Data loggers...

Re: Data loggers...

V=I*R
P=V*I=V^2 /R=I^2 *R
R=V/I

For a reasonable approximation, the Vmp of the panels is set by temperature (as temperature rises, Vmp falls). As a second order effect, current rises a bit with temperature.

Say you want Vmp=15 volts and a 50 watt panels:

• R=V^2 / P = (15 volts)^2 / 50 watts = 4.5 Ohms

Now, you are trying to show that a solar panel with cooling will output a bit more power... And for the most part, that decrease in temperature causes a slight rise in Vmp--Which, would seem to indicate a slight change in the Resistance value...

In practice, perhaps you can simply monitor Vmp/Imp at a fixed R value and come to some conclusions.

To prove (or disprove) your theory, in the end you will probably have to setup a pair of matched MPPT type charge controllers with appropriate load banks/batteries+loads.

Notice that your setup will only produce "additional power" for MPPT type charge controllers.

PWM controllers will not see any additional power (other than a higher Vmp because of the "panel cooling" may charge/equalize a battery better on a very hot/windless day if the panels have marginally low Vmp ratings).

-Bill

Re: Data loggers...

Well--Filament lamps tend to be constant current devices (as voltage falls, filament temperature falls, resistance falls).

Off-Grid Inverters are constant power loads. P=I*V... As voltage falls, current must increase to keep power constant.

Perhaps, your "cheapest" and most efficient load would be to look at using a few Enphase Grid Tied Inverter "bricks". They are true MPPT devices, one per ~200 watt panel, and will attempt to gather 100% of the available power and drive it to your grid (so you don't waste power with dummy loads, battery maintenance/control, etc.).

Since you are looking at cooling the panels--you could do a "quick screen" of what works best (say you have 4 different cooling system designs and an unmodified panel for baseline).... Just setup the five panels and monitor panel temperature. You already know the temperature offsets for Vmp, Imp, and Pmp--so just monitoring a panel set for cell temperature (thermocouple or IR system) with either no load or even a dead short (some worst case self heating). Figure out the "best" one or two.

Then get 2-3 Enphase units and wire up to the 2-3 panels and power them from 240 VAC. Either monitor the Enphase with your data loggers, or get an Enphase box (not cheap) and see what data you can pull from the data boxes without paying for the web subscription (or perhaps web subscription would be nice).

You could rotate the 2-3 Enphase once a day or once a week to double check that the Enphase units are matched (or determine the wind-age factors).

-Bill

Re: Data loggers...

LED's are very non-linear loads (typically need some sort "ballast" to limit current / make LED function independent of voltage).

I am not saying the lights will not work--it is just that they are not fixed resistance... For some applications they are very nice loads.

The Temp/Current sensor you link to will only work with AC current (uses transformers) and will not work with DC current (at least the model you are looking at).

It really depends on what it is you are trying to measure/emulate/quantify.

For the most part, people that would be looking at your system will be using MPPT type charge controllers--which are not constant power, current, voltage loads... They are a complex load that adjust the operating points based on solar panel conditions (amount of light, temperature, number of cells, etc.).

You can use any load type you want (resistor, lamp, current sink, voltage based load, etc.)--but you would have to do some hand-waving/calculations to transfer that data over to how a MPPT type controller would operate (improvements) with your panel cooling jacket.

The cheapest/easiest method I can think of is to use the Enphase (or similar) type GT inverters. as your active panel load... They have MPPT front ends and use the utility network as an unlimited load. Their remote monitoring may be accurate enough (and provide enough data) that you could forget the rest of the data logging stuff (don't know how accurate they really are--but measuring kWH output from each Enphase brick would not be hard or expensive (to compare with logged data).

-Bill

Re: Data loggers...

silent wolf wrote: »

Would this work for a resistance???? It is a MPPT.

http://www.rvfunproducts.com/solar-c...mp-p-2236.html

A big problem may be:

MPPT includes a sweep-function which runs through the whole solar-panel voltage range once every 2 hours to find the point of the absolute maximum power output.

That might confuse your monitoring if it does a "re-track" after two hours (not sure that this makes sense other than there may be an issue where the controller looses it marbles once in a while?).

In any case, you need to provide some sort of controlled loads to the battery bank to ensure that the controller does not go into absorb/float mode and start to cut back on the energy harvest.

I still look at the Enphase as being cheaper (one $200 unit per panel) and getting rid of the whole battery/bank/load management issue.

