Request for efficiency curves for Classic

[djsharpe]

Are there curves available (as with Outback) for Classic so configuration efficiencies can be compared?

[Robin]

We do not have efficiency curves. They are very difficult to do correctly. It is not possible to compare. Here are a couple of issues. When we did efficiency plots at OutBack 10 years ago, we had to use two current shunts and four meters. Then we took a picture of the meters to make sure all voltage and currents were done at the same time. The measurement turned out to be 99.4%. That measurement of course was not accurate. We have some pretty sophisticated power measurement equipment at Midnite, but each unit goes up to 20 amps only. We have restled with how to be able to accurately measure high currents, but have never had good enough success. Remember we are shooting for extremely accurate data. That isn;t the big problem though. If all manufacturers used the same measurement technique, you still could not adequately compare data. The power measurement only gives you data on the power train. More important is how the firmware tracks the sun. You can have the best power electronics on the planet and still have the worst controller. Believe me, they are not all created equal. We have rum side by side tests using identical arrays. Suffice it to say we won. Morningstar was second. OutBack third. Xantrex was far behind. We did not test Apollo or Blue Sky.

[Vic]

Hi dj,  Robin,

Think that dj is looking for the efficiency data,  comparing one model MN Classic vs string Vin,  and perhaps comparing this data to another Voltage-Rated model of Classic.

The OP,  and I had a Forum conversation on another Forum.

I was trying to make the point that running a string Vmp of 122-ish volts,  IIRC,  in a moderate tropical climate would reduce the efficiency of any CC that I was familiar with in a significant way.  ANd if the STC Vmp was 122 V,  if the system went to Float,  with nominal inverter loads,  that the Vin might well be in the mid 130's,  and this plus the tropical climate could take a toll on most CCs.  Raealize that the Vmp with hot PV is lower than STC,  and so on.  But warm climates might also mean the ambient  temps for the electronics is also increased.

It is my personal opinion that the differing max output currents from these charts:

http://www.midnitesolar.com/images/classicFrontPage/graphs.php

Tell much of the efficiency differential,  which is stricking to me.

Whatta I know ??  GL,  Vic

[LostLoon]

Robin,

Can you share the data (graphs, etc.) from your side by side CC testing? Perhaps it is on the website and I just missed it? I am always pushing the Classics and this data could help justify the extra expense to those customers not already convinced by the praise I heap on...

[Vic]

AND,

This topic has come up on this Forum before.

There are several purposes for having efficency data.  To me,  the most useful is to aid in making the tradeoffs between string voltages on a given voltage-rated model of the Classic CC.  And,  there is the desire to comapre  different CCs in the market.

I still do get hung up in trying to make sense of the Power Charts on the MN site.  The current available is reduced as the string voltage increases,  BUT,  have been told that this reduction cannot be used to calc the efficency,  and the reduction in current is fairly large.  The Classic probably cannot rid itself of as much power as the reduction in current represents.

This is an earlier Efficency Curve thread:
http://midnitesolar.com/smf_forum/index.php?topic=239.0

This does not really matter to me at this point,  as will be running string Vmp of about 93 volts. FWIW, Vic

[Robin]

I will ask the techs about doing efficiency graphs once again. As I have stated, it is not easy. It would certainly not be useable to compare against anyone else. You would be able to compare one classic against another though. That would be useful. I agree that it is sometimes hard to pick the best Classic for the application because there is a lot of overlap. We have not kept the data from the other controllers. This scoming summer, maybe we can set up identical arrays on the roof of our plant and put it on the web. Sounds like a big job though. As far as justifying the additional expense of the Classic..... you do need to include a DC-GFP on most others. We will be introducing a new set of Classics in another month or so. I think you will be pleasantly surprised at the prices.

[Vic]

Robin,

Thanks for considering the curves.  THis should help some of us make the tradeoffs in string voltage,  and Classic model.

