Hi. We've got another solar power video and this one's rather interesting. I'll link in the video I did here if you haven't seen it. For my uh, I expanded my home solar power uh system.

I had an existing three kilowatt system and I actually, uh, this is it here on this Uh roof here and I actually moved this to the other side. This is on the western Uh side roof and I moved it to uh, the eastern side roof over the other side because I had to make room for a new five kilowatt um, en phase system with micro inverters and new uh Lg 370 watt panels. I put 14 in them. so I've got a total eight K nominal eight kilowatt uh system now.

I kept uh, the original panels of course and uh, all the racking and everything else and I kept the same Sunny Boy Tl 3000-21 inverter three kilowatt inverter that I've got that's over on the western side roof, right? I like the shaded over on this side, so I kept all that. All the wiring was the same so I just reused the existing system. Why not? It's not going to be as efficient on the western side roof of course, but it'll get the morning sun which adds to my total output. So it's still a useful system.

So I kept it, moved it over, That's all I did. and then I it got like a lower output as I expected, but then something interesting happened. Now, I've actually got three ways that I analyze my solar power system because I've got a hodgepodge of two different systems. I've got the Sunny Boy string inverter We're looking at uh, today, the three Kilowatt twelve panel 200 times 250 watts.

Uh, three kilowatt system that still uses the Sunnyboy inverter, which then I feed the data into Pv Output.org And I've been feeding this data into here for uh, seven years now. So I'm still keeping that as a separate monitored system. But of course, I've got the new end phase uh system which is, Um, the micro inverter. Uh, system? Here it is here.

Here's the live data now and this is actually a photo of my roof. So here's the 14 new panels. Um, and these all have n phase micro inverters on them so I can get the data and the graphs for each individual panel like this. And the good thing about the micro inverters of course is that if one of them gets, uh, shaded or it fails or something, you know a bird poops on it and it you know the bypass diodes kick in and it's like it's gone ski.

It won't bring down the entire string whereas my, um, older three kilowatt system is a string inverter system. So if one panel goes down, it will take the whole Cis. It'll degrade the entire system. Well, one panel gets shaded.

It degrades the entire system. But anyway, here's a photo of my new roof. So I moved my old three kilowatt from here over to here. so I got a perfect cloudless day and I thought, well, my old three kilowatt system here is not giving out as much as I thought.

I knew it would be less, but it was like significantly less than what I was, uh, expecting. So I went to investigate this on a perfectly cloudless day when I was getting a result on the new end phase system like this. The blue one is, uh, the production data here coming from the paddles. as you can see, it's a basically a perfect curve you'd expect on a cloudless day and I'd expect to see exactly the same graph on my three kilowatt system.

It'll just die out quicker because it's on the, um, other side of the roof and you know it won't give me high output like right into like 4 p.m 5 p.m in the afternoon like this. But this is what I got. This is the exact day this is the 17th last month. there's the exact same day.

Why am I getting data that has all these Wiggle Wiggle Wiggle years in here? What the heck's going on? This is the blue graph here. but the interesting part about this is right around 9 40 a.m Every day the big this this drop starts to happen. Here's the same data over a couple of days around that period. We just happen to have a lot of good sun.

Yeah, this one's 9 40 and the Wiggles start to happen. There you go, But like even on cloudy, uh yeah, look so it's all the same, right? It looks very similar. what the heck's going on and then you know other ones, right? Yeah, this is like a cloudy day. You'd expect like big drops coming in because the clouds come in and out and stuff like that.

But once again, around 9 40 a.m exactly the same thing starts to happen and it's just the same thing. So I thought, what the heck's going on here? So I started to investigate this and I was updating people on Twitter. I was like, you know, speculating about stuff and uh, things like that. So and of course everyone was coming up with their theories.

I was coming up with my theories of why okay, the panels are being shaded. I've moved them. Obviously the only difference is that I physically move them from this side of the roof exactly where these panels were to over here like this. But there are large trees over here so maybe early in the morning some shade comes over and you know, stuff like that.

But one of the uh, leading theories on Twitter. Eight out of ten engineers say it is, uh, a mains voltage issue. So I started monitoring, uh, the mains voltage. and then of course your engineering brain kicks in and you come up with like every sort of weird, uh thing you can think of.

Oh, because they're physically moved and the contractors up there have you know, manhandled them and they've like a damage. Maybe there's some micro cracking in the panel and as it heats up one of the cells as being you know, like like all sorts of things. maybe it's a dicky connection or something like that and that's starting to heat up. You know there's a lot of current flowing through these, so as the current like as it goes up there, you know there's something wrong with one of the, uh, physical connections in the entire stream.

