question

trimard avatar image
trimard asked

mppt voltage fluctuating under low light

Hi,


I have following setup:

2x 130W solar panels in series.

1x MPPT 75/15 connected to a 180Ah battery.


As I connected the panels yesterday (very cloudy day) I noticed that the voltage reading ("solar voltage" in Victron App) fluctuates very much. I had readings going from 20V to 36V in just a second. It keeps dancing very much. The sky was cloudy, but there was no visible fluctuation in light intensity. The open voltage of the panels stays steady at 36V. The charging output of the charger was also stable on 13,7V.

I was wondering whether this is normal behavior or not.. Could it be that the mppt has difficulties tracking the MPP under low light conditions? I did not have the ability to check the behavior under 'normal', say, bright sky light conditions.


Thanks for the advise!



MPPT Controllers
2 |3000
Toggle Comment visibility. Current Visibility: Viewable by all users

Up to 8 attachments (including images) can be used with a maximum of 190.8 MiB each and 286.6 MiB total.

8 Answers
wkirby avatar image
wkirby answered ·

Indeed, it is the MPPT performing a "sweep" in order to figure out the Maximum Power Point. The Voltage swing will be more pronounced in cloudy weather because even a small load can cause the PV Voltage to fall lower than it would in sunny weather.

2 comments
2 |3000
Toggle Comment visibility. Current Visibility: Viewable by all users

Up to 8 attachments (including images) can be used with a maximum of 190.8 MiB each and 286.6 MiB total.

trimard avatar image trimard commented ·

Thank you for you answer @WKirby. It is partially reassuring. ;-)

Am I correct to think that this sweeping should stabilize at a certain point? And that it will find a new MPP after a certain period of time by restarting this sweep cycle? Or is this sweeping more of a constant procedure?

Thanks!

0 Likes 0 ·
pmano avatar image pmano commented ·

"even a small load can cause the PV Voltage to fall lower than it would in sunny weather. "

What load are you referring to? The MPPT to Battery ?

0 Likes 0 ·
spirou avatar image
spirou answered ·

I'm having a very similar "problem" with a 2x160W panels in series on 100/30 charging 2x90Ah batteries. I can understand the sweep but the batteries are not getting recharged much. It's a regular overcast day and I'm only getting 5-16W out of this setup at the moment during bulk phase. For the past three days the 100/30 didn't even get out of bulk mode, topping out at just over 13V (current max for today is 14V because I plugged into 230V). On the old van I had a 120W panel on 75/15 & 130Ah battery that had absolutely no problem keeping up. I'd see 20W even on the worst day, now this one is telling me it's unable to charge the battery with double the Wp on anything but a nice sunny day or what?! With the control panel turned on and water heater on anti freeze mode I'm losing 4-5Ah per night and this 320Wp solar setup is unable to replace this "standby" load. How is this possible?



screenshot-20191114-094710.png (73.4 KiB)
screenshot-20191114-100444.png (51.8 KiB)
1 comment
2 |3000
Toggle Comment visibility. Current Visibility: Viewable by all users

Up to 8 attachments (including images) can be used with a maximum of 190.8 MiB each and 286.6 MiB total.

Guy Stewart (Victron Community Manager) avatar image Guy Stewart (Victron Community Manager) ♦♦ commented ·

That voltage swing is too much, I would first look and test all the solar PV wiring. There might be a bad connection or fuse.

0 Likes 0 ·
spirou avatar image
spirou answered ·

Went to the van again, less fluctuation but still underwhelming performance for overcast near mid day. Screenshot was taken before I checked nearly all connections, I'd have to take the panels off to reach the remaining. I've also checked with a multimeter before/after fuse (regulator input side) and there's no difference, "wild" fluctuation. If it is a bad connection it then has to be on the roof although I can't figure out what could have happened as the van was parked since Sunday evening and looking at history it still managed to perform OK on Monday under identical conditions as today.



screenshot-20191114-104826.png (175.9 KiB)
screenshot-20191114-110649.png (69.3 KiB)
1 comment
2 |3000
Toggle Comment visibility. Current Visibility: Viewable by all users

Up to 8 attachments (including images) can be used with a maximum of 190.8 MiB each and 286.6 MiB total.

