Blocking Diodes

[walt]

Jim, some of what I said is from observing the voltages of the controllers.. you may have to go find the original patents (probably expired now) to get all the detail and then that may even be somewhat out of date. If you find something interesting, please post it.. as I said.. I am speculating on some of this. I have not tried googling this (which is usually a good idea first).

Regarding watching the voltages, if you have an MPPT controller, you can often see the search algorithm working with a DVM. The only way the controller can find the maximum power point voltage is to actually "visit" that voltage and measure the power. So the MPPT controller will do a voltage sweep by varying the impedance the panel sees. The sweep voltage range may be fairly small (.5 volts or ?? - the controller that did the sweep every 90 seconds varied the voltage by about 5 volts) and if the actual max power point voltage is outside the sweep range, the next sweep will start at or near the higher power end point and the controller quickly heads to and finds the actual voltage maximum power point. Even if the sweep is too fast for the DVM, you can still see the voltage dithering around the max power point.

Since the thread has drifted.. another interesting thing I saw recently measuring my Victron MPPT controller with a 160 watt solar panel. Pretty much all of these MPPT controllers require that the solar panel voltage be higher than the battery voltage before the controller will operate. This likely keeps the switcher topology a little simpler but it might seem that you are losing out on some of the solar power available by not trying to convert the power at say 8 volts (way below the battery voltage).

But.. my 160 watt solar panel was reaching 12 volts with a load of only 4 ma. By the time the controller turned on at around 16 volts, the current the panel was producing was still under 10 ma. That is a tiny amount of power.. on the order of 1/10 watt and less than one thousandth of what the panel is capable of. Switching converters need some power to just operate and would consume all of that just turning on.

So... these MPPT controllers dont turn on until a the solar voltage is somewhat above the battery voltage - because there just isnt any power below that voltage worth converting.

 

[Sumner]

....So... these MPPT controllers dont turn on until a the solar voltage is somewhat above the battery voltage - because there just isnt any power below that voltage worth converting.

Well the last few posts by you guys have been well over my head but I for sure agree with the above. You can see voltage pretty quick on a panel, but is there any real amperage at that point? I see where people put panels in series with a MPPT controller to get a higher voltage at the beginning and end of the day but personally don't feel that they are harvesting that many amps at the time to warrant the series configuration and feel that with any shading later the series wiring can result in larger losses.

Saying that, on the Mac I do have two sets of panels wired in series since they go to a 24v MPPT controller and from there to two 12 volt batteries in series that can power a 24 volt trolling motor. My house bank is two 6 volt batteries in series with a MPPT controller fed by 3 12v panels in parallel. They are different sizes (wattage) but the same type and voltage output. Not ideal but it has worked fine,

Sumner
====================================================================
1300 miles to The Bahamas and Back in the Mac...
Endeavour 37 Mods...
MacGregor 26-S Mods...http://purplesagetradingpost.com/sumner/endeavour-main/endeavour-index.html
Mac Trips to Utah, Idaho, Wyoming, Canada, Florida, Bahamas

 

[JimInPB]

Well, I’ve done a fair amount of reading in the last day or two. The biggest thing I came away with, is a new appreciation for the apparent wide variety of controller designs that are now out there. It’s really hard to make a lot of good general statements. Most well qualified statements will need to be controller specific.

My original statements in my original post were based on a lot of assumptions & expectations, which is always a bit of a risk. I now see that some unexpected conditions exist.

The biggest surprise that my recent reading turned up was the wide variety of different technologies that MPPT units employ. MPPT does not seem to denote a single type of regulation, but rather it defines a category of related regulators. Right there, the waters get a bit murky & comments about MPPT need to be qualified to specific systems.

Much of what I said in my first post is true if a PWM regulator is used, but is not as likely to be correct if a MPPT unit is used.