There's lots of connections in a you know, a 12 panel string like this and maybe you know that's like that one of the joints is heating up and that's just like causing a higher resistance and that's causing the drop. and then you know that like all sorts of things start popping into your engineering brain. So I guess that's one of the curses of being an engineer is that all these things like you know too much. so all these things start flooding your head and you don't really think of like the simpler stuff you like.

start thinking of all these complex you know, solutions. It could be because you know that you know, like it can't be something simple, right? So what I've done here is I set up my 121 Gw meter to measure the mains, uh, voltage every 10 seconds. Okay, just put it near like a nearby powerpoint. it wasn't directly across the inverter.

So this was the lead in theory. That, uh, because I've added an extra uh, five kilowatt um, solar system which is, uh, of course, feeding into the mains. Okay, it was increasing the main voltage and everyone on my street. I think every house on my street has solar as well, So you know the addition of the extra five kilowatt system, especially close to my inverter.

It's basically wired in parallel because it's on. it's in my house, right? Both systems are feeding in to the 240 volt mains. That's how it works. and that was rising.

uh, boosting the mains voltage a little bit and I knew I was already on the borderline. for I was already like running pushing 250 volts on my mains anyway. So maybe as the sun went up like this, it had just like my new extra solar system plus all the other solar arrays on the street, the mains voltage just still got to a point where the Sunny Boy inverter would trip out and they don't actually tell you that sort of stuff in the manual. And of course, the Sunny Boy actually displays, uh, the mains voltage on the front of it.

Here's a photo of that and yeah, it was going around 250. You know, I think it was going up to like 252 or something like that in theory. Yes, nominal mains voltage in Australia is 230 volts, but we're still 240 volts at my home and my lab here nearby. But it can go up to uh, plus 10.

so that's 253 volts is the actual maximum. But I found settings in the Sunny Boy inverter that indicate that it didn't really trip out at that voltage, and I was able to change those and it didn't really make a difference. Anyway, I was able to graph this mains voltage here with my 121 Gw meter. So every 10 seconds, I monitored the mains voltage and that's the blue here and then I was able to overlay.

This was a bit tricky. I had to figure out how to do this in Excel. This was a pain in the ass trying to actually overlay graphs with different time sample periods. So, 10 seconds for the mains voltage and five minutes for the solar data.

Anyway, that was rather tricky. I just thought I'd mention that details on uh, Twitter actually my Twitter account where I solve this anyway. So I overlaid this and you can see mains voltage here 250 volts right? and then a solar output like this. And sure enough, around that 936, you know it's yeah.

Near 9, 40 am, Boom. It plummets. The output plummets like this. I think it's around almost 15 percent and the mains voltage does actually go down.

actually, right at that exact point, and then it continues to rise up. And then I was actually, I think I was experimenting with something there. So you can see that the mains voltage actually does rise. During the day, it goes up by, you know, a couple of volts and things like that, and then it tapers off and then, um, curiously, it actually goes up again at night around, you know, 7 p.m Something like that, it starts to rise again and it might be interesting to monitor over a couple and get a whole graph over a couple of days.

But so it wasn't like the inverter was shutting down or limiting or anything like that based on mains voltage. So I pretty much ruled out mains voltage for this. And then some people said, oh, it's the inverter temperature. It's got to be the inverter temperature.

It's heating up during the day. the sun's he did it. No, My panel's actually on the shaded side of the roof and it hasn't moved for the last eight years or whatever. So it wasn't that I knew and pretty much from that data, it wasn't the inverter, so aha, it must be shading.

So I decided to set up a time lapse uh, camera. So let's take a look at that data. So this is on a perfect day. let's have a look.

So we're now replaying the footage that we got here and right about now is when the big drop happens. You know about a 20 drop, 15 20 percent and as you can see like there was nothing I thought maybe like the sun might have been low So the shadow. There are some like bigger trees on the uh right hand side uh there but I couldn't like which is the eastern side but I couldn't really see anything. And of course because these panels are on the eastern side it tends to peak around like one o'clock or something like that.

and as the sun sets in the west it's yeah it's it's really on the wrong side here, but yeah nothing that I couldn't see anything there that would explain like a twenty percent drop. that's just crazy. You see the shadows really starting to yeah in the afternoon and then the clouds rolled in but otherwise perfect day. Okay I'm up on the roof and the Wiggles have started about 10 minutes ago.