JohnC avatar image JohnC ♦ commented ·

It's not really such a 'wild' fluctuation. If you multiply out V and A, you'll see W at a fairly consistent ~15W. The sampling times mightn't be exactly the same, so might explain any variation.

Your first graph, earlier in the day, can't even manage that 15W, and the mppt is likely cutting out completely when it can't maintain a sufficient V. That's wilder.

Like Guy, I suspect the panels. Could be anything, even down to diodes in the JBox(s).


0 Likes 0 ·
spirou avatar image
spirou answered ·

I have confirmed it has something to do with connections on the roof, but for the life of me I can't figure out where exactly. Everything seems nice and tight but the live data chart went smooth the moment I detached one panel from the brackets and handled it around a bit. It swung again a few times as I was moving it around but under no condition when I handled and tested (pulled, torqued, twisted) any of the cables/connectors.

Anyone have ideas for troubleshooting MC4 connectors or even a better solution? As far as I can tell Mc4 can't be taken apart without destroying them (as I've figured out when installing the first one).

14 comments
2 |3000
Toggle Comment visibility. Current Visibility: Viewable by all users

Up to 8 attachments (including images) can be used with a maximum of 190.8 MiB each and 286.6 MiB total.

Guy Stewart (Victron Community Manager) avatar image Guy Stewart (Victron Community Manager) ♦♦ commented ·

Congrats on isolating the issue to that one panel.

I suggest you get some new MC4 plugs and sockets (though most likely it is the connection that was made on the cable, not the one supplied from the PV factory).

Cut it off and re-crimp. Do one at time and see if it is fixed as you go.

MC4's require a special crimping tool to be done properly, If you don't have one, you must pay very special attention to the part where you are clamping down on the folded back solar wire to make sure there is a good connection. There are lots of videos on youtube about that.

A weak or insufficient connection, can be different to loose connection. If there isn't enough copper contact, even though everything is tight, when the current increases, the heat increases and the resistance increases and the voltage drops and you get the effect you are seeing in the graph you posted earlier.

If you are going to replace both MC4's, before re-crimping the MC4's, while testing, it would be a good idea to also make a dumb twisted bare connection, and see if things are woking 100% then. There are also diodes and junction box connections in the panel that could be at fault, though less likely.

0 Likes 0 ·
spirou avatar image spirou Guy Stewart (Victron Community Manager) ♦♦ commented ·

I give up. Yesterday I replaced the MC4s (two were indeed crimped loose enough to pull out the cable) and no difference. Today I took the panels off entirely and wired them the "dumb" way, direct wire to wire contacts and no appreciable difference. I tested each panel individually and together in series, both yesterday on the van and today...you guessed it, no difference whatsoever. I opened up the panel junction box for any signs of loose connections or poor soldering and found nothing odd. The 2 diodes also seem to function normally.

To clarify, all my tests were done on uniform grey overcast days so any solar activity would not be much of a factor at such time scale. And to repeat the setup of todays test: 2x 160W poly panels in series (or solo) via a mix of 6 and 10mm2 wires to a smartsolar 100/30 feeding 2x Varta LFD90 lead batteries.

Attached are charts from the series wired test today, start, end and entire time span between 9 and 9.30AM. Battery voltage at start was 12.4V.

If I hadn't experienced what a single 120W poly panel on a 75/15 can output in similar conditions I might be thinking this is normal. But it just can't be. Although for the life of me I can't figure out what the issue is. I'm having trouble believing both panels are faulty and the seller claims there have been no complaints before.

To add, it swings less with more sun but not much.

Any smart ideas on how to proceed are welcome.

0 Likes 0 ·
screenshot-20191124-092838.png (301.2 KiB)
screenshot-20191124-092838.png (301.2 KiB)
screenshot-20191124-092942.png (209.8 KiB)
screenshot-20191124-092722.png (178.9 KiB)
Daniël Boekel (Victron Energy Staff) avatar image Daniël Boekel (Victron Energy Staff) ♦♦ spirou commented ·

Hi @spirou

Those graphs indicate a very bad connection somewhere, might be in the MPPT itself.

To find out: short the cables coming from the panels (without MPPT) and measure the current flowing through the cables. if there isn't much flowing, wiring or panels are bade, if normal current flows, the problem could be the MPPT, try another MPPT to confirm.

0 Likes 0 ·
spirou avatar image spirou Daniël Boekel (Victron Energy Staff) ♦♦ commented ·

Just tested both panels for short circuit current and voltage (same time as yesterday, same overcast conditions). Both panels output constant 0,2-0.21A @ 19.9-20V Seems quite poor but consistent with the max values I got yesterday. I guess the next step is to find a second mppt to try it out.

0 Likes 0 ·
klim8skeptic avatar image klim8skeptic ♦ spirou commented ·

How does your panel produce any volts when shorted?

0 Likes 0 ·
Daniël Boekel (Victron Energy Staff) avatar image Daniël Boekel (Victron Energy Staff) ♦♦ spirou commented ·

@spirou

I assume you measure the current when shorted, and the 20V is open circuit.

Your panels produce 3-4 watt like this...and this could very well be too little to start the MPPT, giving the results you've seen. Something seems to be wrong with your panels.

Please try a different panel, even on a clouded sky a panel should produce 10-50% of it's Wp

0 Likes 0 ·
spirou avatar image spirou Daniël Boekel (Victron Energy Staff) ♦♦ commented ·

Sorry for ambiguity, obviously voltage was measured OC.

0 Likes 0 ·
Rob Fijn avatar image Rob Fijn spirou commented ·

Hello, you can also try with just one panel to the MPPS for a test.

Just to find out if it is only one you the two panels you have.

Rob

0 Likes 0 ·
spirou avatar image spirou Rob Fijn commented ·

I did test each panel individually and there is no difference in the shape of the charts above. If I had another solar regulator handy I would have tried that already.

For this last round of testing I'm 100% certain the connections between the panels and mppt are solid and not the reason for the odd behaviour. So it either has to be an issue with the panels or mppt.

I guess the next step is to find another regulator to eliminate that variable.

0 Likes 0 ·
spirou avatar image spirou spirou commented ·

Performed two more tests today (again overcast day)

Single 160W poly panel @12 = 0.22A Isc

Single 160W poly panel @14.30 = 0.11A Isc

100W mono panel @14.30= 0.06A Isc

60W mono panel @14.30= 0.05A Isc


Voltages were around 19-20V for all. Unfortunately there just is no sun in the forecast to see how it behaves with more light. But I guess the conclusion is that there is no conclusion. Looks like I just have to live with 2% of nominal Wp during winter. And I guess my previous setup was abnormally efficient to have produced at least 10% under similar conditions.

Very underwhelming and dissapointing to say the least

0 Likes 0 ·
Phil Gavin avatar image Phil Gavin commented ·

Get a clip-on ammeter, the current will be constantly fluctuating with the sunlight, but you will be able to take relative measurements while you manipulate MC4 connections.

This may just point you to the problem quickly.

0 Likes 0 ·
JohnC avatar image JohnC ♦ commented ·

Detaching the panel frames from their brackets should have no effect. If your brackets are earthed, then it's possibly a panel fault. The panel frames shouldn't have continuity between either + or - power wires.

MC4's can be disconnected but not rewired (they self-destruct). To disconnect you can get a special tool, but often are useless junk. A couple of v small screwdrivers to hold the two little tabs down, a wiggle & a pull, job done.

