Higher voltage means less current on the PV side inside the controller, so, unless the manufacturer cuts corners, less heat to deal with from resistance in tracks, current through switching components, windings in inductors, shunts etc. Pretty much all MPPT controllers will be using FETs to do the switching, and FETs act like resistors, milliohms of resistance when full on (higher current devices usually have higher on resistance), so the more current flowing the hotter they get.
On the flip side you need to withstand the higher voltage so increased creepage on the board layout makes for a slightly bigger board, higher voltage specs in capacitors and transistors etc.
When talking efficiency it really boils down to the design of the power supply. Most of the buck/boost supplies people will come across use FETs. FETs act like capacitors on their gate pin, and in other ways too. The faster we have to switch them or the more often, the more current we have to push in and pull out to get them to snap on and off quickly (to avoid them being in not fully on, not fully off state and really start to get hot). High efficiency buck/boost supplies don't just turn on and off at the same frequency but varying pulse widths, they also do tricks like varying the switching frequency, drop switching cycles completely as they start to get towards their extremes of operation. For a buck this can mean the lower difference between in and out will result in less switching of the FET and presto you have more efficiency based on voltage difference. There's also losses in diodes for the same reason, but high efficiency buck/boost circuits will replace the diode with another FET, synchronous rectification.
There are so many many weird and wonderful switching power supply topologies that you could write (and there are) page books on the subject. Even things you may not think of such as using the resonance of the output inductor to increase or decrease the voltage depending on the frequency of the switching action. Panasonic (I think it was, time fades memory) loved this type of switching regulator in their VCRs.
The disadvantage is that the greater the voltage differential (Vmp to Vbat) the lower the efficiency. It could be a few percent. Once you get above 50V then you need to take shock hazards into account.
You are referring to the efficiency of the DC-DC conversion here, right?
12V arrays have some disadvantage at extreme temperature. Once panels are above 12V then there's no temperature advantage of a 48V over a 24V panel.
This has more to do with relative voltage, right? Differential between Vmpp and Vbat.
For shading, the advantage goes to series 12V panels as opposed to 12V parallel panels. If you have a 12V battery with 24V panels then there shouldn't be any advantage. Bypass diodes will work equally well in both configurations.
I don't understand why 12v panels in series would be the best option for shade. This seems to go against conventional wisdom for shade tolerance unless I'm misunderstanding something. Can you clarify why this makes more sense than parallel connections?
You understand the concept of bypass diodes?
I thought I did, now I'm realizing I understood how they worked in concept, but may have misunderstood the context in which they are beneficial.
Lets start over. Given a 12V battery and three 12V panels then the series option is better for shade.
You understand the concept of bypass diodes? If one section of a panel is shaded then those cells may produce just 20% of the current the other cells produce. Since the 12V panel contain 36 cells in the series string, the unshaded cells force the higher current through the shaded cells. The shaded cells act like a resistor and will heat up and possibly burn the cell open. The 36 cell panel may contain 2 bypass diodes, each one across 18 cells. The diode associated with the shaded cells will conduct. There's no force feeding the shaded cells with high current. The panel voltage is now cut in half. Vmp decreases from 17V to 8V. However 8V isn't going to cut it charging a 12V battery. Two of the three panels are producing 66% of unshaded power.
If you got the 3 panels in series then 2.5 panels are producing 83% of unshaded power. Vmp drops from (3 * 17V = 51V) to (2 * 17V + 8V = 42V) at full current.
All clear?
Well your explanation is quite clear and I can see the logic in the example you gave. What I am struggling to understand is how to reconcile your example with the conventional wisdom that I've heard many times over of wiring in parallel when shade is an issue, and the little bit of real world testing I've seen where shading reduces current and voltages remains relatively constant, and parallel wired panels exhibit less power loss when partially shaded (one example here).
Can you explain the discrepancy, is the conventional wisdom outdated wisdom that doesn't apply to panels with bypass diodes built in? Or is it possible that your explanation is outdated or overlooking some factor? I feel I am still misunderstanding something as I'm struggling to reconcile what you are saying (which makes logical sense) with what I've heard and observed elsewhere.
*IF* you are making high volts/low amps keeping the volts as close as possible (or higher) means efficiency.
As Gnubie said, less losses to heat, induction issues, etc.
Since I switched to AC Coupled micro-grid, and I started with smaller panel strings/voltage and moved up,
The higher panel string voltage is more efficient in the inverter.
While it's DC voltage that has to be switched to produce AC, it doesn't have to be 'Stepped Up' or 'Stepped Down' through Inductors, so the less it has to be 'Worked' the less losses you have.
What I'm currently doing is looking for where my panel voltage gets through inverters with the less losses.
I add panels until losses start to rise, then reduce the panel string voltage until efficiency starts to rise again.
'Clipping' is wasted power, heat is wasted power, so I simply look for the point where I don't have heat or clipping losses.
NOT running the inverter at it's absolutely maximum, around 80-85%, seems to be where I get my best efficiency,
The thermocouple says the heat sink is happy (but I haven't got into summer heat with the new system yet), and heat is lost power...
According to the manufacturer, a cooler running unit is a longer lived unit, it's supposed to be built to take maximum input and 'Clip', but according to the factory engineer maxed out isn't good for anything...
