Super Simple Solar Energy

Yeah, this can work. But when you say "most", that means more than half. Your choice of trip voltages will pretty much keep operation on the constan t current portion of the curve. That will certainly work a lot better than a simple fixed resistance directly tied to the solar cells, but it does gi ve up a significant portion of the available power.

I'm not adverse to using an MCU device to work with the direct tied load th rough a switch. It would monitor the voltage and current and control power to the load to get the full potential power from the solar cells. This ca n be done with fairly simple components on the low side only requiring a fi lter cap on the solar array output.

simple. They are made from electric furnace heating elements mounted so th e air flow is up. They will work fine with convection air flow but each ha s a 10 inch 240 volt industrial fan which helps to spread the heat around. One is at each end of the basement and plugs in a receptacle for a machin e tool. With a single heater running it takes about 20 minutes to have the basement comfortable in Jan.

How many watts are these heaters? I'm willing to bet they are a lot larger than 300 watts!

ote:

(MPPT) load. If you try to draw a fixed amount of power and the solar cell can't deliver that, the resulting power provided goes down a lot more than it should. Draw too much current and the voltage drops disproportionately .

ed voltage and for any given lighting condition you won't get optimum power . So in reality you need a MPPT load device instead. Not hard conceptuall y. It's basically a switching regulator but the thing being controlled is power to the load rather than voltage or current. Use a low resistance loa d and a standard buck topology can be used. A small MCU can measure the vo ltage and current into the load and dither it to find the optimum power poi nt.

ry

the market. That is what is good about SED. One learns things.

I said the combination of solar panel, MPPT and inverter will not work with a fixed, resistive load.

Okay, then, perhaps constant current is not the correct term.

Maybe power as a function of insolation is a better description. If you attach a resistor to a solar panel, its power output to the resistor will be linearly related to the insolation.

Yes, I understand that.

Yes, I understand.

I don't dislike what you say. But please consider the OP:

"So it would just be solar panels connected to a resistive load. No battery storage, no inverter, no temperature control."

Can you please explain why it will not?

Any resistive load needs to be regulated to the power level of the solar array

If you load less, you loose some power (constant current part of the curve). If you load more, no power since you operate on the constant voltage part

Resistive load must be changed dynamically since solar arrays change power point according to ambient temperature and sun intensity (fast transients from clouds etc)

I discovered that doing experiments in sunlight, comparing different possibilities, can be a frustrating experience. That's because the solar radiation is constantly changing. 10 or 20% changes due to a little haze, etc., are seen by the instruments, even though the sky looks clear. Add a few little puffy clouds, forget it. I came to realize that eking out every last 10% was an exercise in futility, and simplifying the electronics instead was a reasonable goal.

Of course I have. But in the brutal reality world of engineering, numbers matter. Some things won't work.

It's funny to hear people talk about how they would re-imagine our power systems, with solar cells and batteries and power lines from Florida.

e:

ote:

One of them is burning every more fossil carbon to power the ever larger ai r-conditioners requred to cope with rising global temperatures.

There aren't talking about "re-imagining" it. They are talking about re-eng ineering it. John Larkin find's it funny because he doesn't understand that it is necessary now, and getting more necessary with every year that every American dumps their 15 tons of CO2 into the atmosphere.

The fact that he and Donald Trump can't see the necessity isn't in the leas t funny. If they weren't such fatuous clowns it would be tragic.

the inverter-resistive load pair poses is a newgative resistance load on the mppt.

panel-mppt-resistor should work.

if you don't use antifreeze it can be directly connected to the house water supply.

e:

efficient. And plumbing solar collectors is a lot more work. Here ( Dela ware ) you either have to drain the system at night so it does not freeze or have an expansion tank to so you can use antifreeze .

You don't really want to. The "water" circulating through our radiators in our houses in Cambridge UK and Nijmegen in the Netherlands was dosed with a nti-corrosion chemicals as well as anti-freeze - I bought mine from Fernox.