I was unaware that most MPPT units have transformers in them. Some have just one. Some appear to have a range of cascading transformers or cascading taps on a single transformer to improve current boosting performance over a range of input voltages. I had expected that the controllers would be purely solid state devices. The transformer is where their ability to convert excess volts to increased amps originates. I had missed that important detail in the beginning. I had thought that we were talking about blocking regulators, not converting regulators. That was a bad assumption on my part.

My assumption that a regulator would have a minimum voltage drop of .7vdc also turns out to be incorrect in some cases, as it appears that some controllers employ a bypass relay to continue charging when primary voltage from the panel drops near the borderline region & then the relay drops out again when actual battery voltage is reached. Again, I had expected a purely solid state device, so I expected to see at least one PN junction drop across the device at all times. In some controllers this appears to be true. In others it does not.

There were a few physical properties of the solar panels themselves that I was surprised to learn of. I was completely unaware of the Fill Factor effect, which would seem to be part of the reason why MPPT would be important in the first place. I also grossly underestimated the effect of temperature on the panels’ output, which is not so much an issue related to the efficiency of the controller, but is certainly noteworthy when planning the size of your system. Cold panels work much better. Those of us in warmer climates can expect decreased performance.

Panels seem to vary as much as controllers do. The relationship between open circuit voltage & available current does not seem to be consistent from one type of panel to another. Here again, it gets tough to make reliable general statements due to variations in panel types.

I still have a lot more reading to do. It appears that I am just scratching the surface on the details of this subject.

All of this has been very interesting for me, but at this time, I’d like to get back to the original question regarding the need for diodes before the controller. I think that it would be interesting to set up two identical panels in parallel, with an ammeter on each & then shade one of the panels. The resulting amp readings would be a good basic piece of information to have in order to decide if any diode is needed at all. I’d like to know how much the panels actually do back feed in the real world when directly wired in parallel. I suspect that it probably varies with different panels. I suspect that the original controller manufacturer’s warning to always install blocking diodes may have been a blanket safety directive. If this is the case, then the diodes may not be needed in some cases, but at this time, that is just a guess. I’d love to do some experiments & develop some hard empirical information from which to derive more reliable conclusions. At this time, I’m left with more questions than answers.

 

[dLj]

I'm watching this thread with great interest as I may be installing a new system next year on my boat. It does not have any solar system at all so I'll be starting from 0. Which may be a good thing, I guess. As there are a number of other things that will need attention, might be the year after next, time will tell. But I am very interested in learning as much as possible on this subject as most certainly sooner or later I'll be installing a new system.

As I understand it, panel technology is changing quite a lot with much more efficient systems coming out in the not far distant future. I would most welcome posts with links so that others can research multiple aspects of these systems.

So far this has been a very interesting/informative discussion.

dj

 

[Bill Roosa]

MPPT is just a way to get max POWER out of a panel. It is an impedance matcher as it sets the panel voltage and current for max I*V (I and V are related in panels) and then delivers it to the battery at an appropriate voltage for charging (depending on SOC)
That impedance matching is done in the same fashion as a boost/buck converter to step up or down the voltage (and also current since V1*I1=P=V2*I2 still applies)
Not at all sure why every one is concerned about shading as those diodes (should) already be installed in the panel every three cells. The blocking diode is just to eliminate night time leakage from the battery or prevent panels from back feeding when one is putting out more voltage than the other. Most controllers already have these in their design so adding another one is pointless.
Now an clever EE could design a circuit that senses the light level and use it to close a mechanical switch at sunrise and open it a sunset if your really that concerned about <4% waste from a blocking diode. Not thinking an twin comparator/opamp would draw more than 0.01 amps and the light sensor draws nano amps. the rest of the circuit would only be on during the switching operation so it could draw 10 amps for 0.1 seconds (10*0.1/60=1/60th AH) aka noise.

 

[Brian D]

Bill, no need to be clever. Just go buy a garden solar light and use the light sensor from that device on a panel of relays.

Or even better, get one of those mechanical garden timers that turn on/off the garden lights. Set it for average sunset and average sunrise.