New Solar Panel Test -

[Dunk_c]

Hi

I just received a "60w" panel and tested at ambient temperature before the glass got hot, in full sun, no clouds, 11:35 am, Qld.

The product label states 21.6v VOC and 3.77A ISC, Vmp 17.7, Imp 3.39, 60w.

When I tested the panel the VOC is 23.2V and the ISC is only 2.1A.

When the panel is attached to a 10ohm 200w power resistor the panel only delivers 1.8A at 18v, or 32 watts, when cool at midday. I should have measured the solar irradiance but forgot.

These panels are made from 125x125 Sunpower Maxeon cells cut into 1/3 (125x41); 36 joined together in a 36 cell string.

The spec for the Sunpower Maxeon 125x125 cell is 6.28A Isc. Dividing this by 3 = 2.1A. This is the max SC current produced by the new panel.

According to some specs I found online, the full 125x125 cells theoretically produce 0.687v Voc. Multiply that by 36 cell in the panel and you get 24.7V, similar to what I measured 23.2V, plus some drop across the diode.

The label on the panel and the rated output cannot be correct. Seems like half the amps are missing!

Comments wanted before I question the seller.

Thanks

[HackedFridgeMagnet]

As long as its not cloudy you don't need to worry about irradiance. It will be close enough.

Vmp 17.7 / Imp 3.39 = 5.22 ohms. This load will give closest to maximum power.

Your ISC should match the specced ISC.  Maybe half the panel isn't connected?

Maybe you had a shadow on the panel or maybe your panel wasn't perpendicular to the sun or maybe your panel just not as good as it says.


I can't think of another reason.

[Ian.M]

Do an A/B test with a known good panel side by side with the suspect one, and video your test procedure, so (assuming the good panel meets specs,) the seller cant claim "It must have been hazy when you tested it". 

[HackedFridgeMagnet]

These panels are made from 125x125 Sunpower Maxeon cells cut into 1/3 (125x41); 36 joined together in a 36 cell string.

The spec for the Sunpower Maxeon 125x125 cell is 6.28A Isc. Dividing this by 3 = 2.1A. This is the max SC current produced by the new panel.

Yes this sounds significant. I missed it before. Your calculation seems right and their spec is wrong.

[Dunk_c]

Definitely no clouds, clear blue skies, but will check again in case it was a little hazy.  Recorded max values as the panel was moved about.

Don't have a 5ohm resistor.

Clearly the Isc is only a little more that half the quoted value.

The panel voltage would suggest that alll 36 are connected.

Don't have a an identical second panel for confirming.

[Ian.M]

You don't need an identical panel - just one using a similar technology with trustworthy specs publicly accessible online. It wouldn't matter if its only a single cell of 10 sqcm area as long as you can show that under the same illumination conditions it delivers >90% of its rated output.   Apparently the Sunpower Maxeon cells are mono-crystalline silicon.   Alternatively use a calibrated Lux-meter.

Theres been some discussion of them here last year: https://www.eevblog.com/forum/chat/sunpower-maxeon-solar-cell-discussion/

[Dunk_c]

A standard panel would be a good idea!

Even without testing performance, the smaller panel size (about 2/3 the area) and smaller individual cells (about half the area), compared with other 60w panels I see on the web, tells me something is wrong.  The manufacturer has incorrectly labelled the panels and the seller is unwittingly selling them.

[Dunk_c]

More testing today and feedback from seller.

Measured irradiance at 1100 W/m^2.
Voc = 22.7V
Isc = 2.25A
Power = 51W

Panel loaded with 10ohm resistor
V = 16.6V
I = 1.7A
Power = 28W

Seller has said I can keep this one as they feel it is underperforming, and they will send another, if I choose.  Worst case scenario is that the second panel is the same as the first.

My question is,  should I return item for refund and find better panel (I need 60W), or cobble they two together?

The Imp for the panel on the label is 3.4A, looks like two paralleled panels would equal this.

I am using a PWM charge controller.

Thanks

[HackedFridgeMagnet]

More testing today and feedback from seller.

Measured irradiance at 1100 W/m^2.
Voc = 22.7V
Isc = 2.25A
Power = 51W

Panel loaded with 10ohm resistor
V = 16.6V
I = 1.7A
Power = 28W

Seller has said I can keep this one as they feel it is underperforming, and they will send another, if I choose.  Worst case scenario is that the second panel is the same as the first.

My question is,  should I return item for refund and find better panel (I need 60W), or cobble they two together?

The Imp for the panel on the label is 3.4A, looks like two paralleled panels would equal this.

I am using a PWM charge controller.

Thanks

Isc x Voc != Power (does not equal)

I think if you got 2 panels for the price of 1 you would be very slightly ahead power wise.
But if the most power you are going to get out of one panel is about 2.2A x 16V which is near 35W. (guesstimate from the figures you gave me).

