Solar PV regulators/controlers

Your caps are too big, or you're sweeping too fast. Insolation changes very slowly. Sampling charge current averaged over a second is way faster than you need.

We used a tiny microcontroller. Initialize the PWM controller (aka buck converter duty factor) to nominal vsolar / nominal battery volts. Measure the current via your favorite method and stuff it into the A/D converter. Change the duty factor by one increment and keep going until the current drops. Reverse/repeat.

Nice thing about solar is that it's self-current-limiting. Unless you do something stupid, you can't hurt the switcher or the cells or the battery(in the short term). But you do need a cap on the solar supply.

Into current driven loads like batteries or LEDs going from linear to PWM just moves the energy loss from the pass element into heating the load.

Thinking about all the responses, and refreshing my knowledge a bit too, it sees that a boost converter might be good for me, albeit one that maintain s a near constant input voltage at the MPPT, and near constant input curren t. I have 12v panels, but 24V/28V electrics and batteries. I had anticipa ted connecting the panels in pairs for 28V MPPT, but it now seems that 14V with boost is the way to go.

How much does panel-to-panel variation matter in series strings?

Imagine one part in shade.

Parallel might be better.

What we did was something like this:

A non-microcontroller system, which is controlled by small pertubation around the MPPT point. MPPT point is in principe found when Di/dt = dV/dt

For space applications we could not rely on, and in fact did not have a microcontroller handy

The converter was a super-buck converter, with low input ripple current

Yes, but with CCM you define that pertubation. Keep it below 1%

We were always below 1% from the MPPT point, even with excessive array capacitance

Cheers

Klaus

Can you suggest a reference or two that describes your specific super-buck converter design? Or one that's close? They look rather complicated.

it sees that a boost converter might be good for me, albeit one that maint ains a near constant input voltage at the MPPT, and near constant input cur rent. I have 12v panels, but 24V/28V electrics and batteries. I had antic ipated connecting the panels in pairs for 28V MPPT, but it now seems that 1

As within a panel, the output is limited by the least effective cell in the chain.

The one in the shade can actually get back-biased by the others. That might confuse an MPPT controller. Maybe it has the same photocurrent even when reverse biased.

Aren't single panels already series:parallel arrays?

In an array of large-size panels, we're often happy to lose one, so long as it doesn't affect others. It's more efficient to place panels in series, for higher-voltage lower-current output wiring. But then we add an optimizer in parallel to each panel. Alternately, each panel can have a micro-inverter, and the 230Vac outputs are wired in parallel. That's what I have. But I cannot get power from my panels if the grid is down. OTOH, a DC system could be running charging a battery, and there's an inverter for the house.

What does the optimizer do?

One might put a tiny boost converter on every string, or even every cell, all dumping MPPT current into a common DC bus. Glob-top them on the back of every cell.

a diode would do

Probably easier to parallel the panels though. The current needn't have far to go.

NT

Do what?

Getting optimum power out of every cell would be cool. The chip would sell in the billions.

Running electronics off 0.5V is doable, but not easy.

There are several boost converter chips around that work at or well below 600 mV in.

The manufacturers usually provide a diode inside the junction box of the panel for that purpose I think.

once you get it started it's easy

Wouldn't a chip on every cell use a lot more quiescent current? Any given cell doesn't output tons of power.

Rick C.

Figure 9 on this page:

But, I did not work on the converter, so I have no theory on it. I was responsible for the MPPT control

I will see of I can find anything.

Cheers

Klaus

Very close to our design: