The downside to in-series panels

[erik.calco]

I've been running my main system with 2 panels per string for a couple years now. One thing we learn as we're building a system in series vs parallel is that in-series has a bigger negative impact from shading.

But, today is the first time I discovered how big the impact was as I received a Bluetti AC200P today and hooked up 2x320W spare panels I had collecting dust. Monitored output as the sun moved, and over time, less and less shade. But, when one panel was completely sunny, and the other nearly completely shadow, the output was only 90W. Hmmm....

So, I disconnected the shady panel running just one, and the output immediately DOUBLED! I would of been better off running 1 panel until the sun cleared the 2nd panel.

Now that the sun is complete on both panels, put them back in series, and immediately shot up to 500W. The sun still is off to the side, so the shortage now is just due to sun angle.

 

[Crowz]

Do the panels have bypass diodes?

 

[Solar Guppy]

Series panels handle shading just fine, with one caveat, the remaining panels must be above the minimum voltage of the mppt controller AND is NOT in parallel with other strings.

Your issue is your doing both, running only two in series and in parallel, that is not how one should run series configurations of solar panels ( if there is shading )

 

[efficientPV]

It just might be a stupid controller. This is a common problem with many controllers software where they lock in on the first minimal power point. I have this problem and will be doing a work around. As was said, Bypass diodes would have given nearly the same power with a shaded panel in series. I bet if you disconnected the panel for 15 seconds it would find itself. When mine locks in at wrong voltage, a disconnect returns it to normal. Though, I have the opposite problem of locking in at the lower voltage. Most solar products are crap.

 

[erik.calco]

Do the panels have bypass diodes?

No clue. It's a black box, literally.

If you want to look it up, the panels are REC320TP2M

 

[erik.calco]

Series panels handle shading just fine, with one caveat, the remaining panels must be above the minimum voltage of the mppt controller AND is NOT in parallel with other strings.

Your issue is your doing both, running only two in series and in parallel, that is not how one should run series configurations of solar panels ( if there is shading )

How am I running in parallel? Only two panels in this test of the Bluetti.

 

[erik.calco]

It just might be a stupid controller. This is a common problem with many controllers software where they lock in on the first minimal power point. I have this problem and will be doing a work around. As was said, Bypass diodes would have given nearly the same power with a shaded panel in series. I bet if you disconnected the panel for 15 seconds it would find itself. When mine locks in at wrong voltage, a disconnect returns it to normal. Though, I have the opposite problem of locking in at the lower voltage. Most solar products are crap.

I'd love to hear @Will Prowse input on the Bluetti AC200P controller. Bluetti used to be his favorite mobile power station.

 

[erik.calco]

BTW, production was consistently over 500W in full sun. But, as soon as a tiny bit of shade came over a tiny portion of one panel, dropped to 320W.

 

[Will Prowse]

I'd love to hear @Will Prowse input on the Bluetti AC200P controller. Bluetti used to be his favorite mobile power station.

I've actually complained about multiple issues with them. And they have the worst customer support around. Good units when you get a good one.

The amount of times I've complained about them, I can't believe anyone would think they are my "favorite". Did I say that before?

The first models five years ago had zero issues. The new stuff has had tons of issues.

 

[Hedges]

An array 2s3p, a shaded panel probably means zero from that string.
Your 2s only, if one panel shaded it ought to work at well as 1s. If it doesn't, either bypass diodes aren't working or stupid MPPT is happy with 0.5A from a bit of light on the shaded panel, doesn't reduce voltage to find 10A from single panel.

It may be that with a longer string like 10s, shading one panel and dropping current to 0.5A raises voltage of remaining 9 panels and it is easier for MPPT to find the 9s solution even without doing sweeps.

I think (but haven't tested) that half-cut panels are more likely to confuse MPPT. Partial shading could drop current 10A --> 5A and MPPT is dumb and happy.

 

[Will Prowse]

Here's another:

I don't think I can be any more clearer than this

 

[pollenface]

Get a 2nd SCC to deal with the shaded panel.

I have 4x arrays with their own SCC and have shading issues, never lacking power as each array takes a turn at making the power.

 

[erik.calco]

I've actually complained about multiple issues with them. And they have the worst customer support around. Good units when you get a good one.

The amount of times I've complained about them, I can't believe anyone would think they are my "favorite". Did I say that before?

The first models five years ago had zero issues. The new stuff has had tons of issues.

LOL Yes, you said that before, back in the RV days. Don't worry, I know you're 100% objective. Love your integrity!

 

[Solar Guppy]

How am I running in parallel? Only two panels in this test of the Bluetti.

Sorry about that, I re-read the posts and it would seem your charge controller is likely not smart to deal with shading would be my best guess

 

[erik.calco]

Sorry about that, I re-read the posts and it would seem your charge controller is likely not smart to deal with shading would be my best guess

I should get some Y cables and test 2P, though. Only thing is the MC4 cables that came with the Bluetti seem kinda thin to me. But, it is rated for 700W, so who knows.