You can see the shadow of the uh antenna uh pole here is not over them but I can see how as the shadow is going to slowly angle that way and it will actually come across the panels. and because these are string inverter they're all in series. If you upset one of them um it's going to upset the apple cart. you can see that the hopefully that the shadow from this wire is over the panel.

So I think the problem is the slight very slight shadow from the antenna here. and of course we never saw that when these panels were on the other side because um you'd only get that shadow in the morning and well, you hardly get any power at all out of these suckers in the morning. So yeah, I think it's just a subtle shadow, which I didn't see on the time lapse. Yeah, I find it hard to believe that a tiny little shadow like that could cause a problem.

Very diffuse shadow. but it's there. It's visible. It isn't the guy wire.

Well, it actually I can see the guy wire. Ah, just there. Not sure if this is going to show up on video, but I can just see the faint guy wire. But the coax of course is thicker and it's coming across.

And the shadow of the mast of the antenna is, uh, is just about to come across what is it now? Half past a freckle. There you go. It's a quarter past 10. so I reckon at about 20 past 10 that shadow is going to start come across.

and of course that shadow of the mast will be bigger than the guy wire. and uh, yeah, it's only covering a couple of cells there. But I guess that is enough to, uh, kill the output of that. Now of course, if we had micro inverters on these panels, we'd be able to see the output from this individual panel and then we'd be able to go.

Yep, it started on that panel and uh, yeah, but unfortunately, this is a string inverter so that just that tiny little shadow. There seems to be enough to, uh, you know, take off like 10 20 of the output of the string. That's pretty remarkable. I can understand the shadow from the pole, but shut a little shadow from a coax.

Unbelievable. All right. It's now 11 a.m You can see that the shadow the mast has really come across. Um, what one? two, three panels.

Although the further you go out, uh, out there, of course, the more, uh, diffuse it actually gets. I mean, you can see the sharp shadow across there. It's it's still fairly diffuse like. and you can still see the coax one.

I can. anyway. I can still see the coax going across there. Of course it'll be the orientation sorry about the wind.

It'll be the orientation of the shadow as well in relation to the cell. whether or not you know it kills that cell. And then of course the panels have bypassing in them. so the panels will bypass and you'll lose like a a paddle or two out of this thing.

anyway. So what I'm going to do now is take down the mast. If it goes away as I predict it will, then yeah, it's one of those log periodic. uh, jobbies for those digital Tvs.

One of them snapped off Bloody Kookaburras. It's definitely worth the test. And uh, if I'm losing this amount of power, I'm definitely going to move the antenna. Okay, 10 past 11 and the antennas down Ski no more shadow so I'll be able to look at my uh data now pretty immediately.

Next five, ten minutes. and if it jumps right back up to the uh, what is 17 1800 watts that it was at? uh, before then that? that was it. the diffuse shadow of the Master. We, you know, I kind of expected if it was like a like a big sharp shadow.

but it's you know. Especially that little coax. And yeah, you certainly would have been able to see that if you had micro inverters on each panel where you can monitor the graph for each individual panel. Well, that's the problem with these string inverters and I've just come back down.

What do you know? it's back up. 1816 watts. It was down like 1400. Something like that.

Uh, before I don't even have to look at the graph to know. Yep, that was the problem that fixed it. A tiny diffuse shadow from the mast. Some will say, oh yeah, that was huge Dave.

Obviously it was that. Um, yeah, you couldn't I couldn't actually see that on the time lapse video put up. I have to have another careful inspection of it anyway. Um, yeah, I reckon that's a winner winner chicken dinner.

So I'm gonna have to move my antenna and we're getting. there's the two kilowatt graph. Give it a knock as I'm getting like ten and a half kilowatt hours a day or something out of it. And we should get a lot more now, because, um, the antennas down.

so I have to mount that somewhere else. That's really annoying because I'm actually down in the gully here. So uh, just the bloody antenna shadow Mass and everyone on Twitter had, including myself, had all these wild theories about how it's you know, the inverter temperature? Log your inverter temperature. I'm currently monitoring the mains voltage.

Yeah, Like you know, I think eight out of ten people thought it was the mains voltage going high. Even if there was a correlation there, it's not causation. Correlation does not equal causation. So as you can see here, it easily dropped like 15 to even 20 percent right at that? That at 9, 40 A.m with just the diffuse shadow of that Coax cable coming across.

Absolutely amazing. Now that was only shading one panel. So even if the bypass diode is kicked in and that entire panel switched off due to just the shading of, uh, you know, a couple of cells on diffuse shading on a couple of cells on there, it still doesn't explain why it took out that sort of percentage because I've got 12 panels up there if you bypass one of them due to shading. However, however slight it was, that's only like just over eight percent of my output, but it's dropping almost double that.