It sort of depends on whether you are ok with the extra space the less efficient panels take up.

[Ian.M]

Is the build quality good enough for you to be happy with it as a pair of 35W panels? 

I don't think the extra area is that significant as the area of Sunpower Maxeon cells for a particular noon sun MPP wattage is essentially constant,  and for the same VOC, a 60W-70W panel would need 2/3 of 125x125mm cells, so would probably be implemented as two strings of 1/3 cells, so the only real loss of area is the frame and border.

[HackedFridgeMagnet]

I was thinking that the cells themselves may be the lower grade, lower efficiency ones. I think they sort them and the worst go on Ebay.

Not to say the OP wont get the power he is measuring but it will just take up a bit more space than higher efficiency ones.

Rough efficiency calc.

.125m x .125m x 1/3 x 36 1100W/m/m = 206.25W  (each cell cut into 3?)

rough efficiency = 35/206  = 17% not too bad.

[HackedFridgeMagnet]

that was a very rough calc as there hasn't been any mpp data given. and hardly at STP.

[Ian.M]

What I think happened is the OP just got a defective panel that half of the cells are not connected.
Using current generation technology, even the shittiest quality cells should not be just of 17% efficiency.

From the O.P.'s first post, he's getting 1/3 the full undiced cell current from 1/3 the cell area, with an O.C terminal voltage within 1.5V of the expected voltage for a 36 cell string, so there's no way it could have an O.C. string.   It might have one or two shorted cells but that would be easy to check by shading the cells one at a time.

[Dunk_c]

Thanks for the correction on the power calc. Best to keep it real😬

The build quality is fine and I already have some MC4 connector to parallel them. Going on roof so size jot a big problem, just more fixings.

Hadn't thought about low grade cells. They are advertised as grade A+ Sunpower Maxeon cells, the panel made by Sarconic.  There is a Sarconic.com.de web site that looks genuine.  I did email them with the serial number to see if they could verify the origin - all just an academic exercise really.  Paid the price for buying the cheapest I could find but had success previously on the Bay.

Total output current will be less that the 5A that my Genasun GV-5 can handle.

I'll let you know whether I keep or return. 

[Dunk_c]

Reality check - Anyone ever seen a 60w panel in a 550x440 frame?  Even without any testing this seems to be small and more like what I see for 40w panels. Also, some other panels in this wattage I have seen use a 36 string of 2/3 cut 125x125 cells (as far as I  ould measure from web pictures).

I have asked the seller if the replacement will meet the label performance. Waiting in reply.

[Someone]

Reality check - Anyone ever seen a 60w panel in a 550x440 frame?  Even without any testing this seems to be small and more like what I see for 40w panels. Also, some other panels in this wattage I have seen use a 36 string of 2/3 cut 125x125 cells (as far as I  ould measure from web pictures).

I have asked the seller if the replacement will meet the label performance. Waiting in reply.

This is the easy answer to your question! Other typical panels have less than 150W/m2, so their claim of 250W/m2 is very unlikely and only possible with bleeding edge single junction cells under ideal conditions. Its like battery capacities from grey market sellers, always suspicious.

[Dunk_c]

Seller gave the usual story about panels never meeting quoted specs but I accepted offer of second free panel.  I'll just bolt the two together.  Not worth the hassle of return and finding another. Cost will be less than $1/W.  I don't have any placement constraints.  There were more than sixty endorsments for the item,  99% positive - I wonder what they actually got!

I will report back perfromance of second panel.

[Seekonk]

Most don't really know what their panels are doing.  Many solar systems have excess potential and it is difficult to determine when some start to go bad.  I have an Arduino based solar system and dump excess power to heat water.   I have a routine in TOOLS that I can call to shut everything off for a moment and operate the PWM heaters to determine power point and power of array.  Had enough panels go bad to know you can't just set and forget.

[Dunk_c]

As expected the replacement panel performed similarly. 

[Dunk_c]

I tried to determine a rough MPP, this time tracking MPP with a variable dummy load

2:30 pm, clear sky, hot panel, perpendicular to the sun
W/m^2 = 1070 using uncalibrated solar irradiance meter
60w electronic CC dummy load

See chart, MPP = 32w, half the advertised rating (under my conditions admittedly)

I expect the second unit will behave identically.

So the two connected in parallel should yield 60w.

[HackedFridgeMagnet]

Thanks for the follow up.
So the cells are 1/3 or a 1/2 of a full wafer, or something else?

Then I guess you could calculate the efficiency.

[Dunk_c]

I measured the surface area of each cell and multiplied by 36 cells to give 180w/m^2.
I measured 1070w/m^2 solar irradiance.
So efficiency = 16.7 +/- error % efficiency.under hot sun, in the field conditions (not 25 C).

[IconicPCB]

Try a 5 ohm load in preference to the 10 ohm