He said what I found, about 80% is where the unit will be happiest, and I want longevity.
Since this ISN'T the typical small solar system, different rules apply,
The 'Grid Tie' inverters produce directly in 240Vac, are parallel connecting on the AC lines, and communicate through those AC lines.
It's really stupid simple to connect these together, but it's infrastructure intensive...
I run the Grid Tied inverters to produce AC power directly from panels and put it on common AC lines,
BUT, and it's a BIG BUTT (like HUGE), I need a hybrid (battery) inverter to produce the sine wave the grid tied inverters need to syncronize with.
That inverter MUST be at least 10% higher capacity than my total grid tie inverters (plural) combined,
And it MUST support frequancy shifting, this is how the inverters communicate via AC lines.
Having the big (and costly) hybrid inverter increases inverter cost,
But it reduces battery cost, the battery only having to support your system while there is no sun situation,
You would be surprised what a 96% efficient panel string & grid tie inverter will produce on even the most overcast days...
(No 35% losses to battery & battery support)
I don't know if this will work for anyone else, but I suspect anyone that needs to power home, small business, basic homestead (not big farm) it would work for them like it is for me.
I'm going slow, trying to figure it out as I go, squeezing as much information out of the manufacturers as possible to maximize this,
I'm by no means the 'Last Word' authority, this is all new to me, but I do have spare inverters if something cooks, so I'm experimenting a little...
A few extra inverters,
Literally a truck load of inverters...
How does one choose a panel?
I have a 400ah lithium battery, 13.3 resting voltage, 14.4 charging.
I was looking at the panels available. I would like 2 panels of 200W each (that's pretty much what fits on the roof).
Most panels come in 18V and 36V version.
I guess it's for PWM controller in 12V or 24V setups.
But, what about MPPT?
I have a Victron 100/30.
Should I get the lower voltage, but higher amperage wired in series, or the higher voltage, and lower amperage wired either in series or parallel?
Would there be an advantage to either?
Thanks!
I would always choose the higher voltage panel (all other things being equal). I am running 250w panels (30.3v / 8.37a) in series sets of three to bump the voltage up to ~92-100vdc and then combining three sets (parallel) into a combiner box.
Ultimately you just need to make more voltage than 14.4v, so either style panel will work. With the Victron 100/30 you will be able to run your panels in parallel or series, so get the less expensive option?
With MPPT, you don't really care about the "voltage" of the panels, meaning you don't care if it's a 12v or 24v panel.
Instead, you are primarily concerned that the total voltage of your panels when combined in series does not exceed the Maximum PV Input Volts of the solar charge controller. If you are in a location where it gets cold, you need to leave room for the voltage to climb when it gets cold.
Panels in series, the voltage adds up and the amps remain the same. Panels in parallel, the voltage remains the same and the amps add up.
How does one choose a panel?
For me, in order:
- what the controller will tolerate. Usually not an issue, but some DC-DC chargers have very low input voltages (~23-25v); in that case the 24v would not be acceptable.
- what fits the allotted space
- $/watt
Most panels come in 18V and 36V version.
- nominal 12v / 36cell / Vmp ~18v
- nominal 20v / 60cell / Vmp ~30v <-- somewhat less common, but they do exist
- nominal 24v / 72cell / Vmp ~36v
Would there be an advantage to either?
24v is more common in 200w panels than 12v, so I'd expect them to be a bit cheaper by the watt. I suspect 12v 200w is a niche product that exists to meet the needs of people running 12v systems and PWM controllers. Hopefully someone who knows more than me will chime in.
I like paralleling all panels so if one panel is shaded, the others still push amps. This sounds like a great idea, until I balanced that agains wire size for a 50’ run with a 36 amp run for 3% loss and the fact that 10 gauge wire is easier to work with because among other reasons, you can use MC4 connectors, which are limited to 30 amps.
With my six panels, I ended up with a 3S2P set up with 100 watt panels. That’s just the way the math worked out for voltage loss on the wires balanced with high amperage. By going with three panels for a total of 48 volts in series, that made the voltage loss less than 3% for the length of the run.
Others are not that concerned about my 3% loss number for my panels, and find it better to add an extra panel or two to compensate.
I then came up with two other arrays with two separate charge controllers. The second was a portable array. The third was 350 watts of flat panels on the roof,
My second array I added to power my RV was four portable panels, 100 watts each, all in parallel. This works fine. I even tilt it into the sun and because the tilt angle allows more early morning production and evening production, these 400 watts of panels throughout the day can produce more power than the 950 watts of panels I now have on my roof.
alright.
on a voltage drop only stand point, it would seem that using two 36V panels rated at 5.5A each, wired in series will keep the amps at 5.5, and ramp the voltage to 72V, which my MPPT can handle.
That way I can use awg10 wire, with only a 0.2% loss over my about 4.5m of wire.
I'm usually parked at the beach, and not much shade, so both panels would always be in the sun, but not tilted.
Reasonable thinking?
My box is already at the highest legally possible, so, adding a tilt mechanism would make it too tall.
Also, it's almost perpetual summer here. 25C for about 8 months of the year, and above 30C for some 6 months out of them!
We got 38C today....
Here's a capture of the panels I am looking at. These 200W basically all have the same ratings.