I've completely rebuilt the MPPT controller test rig so that the Arduino has control over the PWM mark/space ratio. This opens the way for complete closed loop feedback control and automatic tracking of the maximum power point. Then the sun came out!!
By Julian
Youtuber, shed dweller, solar charge controller aficionado
that's a nice 'board' you have there ^^
Did you make a computer simulation with the circuits?
The Buck Converter should have a Low ESR cap directly at the MOSFET Source to Ground, then also at the output side of the Inductor… not after the current sensor. High-side MOSFET drivers… remember, you need to turn the Gate OFF fast, as well as ON fast. Look at the Fairchild FAN7371MX or Micrel MIC5014/5015. Also consider shielding the Hall current sensor in some ferrous metal. And… of course, there are other ways of measuring DC current… differential voltage drop across a shunt… but I'm not sure if the differential ADC is cheaper than the hall effect sensor. Great project… still watching the entire playlist…
I've seen a few videos of your MPPT Solar charge controller but never really understood how it works. But with this video I finally understand I think.
Changing this PWM cycle would increase and decrease the voltage in the buck converter. I've found this out myself playing with a 555 timer DC to DC converter. The same is true for an SMPS as well I think. Is the idea here to keep the output voltage the same no matter how little and how much light the solar panel receiving? So the battery is always charging? and when the battery voltage gets to around 13.5v the charge controller just trickle charges the battery to always keep it topped up.
Julian is it possible to send me a copy of the part list for the Aarduino you are using I would like to try something like it here in the states (ohio).
It's a simple potential divider designed to scale 25v down to 5v. It has a 303 (30k) resistor on top and a 752 (7k5) resistor below.
I modified the sensor on the solar panel side to scale 47v down to 5v. I added a 33k to the 30k making 63k on top.
63+7.5=70.5
70.5/7.5=9.4 (47/5 is also 9.4)
What are the details on the voltage sensor board?
Hey thanks.
It's certainly true that using a linear regulator to take 12v down to 5v isn't very efficient. But at these current levels, the Watt loss is pretty small. Accuracy isn't affected much at all. Looking at the data sheet for the 1117 regulator, load regulation and output stability only change by a fraction of 1% as the temperature increases. But the 1117 has an absolute maximum input voltage of 20v, so if I want to make this work on a 24v battery, I may have to use a switch mode regulator.
1. not yet 2. limited only by buck converter components 3. yes, with suitable profiles.
Next steps are: calculate efficiency and display, build simple MPPT control loop, build battery voltage regulation control loop (your point #1).
Julian, some questions: 1) at this point are you monitoring the charge state of the battery (in order to avoid overcharging)? 2) what would you estimate is the largest pv panel the system could handle (watts)? 3) would the system be able to work with various lead-acid battery architectures, e.g., deep-cycle and standard auto batteries (not at the same time), etc?
PS great series of vids. Thanks.
I've not really thought about that yet. But I'm not planning to turn this into a product to sell – the Arduino MPPT will be open source with code freely available. Inevitably it will change and evolve, so it may get smaller, but I can't say whether I'll ever put it all on one PCB. Someone else might do it though.
are you planning to put this in to a small(ish) pcb package like your other charger you are selling?
or would it be too much work to cram all this on to a single board?