The pipes were mostly copper, and most of the radiators were pressed steel, but there were also a couple of cast aluminium radiators below low window sills. I still had to vent a few millilitres of hydrogen gas from the syste ms every month or so, but the additives were pretty effective.

s slightly higher level heat to your environment.

vested, but any "efficiency" calculation would have to take the change in o utside world into account.

The bottom line is to get heat where you want it. While you're correct abou t efficiency you're being a bit pedantic. The heat pump will get more heat into the house than the resistor. That's all most folks want.

Ah, but it won't. That's the problem. Look at the curve for the current v s. voltage. Win posted a line for a data sheet of a panel. It shows the I V curve for different conditions including insolation. If you pick a resis tor that is optimal for one level, it won't be optimal for any other insola tion level, not even close. Win suggested a constant voltage on the output of the panel which is still not optimal, but it's not bad, much better tha n a resistor. The main problem is it doesn't compensate for the variation in voltage. It is likely a suitable thermistor could be incorporated into Win's circuit to adjust that threshold voltage to suit a given panel for te mperature variations and be very close to an MPPT device without actually m easuring the output power. It would require a larger capacitor on the inpu t to the regulator than otherwise, but it is simple.

range that is constant voltage (or nearly so) and a knee that joins the two . Forget about your labels and consider that curve. The knee is where the maximum power is obtained.

ally. Maximum power is at a point on the knee between the two.

at they say.

Yeah, so? If he does that he will get a really crappy system. He asked an d he has been told it will be crappy. Now people are trying to help him fi gure out how to make a better system that is still pretty simple.

I'd rather bang out a little code on an MCU or FPGA than construct a frame for some solar cells. I think that is the part of the system that will nee d significant attention and effort. So rather than design and construct th e array myself, I'd save on the money for the electronics by building that myself and buy the solar panel.

Just my opinion.

with a fixed, resistive load.

Because the inverter operates with a fixed voltage output into a fixed resi stance which will be a fixed power level. How can that possibly track the optimum power point?

An MPPT designed to charge a battery might work since it is designed to dri ve different currents. It depends on whether the MPPT is designed to outpu t a constant voltage or if it just lets the battery set the voltage.

The easy way to design an optimal circuit would be to optimize the power go ing into the load... or have no MPPT converter at all and connect transisto rs as loads with the control coming from a circuit designed to optimize the load power which in this case would be exactly the same as the power from the solar cells.

While measuring power is not so easy in discrete components, it is duck sou p in an MCU. Solar cells on roof, two wires to the basement, an MCU and so me transistors are all you need... oh, and a heat sink. A linear regulator would provide power to the MCU. As soon as the solar cells start to make power the MCU boots up and drives the transistors to the MPPT.

I don't understand your point. Eking out the last 10 or 20% is exactly wha t a MPPT does for you. In this case it is even easier since you are only t rying to make heat. You don't even need a regular MPPT converter. You jus t need a MPPT load. The required circuit couldn't be much simpler. Switch ers make noise and your design is suboptimal because it leaves power in the solar array. A linear MPPT load would not make significant electrical noi se, use very few parts and be very easy to design and build.

The poor intrinsic efficiency of the PV cells means that a well insulated black painted flat radiator under glass perhaps followed by a parabolic trough with a black pipe at the centre will capture a far greater proportion of solar energy and deliver it as hot water (or antifreeze mix). Kingspan, a pump and mild steel radiators are cheaper.

Plenty of designs about usually for heating swimming pools or bulk thermal stores.

That may be true, but neither does resistance heating or hot water heating. So, if is is physically possible to place the solar air collectors on the South side near his basement, that will probably be the cheapest way to go. Then if needed, he can get a dehumidifier.

Hundreds of solar projects here.

Mikek

te:

uth-facing

r area is

as a

Actually on the ground on the west side of the house near the basement. N ot visible from inside the house. I already have a dehumidifier.

Dan