It would also seem that if you are into monitoring--you need to log wind speed, air temperature, panel temperature, water temperature, water volume/flow, etc... There are a lot of variables and if your panels/measurements end up in the 5-10% accuracy range--and you are looking for a 5-20% improvement in panel output with cooling--it will be difficult to separate the improvement from the noise.

-Bill

-Bill

Re: Data loggers...

The Enphase (assuming that it works as advertised and it has a relatively effective MPPT front end) does address your "active load" issue very nicely. It is the only small / low(er) cost single brick device that I know off (or the SWEA GT inverter and equivalents might be an interesting option--probably not any cheaper, and they don't have any monitoring that I remember) that I can think of off-hand that I would use to do this experiment (as I understand your needs).

Here are a couple discussions about micro GT inverters:

Micro grid tied inverters on E-bay
Enphase

Now, the Enphase does have lots of data locked up in a proprietary interface (power-line communications / monitoring system)...

The communications box is $350 and has some "local" information available--but, as I understand, probably a crippled output so that you will pay some $x per year per inverter for monitoring (data is shipped back to Enphase subscription website and represented back to the GT system owner)...

So, you are left with monitoring the DC input or AC output with a data logger, or getting an "Envoy" data hub for $350 and pulling what data you can out of it, or getting an Envoy and a web subscription and using what data the web interface supplies...

There are a few folks here trying to figure out how to get more data from the Envoy--perhaps you can talk with them (I don't think anything has come of that yet).

The Amazon link is broken--so I could not do anything there.

Regarding the 2 hour re-track---That is just it--I don't know what will happen.

Not all MPPT algorithms are created equal. And it can be very frustrating for you as you dig through the data and start trying to separate the effects you are looking for (cooling of solar panels) from the confusion of a mix of equipment--each with its own peculiarities.

-Bill

Re: Data loggers...

silent wolf wrote: »

What does this necessarily mean?
three phase power

All it means (I believe) is that the AC output voltage of the Enphase GT inverter your link pointed to is designed to operate at 208 VAC nominal voltage (a voltage typically associated with some versions of three phase industrial power) instead of 240 VAC nominal line voltage (183-229 Vac vs 211-264 Vac).

All you need to do is identify if you have 208 or 240 VAC service and then pick the correct Enphase Inverter for your service.

So would you think I still can get accurate readings from these smaller panels during the day? I dont need it in the morning right away, just as long as it starts around 9:00.

Again, I have never even held an Enphase inverter in my hands--As a rough estimate, the inverters are 95% efficient, so that means their worst case operating loss is ~5%--Or 0.05*200watts=10 watts of tare losses.

So, my guess is that any wattage (P=V*I) that is less than 10 watts output will give you variable results (i.e., the Enphase internal electronics will be cycling on and off while attempting to load down the panel enough to start selling power). ( I would guess that they could (again, worst case guess) need upwards of 20 watts of panel power before the GT inverters stabilize their "MPPT Load" for your needs.

That is why I suggest that you only get one of the micro inverters before you layout "bigger money" for an entire set to ensure that they meed your needs.

Overall, Tomorrow I am leaving for isweeep early morning, I am hopping to get with solar companies and get 200 watt solar panels. I have already been talking to SOLON since there is a manufacturing plant in tucson where I live.

I would prefer the 200 watt panels--that would give you a wider "active load range" vs panel size to minimize other sources of errors (should make the start-up errors 1/4 as large).

So couldn't I use a fuse for a load then?

I believe the comment is for fusing the voltage measuring leads on the shunt resistor--The issue here is that the shunts may be at some other voltage besides zero volts... The remote/data logging volt meter leads come from the shunt to the DAC unit... Any short circuits in the voltage leads/DAC/etc. side of the system could bring the full amperage from the shunt down through your volt meter leads (typically small gauge wire because there is no current).

What is typically done when connecting to the voltage points on a shunt is to use small fuses (1 amp or less) or some resistors (100-10,000 ohms -- depending on the input resistance of the DAC/voltmeter--and they make fusible resistors too) to prevent any "down stream shorts" from causing your logger wiring from catching fire.

If you are using shunts on individual small solar panels, the fuses may not be needed. If you are using shunts on a DC battery bank and the shunts are not at ground potential (i.e., in the ground leg of the battery bank)--then you very much do need voltage sense lead fusing/resistors.

-Bill