Charging 24v lead-acid with 36v PV cells

[basement_dweller]

Hello there,

First of all, I would like to apologize for my sloppy English.
I hope this is the correct place to ask for an advice from experienced folks.
I'm trying to design a PWM solar charge controller. It's supposed to charge an 24v lead-acid deep cycle battery from an 36v photovoltaic panels.

As far as I know, advanced controllers implement buck/boost converters right after PV panels to maximally utilize panel power. Some cheap PWM chargers just use buck converter to regulate voltage enough for battery charge.

What I have in plans:
1. Charge a 24v lead-acid directly from 36v panel.
2. I can use PWM drive to limit the current into the battery/load (read cell voltage and slowly increment duty cycle till desired current limit is reached, same method to prevent PV overload).
3. To control battery overcharge, i will stop PWM drive once a few seconds and read battery voltage (subtract Battery_U from Cell_U).
4. I will use isolated DC/DC converter to power MCU from battery. DC/DC secondary side will be ground referenced to PV panel. (I think that way it will be easier to drive PWM and read voltage/current on PV side).
5. I will definitely add some current sense on load side to make sure charger supplies enough juice for both battery and load.


My main question to you: Is this desing worth a try and is it safe to rely on PWM for safe charge of that battery? (I'm afraid that battery may start gassing, regardless of PWM limiting)

And what is the easiest method of sensing actual current? I can limit max current based on PV voltage and duty cycle, but how to measure it?
One of the ideas was to sample voltage across current shunt resistor when MOSFET switches on and calculate using that voltage and known duty cycle, but there will be voltage spikes. For now i'm planning to put large LC filter to smooth things out.

Attaching JPEG image of charger's power stage.
Best regards

[Seekonk]

Do you want to go cheap? Do you have a lot of electronic junk?

I have a 36V array system I run my cabin with. It has run from a UNO many years.  Simple bulk charge like a car.  If you have an active system most controllers never get out of bulk charge mode. I just control PWM by max battery voltage, simple buck converter.  I also keep panels from going below power point voltage.  It is a simple up down counter on the PWM.  You can manually set power point seasonally or track panel temperature.  High side FET driver is just an opto isolator.  36V array is about 50VDC.  Many small wall warts run at 50V at reduced power.  That drives the opto.  An opto is slow, but it just doesn't matter at 490Hz, FETs won't heat up.  I use a power transformer from a UPS as the inductor.  Main cap is every capacitor I can find in parallel. You don't really need current monitoring. Less than $10 outlay and it supplies all the power needs of my camp, even hot water.

[uncle_bob]

Hi

Using pretty much the same parts you have shown (mosfet and inductor) you can make a buck converter. Yes you will need a diode.

With the circuit as you show it, the Mosfet PWM pumping the battery ground up and down. If you want to get power off of it while charging , this will be a bit of a mess (think of the load voltage).

Far easier to use a readily available (cheap) buck converter.

Bob

[mtdoc]

You need to specify what you mean by "36V PV cells"

Is that the open circuit voltage (Voc) or the maximum power point voltage (Vmp)?

In order to reliably charge a 24V nominal battery bank the Vmp will need to be at least 34V.

The vast majority of solar panels built these days are for grid tied batteryless systems. They typically have a Voc of 36-38V and a Vmp of of 29-32 V. One alone will not reliably charge a 24 V nominal lead acid battery bank.  Any buck conversion will only make matters worse. If you have more than one they can be placed in series. The voltage will then be high enough above the necessary charging voltage to justify buck conversion - most efficiently done with a MPPT controller in that case.  If charging a 24V battery bank with PV Vmp of 34-40V or so, a simple PWM charge controller will be fine.

[basement_dweller]

Thank you everyone for the input. It is indeed is going to stay in the bulk charging stage most of the time
Yes, I understand that load voltage will jump up and down, but it's not a problem, as "subscribers" will have it's own buck converters.
Panels have 36-40V maximum power point voltage, and about 48V open circuit (well, according to seller, haven't seen a datasheet yet).

Anyway, thanks for the suggestions, for now my design stays unchanged, will cook up a prototype tomorrow.
I guess I can always add buck/boost converter to existing board later.

[Seekonk]

I wouldn't go with that design, but that is how you get an education.