Open Source MPPT

[OceanSeaSpray]

Here is an interesting paper about GaN FETs, some of you might already be familiar with it. They do have some notable difference from MOSFETs. They are more sensitive and not always as forgiving. They (those?) are not (cannot be?) avalanche-rated either and so not exceeding VDS is even more important.

We shouldn't let any charge control issues come in the way of the MPPT project, there is time to find pathways.

Regarding specs and after looking into GaN FETs, I would be increasingly tempted to favour a design intended for say ~300W. There are lots of smaller installations that would have no use for a large unit and would incur unnecessary cost. Even just cost+complexity aspects could well warrant designing a more modest converter and installing two instead of one large one, splitting the solar capacity.
Of course, that also means 300W panel max, but realistically, on a boat...?

I can see that an interleaved buck converter of that capacity could be small and low cost compared to a large unit, it would look like two of the power stage of the small Genasun unit, off-the-shelf toroidal inductors hopefully and within the thermal capabilities of the GaN FETs without having to deal with significant heat removal. GaN FETs are not as well set up for heatsinking compared to MOSFETs.

[whiskthecat]

Okay, well I have the panel now, so that's a pretty decent start.

[whiskthecat]

Could a mod please remove the poll? Going with 100v.

[thomasow]

Originally Posted by whiskthecat

Okay, well I have the panel now, so that's a pretty decent start.

Pictures

And how are you looking to mount this, 96" or so isn't it?

-al-

[whiskthecat]

It's sitting on a ping pong table which is sitting on a pool table just to give a reference for scale in addition to the soda cans. I've been having nightmares about how I'm going to mount it but none of them have been all that useful. Any suggestions beyond what I'm seeing on google images would be more than welcome.


So it looks like during our discussion the next gen gan fets are now available for purchase but in 10qty at $10 per unit, ouch. They are substantially better though.

[Lake-Effect]

here's a MPPT project using Arduino

[whiskthecat]

Thought you guys might wanna check this out. Quite impressive.

http://www.gansystems.com/_uploads/w...0converter.pdf

[OceanSeaSpray]

Thanks for posting it. It is impressive, especially the lack of ringing considering the slew rate of the signals. The PCB layout might have given them a bit of work. The inductor on the photo must be hiding some HV capacitors.
One difference is that it is a low current application, very suited to GaN FETs.

[sailormed]

whiskthecat: This is the way i mounted my solar panels on the SS tubing of the frame. The model depends on the tube diameter, you can find them in several sizes.

[whiskthecat]

Nice mount. I just need to figure out how to get my stainless railing from waist height to above head height.

[whiskthecat]

Okay, a couple of updates. I've yet to make the plunge and order any parts but in the meantime I have acquired a large 11 inch pulley for a prop shaft alternator and I've repaired my old Rutland wind generator and will probably replace it with a newer model. So, the need to MPPT things is increasing fast.

[whiskthecat]

It seems I will be putting this off for quite a while now as I've found a workable solution. I was messing around with a 20$ 385 watt PlayStation power supply to make a cheap 13.2v (Adjustable) battery maintainer for my bilge pump when I had the realization that 90V DC from the solar panel wasn't too far away from 100VAC that the supply was intended to run on. Sure enough I hooked the panel to it in bright sunlight and it worked. For now I will just use a couple of these units on my panel and prop shaft generator to at least maintain battery voltage and I'll likely add on a microcontroller and some current sensing to handle charge and maybe even MPPT. You can't beat 20$

Here is a link to what looks to be the same model I am experimenting with.

[whiskthecat]

If anyone is interested I can document the process. It appears these power supplies have over current protection so I can likely tap in to current sensing that is already present for mppt. Right now all I have done is modify a resistor value near the 2.5v feedback reference. This allows me to adjust from the default 12.4v up to about 13.7v when the over voltage protection kicks in. This can easily be disabled to allow any voltage up to the 16v the output capacitors are rated for. I suppose it would be best to modify the OVP to kick in at 15v or w/e would be appropriate for flooded lead acid charging. I have access to a lot of these power supplies to experiment with.

[whiskthecat]

Just a quick update. My modified ps3 power supply has been keeping my battery floating nicely at 13.4v with the bilge pump running every 3 hours for 3 minutes. (The magnetic water level sensor failed closed circuit so I wired up a small 16 bit MCU as a timer) I've just ordered a 48v alternator @ 266rpm (192v @ 960rpm) that I will be using as a propshaft generator and I'm fairly sure a ps3 power supply will work nicely here as well. I will report back when I know more. For now the plan is to modify a few more of these power supplies and get all of my power sources (solar, hydro, wind) attached to keep the battery held at float voltage.