Am getting ready to configure an array for a new solar system. In order to choose the optimum array volatge, efficiency curves would be helpful.
Have looked at the Outupt current vs PV in voltage at a constant battery voltage, but all of the Outupt current difference twix different PV In voltages cannot be attributed ONLY to changes in efficency ... can it ?
Had considered running 120-ish volt nominal PV inut to a Classic 150 (in my moderate CA climate). The PV combiner to CC run is fairly short, but 120 vs 90 volt PV would reduce the number of strings, saving a couple breakers and conduit/wire runs.
Curves, or at least efficiency data would help with that tradeoff.
Thanks, Vic
Yeah, we really do gotta get those efficiency curves up. They're not all done yet.
You are correct though that the power ratings do reflect the efficiency of the Classic... sort of.
Since the maximum power ratings are based on maximum temperatures inside the unit, the curves will
show that higher input voltage give you a lower output power rating and that's because the
hardware efficiency is a bit lower there.
But wait ! There's more ! The efficiency goes up as the output (battery) voltage and output power
go up, but the output current goes down because the temperature goes up with output power.
Hopefully I've confused someone. I confuse myself some times.
The Classic follows similarly to some other charge controllers out there that show efficiency with varying input voltage and output
current vs. output voltage of course.
We'll get some efficiency curves soon I hope. I've been spending time in production lately and doing software as i can.
boB
Morning, boB,
You ARE a Night Owl !
OK, realize that the Classic topology is probably considerably different than the venerable MX-60. In both of the existing systems here, run 106 V PV in to the MX. During Bulk or Asorb, the MX's fan does not run, even delivering 2500, or so watts to the battery side of things. But in Float, Vin usually rises to just above 120 V in, and even delivering about 120-150 watts, the fan does run. Has always seemed that the conversion losses rise at some exponential rate.
I assume that this goes with the territory, in any of these big switchers.
Currently, have one Classic 150.
The new array is composed of SolarWorld SW-245s. STC Vmp is 30.8 V. So, I can run strings of three, or four -- 92.4, or, 123 ish Volts. The array is 24 panels, really too large for just one Classic 150 (5880 STC watts). The one-way distance from the combiners is less than 30 feet. So, my guess is that you would council strings of three -- 92-ish Vmp for my 1280 Ah 48 volt bank.
Thanks for the bandwidth, boB. As a Hardware engineer, always loved to get out into Production to see how my designs really worked in volume. Great learning experience. Know that you have been at this Solar stuff for decades, so Have Fun. 73, Thanks, Vic
Well, the topology is very similar in that it is a buck converter. They're pretty much all buck converters.
So, yes when the Classic goes to Float, the input voltage will rise, BUT the output current is so much less,
(usually), that I am surprised that the fan would come on for that, but who knows ??
73
boB
boB,
1. Thanks ... on the MX-60, the fan would cycle, and now run A/C in the power room (manually) in Asorb/Float during normal Summers, and this, along with a small Aun Fan for the MX keep the internal fan from running. (surprised that you could decode the orig post due to all the TYPOS !!).
2. But, regarding the Classic 150:
Several things, do not quite understand that the ultimate current available at 58.4 V with either a 110 V or 120 volt array seems the same.
AND, it seems unlikely that the difference in max current available with a 90 V, and 120 volt array is COMPLETLY due to efficiency losses at the higher Vin, and therefore shows up as waste heat. Which would be about 400 watts.
Do not wanna waste too much of UR time. Guess am talking myself into running 92 volt nominal strings, as runs are short, and that still gives lotsa margin for EQ.
73, Thanks, Vic
Quote from: Vic on June 04, 2011, 10:03:30 PM
boB,
1. Thanks ... on the MX-60, the fan would cycle, and now run A/C in the power room (manually) in Asorb/Float during normal Summers, and this, along with a small Aun Fan for the MX keep the internal fan from running. (surprised that you could decode the orig post due to all the TYPOS !!).