Why? Well, it's obvious that the, um, it can't affect the other cells. but it's obvious that the maximum power point tracking algorithm in the Sunny Boy inverter is being upset by the fact that that one panel was being taken out. So it's not the fact that you just lose the one you know the output from the one panel. It screws up your maximum power point tracking and you can lose even more.

In this particular case, I lost at least double the amount. Shade one panel even partially and you lose twice the output. Like you lose two panels worth. It's just nuts, but because we don't have any actual data on how the Sunnyboy inverter works if you do, leave it down below.

But yeah, it must be that the maximum power point tracker was just going well. I can't handle it. and it's it. was just dropping the output even further.

So yeah, Amazing, huh? So anyway, let's go to the whiteboard and I'll explain how this uh, diode bypassing thing actually works. So this is the arrangement of a typical residential solar panel. Like I've got. Uh, they're usually in a 6 by 10 arrangement here.

60 individual cells and at around about 0.6 volts are per cell. Because remember a little, a silicon solar cell is effectively just a diode that actually produces voltage when photons hit it when light shines on it. So it's effectively just an individual diet. and they're all strung in series like this.

All 60 of them. So 60 times about 0.6 volts roughly. Uh, it will give you the typical roundabout 36 maybe up to a 40 volts maximum that you'll get from a typical residential panel like this. There are commercial panels that actually have larger numbers of cells in series, but you really won't find many of those in residential applications.

So this is a pretty standard panel. you get across all different manufacturers. This is not like Lg specific like I've got in my Lg Mono X panels and most panels these days. The old ones way back in the old days didn't have any protection built in, but these ones actually have bypass diodes built in and most of them will have three bypass diodes.

And that's what my Lg Monox panels have. So this is actually the physical arrangement of my Lg Mono X panels that we've been looking at. And so these three bypass diodes will allow, uh, current to actually bypass this panel if it actually shades. Because remember, we've actually got in my installation.

We've actually got 12 of these panels all in series. So we've got not only 60 cells like this, we have 60 times 12. So that's an awful lot of cells all stacked up in series. That's why I can get about 450 volts nominal Dc out of my 12 panel string that I've got.

and that's incredibly dangerous. As I've looked at, they can cause fires when Dc arcs over, of course, inside the isolation switches. I'll link that video in here. My one actually failed.

It didn't catch on fire, but luckily, but it did fail. and that is a thing Because anyway, Dc is like a real pain in the butt. When arc's over, it's nasty business. It keeps self-sustaining Anyway, 12 of these panels in series, with each one of these panels having 60 individual cells.

So how these bypass diodes work is that they actually will provide a lower impedance path than a shaded string arrangement like this. So basically, if you've got shade on and in theory any one or multiple Um cells in this, uh, third of a string here, then the current will bypass these cells and go through the diode like this. And likewise, if you had you know Big Bird poop over here covering a bunch of cells, then this diode would provide a lower impedance path and it would bypass this particular string here. Now in theory, the best thing you can do is have a bypass diode on each individual cell.

You could have a diet across there, across there, across there, but they don't do that just because that's really difficult and expensive from a manufacturing point of view. So they put them all at the one end like this: If you actually go and inspect your solar panels, you actually see them wired in series like this. So it's just easier to put three diodes at the top of the array like this, and just compromise that if one, a couple of these cells get taken out here, well, you're going to lose the output from all of these other cells as well. It's just it's just a compromised thing.

uh, based on manufacturing and cost. So let's look at what happens when these actually get shaded. exactly how that works. Well, They're all in series as I said, like this.

So the current for the entire string of the 12 panels has to flow through every single cell like this. And if one cell actually gets shaded, what actually happens in that particular case is that actually because the current is still being forced through that cell because they're all in series you can't like. Imagine if like, one cell failed like this and uh, it would like literally take down your entire string. uh, your 12 panel array, it'd be completely gone if it went open or something like that.

Now, when this cell here gets shaded because the current is still being forced through that cell, it's all in series. Think of it like a constant current source pretty much and it's still being forced through a cell which has no sunlight on it. Then what happens is this cell actually generates a negative voltage. So yes, like it'll be normally be positive here and negative here, but it will actually be forced to generate a negative voltage there.

and this cell will actually heat up. And these are called hot spots. And you can actually detect these with uh, thermal cameras on uh, your panels. And that's really bad.