Regarding the pricing of the  Calssic:  to me,  the Classic is a great building block.  It is not static.  Many other manufacturers of CCs seem to get something shipping,  and then only do bug fixes (and some other CCs seem to have serious BUGS,  (like the Xantrex SCC-150!).

On the other hand,  the Classic seems to be a canvas,  that the engineers at Midnite continue to paint great new functions onto.  And believe that this will continue for some time to come.  There is just no other manufacturer of MPPT CCs that I know of that has this commitment to product function extension.

AND,  the Classic is just a small amount of $ more than the others,  that,  IMHO are seriously lacking in capability.

Whatta I know?  ..  But I sure love my Claccic 150.  Will need more of them this year.

Thanks Midnite for all the great products and support.  All the best for the new year.  Vic

[Robin]

Here is a chart that was made from data taken off of the APRS website. They have some sophisticated monitoring equipment hooked up to Ryan's three identical arrays.
The results tell you more than an efficiency curve can ever hope to. This chart took one week of data from all three arrays and spit it out. The thing that matters is how much energy is put into the battery bank. This chart pretty well sums it up.

[attachment deleted by admin]

[niel]

of course the classic won.

why is it the outback and xantrex split outing each other on differing days?

[Kent0]

While this data confirms what we'd expect of the Classic, it's for only one test condition and we don't even have the information about which condition.

[Robin]

Kent, the snap shot taken from the APRS data was just a generic week. They all look this way. Each PV array is 1024 watts I believe. No shading. I will see if I have a wiring diagram or some other data. I do know that Ryan has a 48V battery bank. I do not know the PV array configuration, but I suspect this is just a good indication that the Classic is just a better charge controller. There was nothing special done to give the Classic an edge.

[Halfcrazy]

Each array was 6 solar world 165 watt panels. Wired 2 in series 3 in parallel. These run through the APRS data logger (Calibrated) and then on to there respective controllers. We did verify the output of each array and APRS considered that. The battery is a 48 volt bank of 16 Trojan L16 lead acids. We also had diversion loads set up so all 3 controllers would stay in Bulk. I can say the XW always tracked correctly it just seems to be less efficient.  The FM seems to get "Confused" a lot and spend a lot of time where it should not be?

The XW and Classic almost always agreed to the volt on VMP where as the FM would vary wildly. One thing this test tought me is Controller efficiency does not mean Jack you can have a super efficient controller and have a poor tracking algorithm and you will not harvest as many KWH per day. We spent a lot of time on the algorithms in the Classic and I feel this truly paid off with a controller that is capable of out harvesting the competition.

If the industry wanted to come up with a number I would suggest something that calculates it based on daily harvest not electronics efficiency. It is sort of like comparing Flywheel horse power to Rear wheel horse power. All that really matters is the horse power at the pavement.

Ryan

[Kent0]

Ryan,

Your numbers don't work in new math;) 2 x 3 = 6 not 9. If it is actually 9 modules wired 3 x 3 then the nominal max power voltage is 106 volts. If it is six modules they could be 2 in series for 71 volts or three in series for 106 volts.

Since the arrays are nominally the same but not necessarily exactly the same, a good test procedure would also rotate the charge controller and array combinations to make sure that differences in the arrays aren't 't partially responsible for the differences seen in the final results. That may have been done. The small amount of information provided just isn't enough.

Even if "harvest value" comparisons are the best way to compare charge controllers, actual efficiency numbers have more value than just comparing charge controllers and they are certainly part of the equation for comparing charge controllers. For instance: to evaluate a 2 series x 3 parallel array vs a 3 series x 2 parallel array one needs to consider both the wire losses to the PV array and the charge controller losses (or at least how the charge controller losses change with input voltage). 

[Halfcrazy]

Good catch I have 6 panels in each array wired 2 by 3. The 3 arrays where rotated regularly and also calibrated by APRS world.

[Vic]

Thanks HC and Robin for the additional info.  Nice job on the Classic programming.

More is better,  but agree that the absolute efficency is not very important,  and harvest is 95% of the ball game to me.

Thanks,  keep up the great work.  We always want more ...  Vic