You mean Aux Fan ? Nope. can't read any of it ! You've read that recent article about how the letters don't really have to match in order to be able to read something ? Well, the beginning and end letters are supposed to be in the right place but I get the drift...
Quote from: Vic on June 04, 2011, 10:03:30 PM
2. But, regarding the Classic 150:
Several things, do not quite understand that the ultimate current available at 58.4 V with either a 110 V or 120 volt array seems the same.
AND, it seems unlikely that the difference in max current available with a 90 V, and 120 volt array is COMPLETLY due to efficiency losses at the higher Vin, and therefore shows up as waste heat. Which would be about 400 watts.
The Classic won't dissipate anywhere near 400 watts. Just remember that it won't be dissipating more that say, 50 to 75 Watts...
~Maybe~ 100 watts MAX (fer sure at 25 degrees C ambient)
Yeah, if all you do is raise the input voltage and keep the output current and voltage the same, it will get hotter because it heats up more....
The input power goes up and the output power stays the same so that means the efficiency goes down and the Classic gets hotter because it's
dissipating more Watts.
Let's say the efficiency is 95%... 58.4V at 70 Amps output is 4088 watts. 5% of that is 204.4... 204 Watts is RIGHT OUT ! (as Monty Python would say)
So the efficiency better NOT be 95% ! It's more like 98% or maybe 98.5% Still pretty hot though.
Quote from: Vic on June 04, 2011, 10:03:30 PM
Do not wanna waste too much of UR time. Guess am talking myself into running 92 volt nominal strings, as runs are short, and that still gives lotsa margin for EQ.
73, Thanks, Vic
No problem ! A bit lower input voltage with a short wire run rather than real high input voltage is probably better. You could certainly play with the different
scenarios and see what works better.
Say, you didn't put that front cover on the Classic that came with it, did you ? I'm finding out that a lot of customers are doing this when it's just not necessary.
That cover is for places like Hawaii where there is salty air. The cover restricts the cooling.
boB
Thanks boB,
Nope would not think of strangleing the ventilation of the Classic. BTW, I do like you guy's approach of knocked-out knockouts, with filter screens provided. AND, to me, the MX-60 you guys did about a decade ago is STILL a very capable box, ... amazin' And yet with that as an example some other companies forgot to even put in the Aux function of 'CC in Float', ... astonishing !
Keep up the fine work, boB. 73, Thanks .. Vic, K6IC
Hello,
A newbie here. I'm considering purchasing a 150 for a new off-grid system. Plan on about (12) 205W panels and operating at 24V.
I understand there is some controller efficiency benefit to keeping the input voltage low. However, doesn't putting panels in series improve the power extracted from the sun? I assume as the light fades, the voltage on each panel fades as well. So having two strings of six panels in series at 18V (max operating) each should output usable power longer than three strings of four or four strings of three; would that be correct? If so, how does that factor in toward maximizing the controller's efficiency and lifespan?
Thanks,
~Dan
Quote from: mmidnnite on June 27, 2011, 11:19:58 AM
I assume as the light fades, the voltage on each panel fades as well.
Not so, the voltage goes up first in the morning and amps follow, reverse at sundown, amps drop then voltage.
And when the voltage from the panels, presented to the batteries, is less than the batteries' there is no longer any charging occurring.
HTH
EJ
Hi EJ,
So are you saying that no time is gained (with sufficient voltage for battery charging) by increasing string length? Time seems to be the single largest loss on efficiency (-22% av, Ca) so if you could keep the voltage up for one more hour you'd gain over 4% right there. It sounds like you may be saying that will not happen and so the only "gain" will be more controller losses?
~Dan
That would be correct. The PV will give a steady voltage when not shaded the amperage is what changes.
Hi Dan, I am probably not doing a good job at explaining this, but here goes... In the early
am you would see 1 V and no amps, then V rises and there are still no amps, once V rises above the batteries voltage there is potential to charge, but the amps are also important , voltage will continue to rise as will amps but they only reach max amps when the sun is higher probably ~ 10 AM for you, maybe sooner. All the while your CC starts 'looking' for the best power point so as to maximize the WATTS that are going into your battery.
