2V, if it has 11.5V, or special cases when it has 12.5V, but no charge, the latter happens to me often, when I measure a battery and I think it is OK, but when I connect it, the inverter shuts down immediately.

Can I improve this circuit somehow?

In the microcontroller, I am censusing 3 voltages, battery voltage, voltage at the power supply, and voltage at the operational output. I am applying PWM to 20khz, and the duty cycle, as you will tell me how and in what percentage.

**broken link removed**

I appreciate yours comments.

 

[KlausST]

Hi,

To be honest: I don't call this circuit a "battery charger". I rather call this a "trial and error" circuit.
It doesn't include a true voltage regulation, and it doesn't include a true current regulation.
Maybe you are satisfied with the function, but indeed the function is not reliable, because it depends on some parameters:
* Mosfet temperature
* battery temperature
* Mosfet type, individually (even if you use only IRFZ44N)
* input voltage
* battery voltage (with a low battery voltage your Mosfet will burn due to overheat)
* battery health status

My recommendations:
* read about battery charging, maybe at https://batteryuniversity.com
* use a true battery charging curcuit. Or use a dedicated battery charger IC. Consider to use a "switching" type rather than a "linear" type. A linear type combined with your big battery will cause a lot of power to be wasted as heat.

Klaus

 

[betwixt]

I have to agree with Klaus. It has no feedback mechanism to regulate anything and a few other potential problems too.

I'm not happy with "I'm putting a resistor in the gate to limit a certain current. " because the static gate current will be almost zero and it isn't reliable to slew the gate signal rise and fall using a resistor and the gate capacitance to make a PWM filter.

Adding capacitors across the LM358 output to ground may cause instability problems.

If you are simulating it's performance, I wouldn't use an AC voltmeter across the battery to measure it's voltage.

If you want to control the charging with an analog circuit, you need to monitor the voltage using a comparator, comparing against the 'fully charged' terminal voltage to cut the MOSFET off and if necessary, monitor the current to ensure it says within safe limits. I would advise a series resistor is added in series with the charge current and you do not rely only upon controlling the MOSFET current as it will dissipate lots of heat.

If you want to use PWM you still need to monitor both voltage and current. I would suggest adding a resistor between the negative side of the battery and ground to monitor the current so it is ground referenced, that way the current is converted directly to a measurable voltage above ground and you can use an ADC to determine it. Similarly, a potential divider across the battery can be used to allow an ADC to measure the terminal voltage. You can do it the way you show in the schematic but more parts are needed, I would go for the easier option! Also, if you are using PWM you want the MOSFET to be either fully off or fully on so there isn't much point in using an analog IC to drive it, a logic level shifter from the microcontroller to the MOSFET would be simpler and more efficient.

Brian.

 

[djdaga]

Greetings. Well I agree with you and KlausSt, I've read the matter a little bit on https://batteryuniversity.com. I'm clearer, though there are some doubts. The truth is that now, I am working on two fronts. I'm making the power regulator for operationals, as well as two circuits for the control of current and voltage then, I show them what they think, and then we'll talk about the implementation challenges and questions.