Help needed with transistor based, solar charge controller

Thread Starter

RIKRIK

Joined Oct 11, 2019
93
Okay so recently i have been wanting to make my own solar charge controller, I had a look round at circuits, They had regulators, dedicated IC's (555 timers, opamps ect) , More expensive p channel mosfets. Basically i wanted something cheap and easy to make, I like stuff that can be made by people on a real budget but that isnt complete garbage (Air quotations).

I decided on using the internals of the lm339 as its a pretty robust comparator.
A3.png

Basically , My know how is basic at best. So if anything in the circuit pops out to be wrong, please feel free to berate me on it or if you have any recommendations on changes i could try that would be appreciated :). Also it means i can make sure i am ordering everything i need.

My 2 main concerns.
1) im using a 12v relay, the panel voltage does go up to 17v, would i need additional protection , i did try simulating a totem driver as it half's the voltage and i thought having a 6v zener regulator with a 6v relay but i wasn't 100% happy with the circuit.
2) my resistor values, I did try and work some out but wasnt sure on some things, like swapping out the resistors for where the diodes should be.

The batteries on the left, solar cell on the right, LED is just there to show the state of the relay.

A1.png
A2.png
 

wayneh

Joined Sep 9, 2010
16,390
update , Opinions anyone :)View attachment 205798
Let’s start over with what you have (your panel specs) and what you want to accomplish. Don’t poison the well with a preconceived solution. Personally, the approach of reverse engineering a cheap IC makes no sense to me. But anyway let’s see the range of potential solutions before setting on one approach.
 

Thread Starter

RIKRIK

Joined Oct 11, 2019
93
Okay i have a 30W 12v panel, But if i could increase it to 50W 12V 4.1A, that would be great. as i got some 5A diodes lying around.. The two transistors im using

2N2907A
https://alltransistors.com/transistor.php?transistor=2579

2N2222A
https://alltransistors.com/transistor.php?transistor=1774

As i have them lying about, but i would use the non "A" versions so 2n2222a, if i wanted to make it easyier to source parts.

Im thinking more for poorer places around the world. Where they can easily fix broken electronics or even a cheap class project.

Any help would be amazing thanks :)
 

wayneh

Joined Sep 9, 2010
16,390
Okay i have a 30W 12v panel, But if i could increase it to 50W 12V 4.1A, that would be great. as i got some 5A diodes lying around.
I don't understand. You can't increase the output of a panel except to get a different panel. But anyway maybe you want to build a charge controller that can handle up to 5A.

Neither of those can handle even 1A of continuous current but I see the relay in your schematic. So your plan is a circuit to charge a battery at full-on or full-off. You'll need to have hysteresis to prevent the relay from chattering. The voltage on the battery will be immediately lower when charging is removed. You don't want it to kick back on immediately. And, you'll need to carefully match your battery capacity to the full current of the panel, to prevent over-current charging of the battery.

The circuit shown has some problems. The first is that the comparator datasheet shows an "equivalent" circuit to explain the block functioning of the IC. That diagram is not necessarily a real schematic. If you want to build a comparator from discrete components, I suggest finding a known-good working circuit schematic. I'm sure someone around here can guide you to one if you post specifically on that topic.

A second problem is that your circuit doesn't see the battery voltage once the relay breaks that connection. Your comparator needs to monitor the battery voltage at all times, I think. I think I'd want to also monitor the panel voltage and only trip the relay when two conditions are met: 1) The battery voltage is low and calling for a charge, and 2) The panel voltage exceeds the battery voltage. You could eliminate the blocking diode (which is placed backwards) if your relay control is working well.

A third problem is the relay. Those tend to be more expensive and less reliable than solid-state components.

A 5A charge controller is a cheap and abundant device these days and solves all these problems with an elegant design. It will ensure proper charging of your battery (very important to the lifetime and safe use of your battery), regulate the voltage to the load, and basically do everything right and efficiently. Building your own design from discrete components will be difficult, costly and will never achieve the performance of the commercial device. Of course you're free to continue, but I felt compelled to "warn" you. Maybe the challenge you could tackle is opening a supply chain of charge controllers to the people that need them.
 

Thread Starter

RIKRIK

Joined Oct 11, 2019
93
Okay thankyou , the battery is on the left , the panel is on the right, the panel is conected to the NC part of the relay. When the battery votage gets to a certain level. The relay will activate too the normally open position.

I knoticed some chatter on the relay in the simulator But when i biased the comparitor with a 2k resistor it slams the relay on and off.

And yep i was on about a bigger panel.

I will have a look into making the circuit only charge when the battery voltage is low.
 
Last edited:

wayneh

Joined Sep 9, 2010
16,390
Top