0 Likes 0 ·
Daniël Boekel (Victron Energy Staff) avatar image Daniël Boekel (Victron Energy Staff) ♦♦ JohnC ♦ commented ·

I find needle nose pliers to work best, you push both clips at once and the connectors pull apart very easy.

I'd do a visual inspection first, if the conductive parts are shiny and clean, it's probably a good connection.

Often seen faults:

-contact not pushed far enough in connector, making for a bad connection

-water ingress due to bad connector mounting.

0 Likes 0 ·
Matias Käkelä avatar image Matias Käkelä commented ·

@trimard did you ever resolve these issues? I’m having similar Symptoms: sunny weather yields over 500w at best, though typically closer to 400w (I have 2x285w panels). However, cloudy days give only 5-10w (today’s peak is around 25w), and the pv voltage is jumping between 45-65V constantly.

As I’m new to this domain, it’s hard to know if this is normal or of there is something wrong.

0 Likes 0 ·
flodis avatar image
flodis answered ·

I have noticed at solar currents below 0.5A it has a hard time starting the charging. It loads down the panels for a short while and indéed reaches an optimum current. But, it does not recognise this as a success.

It returns to non load and starts figuring agan.

If cells are shortcircited for a few seconds the algorithm picks up, but after half a minute it drops out of perfect charging and returns to some attempts again.

I think it may be related to low resolution of the current sensing ADC.

More attention to low light and software could improve.

2 |3000
Toggle Comment visibility. Current Visibility: Viewable by all users

Up to 8 attachments (including images) can be used with a maximum of 190.8 MiB each and 286.6 MiB total.

flodis avatar image
flodis answered ·

Here is an example of 0.4A charging not sustained.


20201219-200934.jpg

The battery is a rock solid 40Ah LiFePO4 that was discharged to 12V. The panels in this case are 2X60W and 2X25V panels and wiring is excellent. When directly connected to the battery they produced as steady 0.4A. When the current capability of the cells reaches appr 0.6A and above the charging is continuous and works as expected.

I have tried to leave the charger connected over winter, but as it does not do well in low light it can not even maintain energy to power the 25 mA to the regulator and slowly drains the battery.

This problem can also be emulated using a power supply set to 17-18V with a non fallback current limit of 0.4A.

With all the information the software has, it is frustrating to see it fumbling.


20201219-200934.jpg (368.1 KiB)
1 comment
2 |3000
Toggle Comment visibility. Current Visibility: Viewable by all users

Up to 8 attachments (including images) can be used with a maximum of 190.8 MiB each and 286.6 MiB total.

klim8skeptic avatar image klim8skeptic ♦ commented ·

The Vmp and Voc of the panel look very low. Please confirm panel specifications.

0 Likes 0 ·
Marcin Mocydlarz avatar image
Marcin Mocydlarz answered ·

Hi, I have similar problem. Easysolar II GX 3000VA - there is mppt 250/70 inside.

During sunny day, when solar power increase up to 1000W or more then power/voltage/current is fluctuating - drops around 20%, it looks like power is dropped and then increase again.

When i have limited charging current to 20A then fluctuating stop. Is there any reason that happends?

1635178803681.png

other mppt 100/20 works fine, power is stable +/- 2-4W.

firmware 1.59

thank you for the advice.


1635178803681.png (35.4 KiB)
2 |3000
Toggle Comment visibility. Current Visibility: Viewable by all users

Up to 8 attachments (including images) can be used with a maximum of 190.8 MiB each and 286.6 MiB total.

Marcin Mocydlarz avatar image
Marcin Mocydlarz answered ·

I have just connected another mppt250/70 to the same string. Problem disappeared...

It looks there is problem with MPPT inside Easysolar II GX :(


2 |3000
Toggle Comment visibility. Current Visibility: Viewable by all users

Up to 8 attachments (including images) can be used with a maximum of 190.8 MiB each and 286.6 MiB total.