The early input is miniscule (tenths of amps) so 'time' is less of a factor during the shoulder producing parts of the day. This is where some of the less reputable sales pitches will say "able to produce power at night or??"
putting more panels in series ups the string voltage and thus reduces line loss. Also allows smaller size wire to be used and cuts copper costs.
Watts and watt hours ( or Kwh) is the real story not just voltage.
HTH
EJ
ps I am leaving CC losses out at this point
Very well, I appreciate that explanation. I guess what I'm really trying to do before actually assembling the system is to determine the balance between controller loss with higher input voltage (longer strings) and shorter strings (with less controller losses). Was wondering if the folks at MidNite, or anybody else, had anything to say about it. Maybe the answer will be in testing 3x4 vs 4x3 myself to see what the difference is, but I was hoping there would be some consensus by people in the know what would work best considering all losses: wire, voltage, controller, etc. Thanks for your input.
~Dan
0 panels, 0 controller, 0 batteries, 1 refrigerator.
Hoping for 12 panels, Classic 150, 8 Trojan-105's on a 24V system.
RE The curves,
I really love the folks at MndNite, ...
Am just now ready to pull the trigger on a new installation, in the next couple of days. Am still mulling the string configuraton. Efficiency Curves would be very helpful.
I'd bet that boB, or someone in engineering knows what this data is right now, they just have to. But guess that this is subject to improvement, so there may be a bit of reluctance in releasing it, when it could well be a bit better later. For me, tho, I guess that I will have to assume that that the ratio of currents available for a given battery voltage between the CL 150 and 200, is the dirrerence in efficiencies, and therefore, will need to run eight strings of three 30.8 Vmp panels, vs six strings of four, (92.4 Vmp vs 123.2 V) or something even higher voltage ... just musing.
Happy Fourth to all, Vic
OK, then ...
Am still trying to study CC efficency tradeoffs between available MPPT CCs. Between manufacturers and within product lines from any one manufacturer, like MidNite.
And, of course DATA helps in doing the tradeoffs well.
Perhaps I am asking too much, in thinking about CURVES ...
Vic
Vic, we have not had the time to make efficiency curves for the Classic versions. Having said this, the electrical efficiency of the Classic 150 compared to the larger controllers, such as the MX60 is not quite as good. The MX has huge inductors and more FETs.
I did do a measurement recently on the Classic with a 12V battery and a 72 Volt array (power supply). The efficiency came out 1% worse than the MX60. Something like 92% for the Classic Vs. 93% for the MX60.
As for daily harvest in kW-Hours go, we have seen the Classic out perform all other controllers except maybe the Morningstar MPPT 60 CC where we were on par. Almost no difference and some days even more.
That outperformance has to do with the tracking ability of the Classic. Even though its electrical efficiency may not be as good, the tracking efficiency more than makes up for it comparatively.
We hope to make curves for the other Classics at different battery and array voltages and do it "accurately" but we have so much work to do here right now that it will have to wait just a bit longer.
I do not want to publish any efficiency curves, even if I had them, unless they are done correctly and accurately. This is not easy to do. I do trust that 92% number though, even though I don't like it. However, lower efficiency is really to be expected at low battery output voltages because the losses are so much more a part of the output power that it just isn't as good as when you are measuring efficiency at higher battery voltages and higher output power.
boB
OK, Thanks boB,
Appreciate the data on the difference between the MX-60 and the Classic.
This helps give me the confidence to move foreward with my new installation, as I TRUST you folks.
The current plan is to use the existing Classic 150 with 12 ea SW-245s -- four strings of three in series (Vmp = 92.4 nominal), and an MX-60 with its own "identical" array of 12 in the same configuration. Later will probably change out the MX for an additional CL150. This generational difference may not be ideal with a common bank, but will see how things go.
73 Thanks again, SRI for being needy. Vic