 

[erik.calco]

I watched the sun come over my main system's panels today, and it did not have this problem when one had full sun and the next panel in the string had mostly shade. I wanted to monitor it as the sun moved over the first panel, but had to reboot the PC with the solar software cuz a data dump yesterday left it in a mode where I wasn't seeing updates. But, overall, this confirms that this other MPPT does not have this partial shading issue.

With the Bluetti, I only connected 1 panel to start. And, even it had horrible readings with 1 panel only getting 90% sun. Would jump between 0 and 50W, so that 1/3 of the time it read 0W, the other two near 50W. Now that it has about 97% sun, finally hitting 100W and no longer hitting 0W.

Overall, the evidence is pointing towards the overall consensus of contributors to this discussion... looks like the Bluetti MPPT does not handle partial shade well at all. It does, however, perform well with full sun. I just have a lot of trees impacting the morning and afternoon.

Note that this is not a big deal on the main system because the time it goes from full sun to full shade is short, and when it is full shade, well, can't expect miraculous production lol. Likewise, not a big deal with the Bluetti. This is primarily useful for helping to plan and design better systems in the future, and learning how MPPT quality can differ, and for overall understanding the Bluetti.

UPDATE: I just added the 2nd panel while it has 50% sun, and the first has 100% sun. The power dropped from 250W to 160W (consistent with shading issue), but this 160W is much better than yesterday, evidence that reconnecting it can "reset" the tracking to get better results. This is the sort of thing that a firmware update could help improve. Now I have to look for open source MPPT.

And I found one...

https://www.reddit.com/r/solarracing/comments/ss6nm9

 

[Consumerbot3418]

the output was only 90W. Hmmm....

So, I disconnected the shady panel running just one, and the output immediately DOUBLED!

but this 160W is much better than yesterday, evidence that reconnecting it can "reset" the tracking to get better results.

I'd be curious to know what might have happened if you'd waited 10 minutes or so...

I guess some/most MPPT algorithms may be proprietary, but I can see pretty clearly that my Victron SmartSolar 100/50 does a full sweep every 10 minutes. At least, I presume that's what's happening at precise 10 minute intervals here:



My Midnite Classic does it every 3 minutes (and I believe also has an option to perform a sweep manually):


I'm guessing that in between sweeps, they venture ± a certain number or percentage of Volts, looking for increased yield, but might not range far enough from the previously determined peak to compensate for a completely shaded panel. If it is by percentage, @Hedges above speculation about longer strings would make perfect sense.

While some MPPTs are better than others, none of them are going to be able to give stellar results with partial shading. 10% shaded does NOT mean you can expect 10% decreased output--depending on the internal wiring of the panel, and how the shade falls on the panel, the losses will be much higher.

I'm thinking about adding some cheap used panels in shaded conditions, but substantially overpaneled, and I'm interested in how to optimize the string config. But I think if I was experimenting as you did, I'd want to give the MPPT controller time to run that full sweep.

 

[Will Prowse]

You do recommend 2 models on your ultra portable solar generators page?

Yeah after the software issues were sorted they are good and have features EcoFlow don't have. Did you watch those videos

 

[Will Prowse]

I posted a lot of videos about this, they are all there.

 

[Hedges]

I just watched this one after searching Tesla on your YT channel. At 4:12 you mention that you can make a system to charge 2 cars for "very cheap"? I guess I have a different definition of "very cheap"?

Less than the cost of two Teslas.

For a guy who wrote the book on DIY solar for beginners, I expected a bit more patience with your replies? I guess I need to spend a year or two watching every video on your channel? I'll be dead before I get through all of them. I just wanted to figure out how to boondock with my pop up and my ModelY? I'll go fuck myself now, thank you.

Will has something like hundreds of thousands of viewers, many thousands of subscribers to DIY Solar.
Kind of like a celebrity with fans. He might sign a few autographs, but can't provide personalized support.
That's where the forum community comes in.

Oh, you have one Tesla. If you can charge it for a fraction of PG&E, that is "very cheap".

Portable boondocking, not so convenient to build.
I would suggest level-1 charging (1500W), 3000W of panels laid out on the ground feeding a batteryless inverter. That's 6 to 10 panels, could fit on one of those little trailers. You'll need space to spread the out without shade.

What you want is a PV direct charger that works at variable rate, so 100% of available power is used. I think there are variable rate chargers but don't have the details. Maybe someone else can provide info.

Me, I don't do videos. I want to read a document, follow steps, not have it pace me on my work or learning.

If you do have the capability to take care of your own needs that's great! (This is a DIY forum, after all.) Most of us can't reach ...