That can degrade the uh life of individual cells in your panel and because they're all in series. If that heats up and internal connections break or whatever, then your entire string is gone. So you know. Well, at least this string here.

So it generates a negative voltage and that might be okay for one little tiny cell. Like here, you'll still get the current being forced through like this. So normally you'd have the current flowing through the string in this direction like this. but it gets to a point where you have one or more cells shaded here that the diode actually represents a lower impedance path.

um, than going through here and forcing it through all of these reverse voltage cells. so you can sort of think of it like it switches on the diode. But that's not really the case. It's effectively.

it's just because diodes are and they have an impedance they have. You know it's they have a voltage drop, but they effect have an effective impedance. So if that's actually lower impedance path than this because of the negative, uh, voltage of these cells, then the current. More of the current's going to flow up through these diodes to a point where like all of it flows through there and you're just going to like it's just going to essentially bypass this entire string.

And because this diode is usually a schottky diode, it's less than the 0.6 volt drop. Then you know it's going to effectively clamp. Once the current starts flowing through there, it's going to effectively clamp the voltage across this entire string to that 0.45 volt. So you know, technically, if you shade your panel, it's still going to heat up.

This bypass diode is going to drastically reduce that and essentially, uh, protect all of these cells in here from overheating and getting hot spots and ruining your day and ruining your life of the panel because these things are supposed to have like 25 year plus life on them. In fact, some of them have 25. My ones have 25 year warranty. So if they didn't have the diodes in there and you're getting hot spots due to all shading and bird poop and whatever it is or bloody coax cables, then um, yeah, these will get hot and they'll eventually just fail completely.

But I know there's people out there with their thinking cap on. Something Doesn't seem right here. Negative here. Positive Here, The diode is in this direction like this, and I've drawn current flowing in here into the negative terminal and going up the diode like that and then eventually out like this.

but does a conventional current flow this way from positive to negative? In which case, these diodes can never switch on? Well, yes, that's true conventional current, but you've got to remember that the solar panel is not a component. It's not sinking current. it's sourcing current. So if if you had a battery over here and your solar panel was a resistor or whatever, it was a load.

Yes, conventional current would flow like this. and you would never get current flowing through those diodes, they'd all be reversed bias forever. But because the solar panel is actually effectively a battery, it's a source. Current flows conventional current, not electron.

current flow. Rubbish. Conventional current flows out of the positive terminal. That's why current flows in like this.

And this is why current flows up and through and bypassing the diodes. Got it? So, as we saw in my particular case, we had the very faint little coax. It was just like a little bit of shadow here. In fact, I don't even think it was going across that top one right at 9, 40 a.m I think it was only going over these uh, two strings here.

So I think it was like at most it was like that or something like that, and that should have only. even if that totally took out. Um, both of these strings and bypass these diodes. I'm we're still only losing two-thirds of one panel, and that's like six, five and a half.

Six percent something like that. But we saw we actually got a drop of 15 to 20. Um, almost up to 20 percent drop. And that's due to the to the inverter that we use in Sunny Boy Inverter which is maximum power point tracking and it's all completely screwed up and we won't go into details that would have to be its own separate video or whatever.

But yeah, some light shading on a couple of cells in there like that caused um, yeah, like a 20 drop of the entire string am I? And it wasn't even over the full cell. It's not like the whole cell was covered, it was just like it had a you know, a fuzzy little shadow over the one cell and that was enough to cause that and screw up everything else and ruin your day and lose 20 of your uh, capacity of your array just by a tiny little shadow like that. Absolutely amazing. And then when we saw the mast actually come over with this big thicker shadow that was actually covering three Uh panels.

and it was a big thicker shadow as well. And in that particular case, it only dropped by another 10. So yeah, it's it's really. it's not just the shade in the panel, it's also the inverter that you hooked up with.

And that's the advantage of using micro inverters for each panel. Your maximum power point tracking will be better on each panel rather than the entire string of, in my case, uh, 12 panels in series, so it'll be able to handle that better as well as uh, you know, not taking down the rest of the array as well if you just got shading on the one panel. So yeah, there's advantages to those micro inverters, but string arrays and inverters are actually a cheaper solution. But yeah, there's some trade-offs, so there you go.

I hope you found that really interesting. I was actually really surprised. Like I knew Shade. I knew this problem, but I didn't think.

even if um, you know it was shade a little bit, I didn't think it would have like a 20 drop that I've seen. I thought that was really remarkable for a fuzzy little bit of coax shadow there. Absolutely remarkable, so I've learned something. Hope you did too.

Catch you next time you.