diy solar

diy solar

Hyperion bifacial 400W panel string drops 80% when one panel shaded

axum

axum

New Member
Joined
Nov 12, 2023
Messages
15
Location
Denver, CO
I bought 12 Hyperion 400W bifacial panels recently and have 8 of them currently hooked up in series and connected to a Sol-Ark 15K (MPPT 1)

The location of the install (ground mount) is not ideal as one panel gets partly shaded after around 2PM. I know I could move the panel to another location but that's not really the point of my question.

So I'm getting 3000+ watts when all panels get direct sunshine (which is pretty good, given that the orientation isn't perfect). However, once ONE panel gets shaded, the output immediately drops to 450 watts or less, even though the other 7 panels are still fully lit by the sun. I always thought that bypass diodes in the panels would prevent such a thing.

Does this particular panel not have bypass diodes? If it does, why is this happening?

Here's the chart of the power output:

1719354948794.png


Disregard the numbers for power output on the left. Must be a bug in the My-Solark website because the actual power was 3000W at the peak. Also ignore everything to the left of the 3000W peak around the middle of the graph.

The purple line is the power output of the panel, it drops to 450W when one panel gets shaded. The red line is the voltage, green line is current.

The red circle indicates when one panel got shaded (started by a little shade and within an hour it was fully shaded)
 
From my admittedly incomplete understanding, bypass diodes can only bypass INTERNALLY within a single panel that contains them. This has also been my experience when I connected a few 250W Trinas in series so I could experiment with a piece of cardboard covering various portions to see the effects. I found playing with full-cell panels that the edges of the panel were extremely sensitive to shading - covering a cell or two on the edges had a much larger smackdown on power than shading near the middle. It all depends on how the individual cells are wired.

If an entire panel gets more or less shaded as you are describing, any bypass diodes within might not be able to compensate.

Before somebody jumps down my neck for even stating the above, I have ONLY experimented with what I've got. Perhaps things are now different.

This is always a weakness of series-connected ANYTHING. If only one of the anythings goes sideways, the whole string of anythings hits the skids.

Try manually shading another panel in your string and see if the same thing happens.
 
Hey Axum, most panels do have bypass diodes, which will help individual panel production under partial shading. However, panels in series will all be affected by the shading of one panel as the bypass diodes do nothing to bypass current from one panel to another.

My recommendation would be to either utilize DC optimizers which should bypass the underperforming panel when it is shaded to prevent it from bringing your whole string down, or to move the problem panel out of the shade. (This exact issue is partially why DC optimizers were made)
 
Carlos_Sol-Ark said:
Hey Axum, most panels do have bypass diodes, which will help individual panel production under partial shading. However, panels in series will all be affected by the shading of one panel as the bypass diodes do nothing to bypass current from one panel to another.

My recommendation would be to either utilize DC optimizers which should bypass the underperforming panel when it is shaded to prevent it from bringing your whole string down, or to move the problem panel out of the shade. (This exact issue is partially why DC optimizers were made)
Click to expand...

I based my assumption on the following video:
. Fast-forward to 13:30

Is this information not correct? It does make sense to me though, because all 3 bypass diodes would pass the current if the panel is blocked/shaded.
 
v_green57 said:
Here's the datasheet (I hope) based on your description:


Page two, upper left:

Junction Box IP68 rated (3 bypass diodes)
Click to expand...
I don't think that's the same panel. There's a model that ends in "B", which does list the diodes, but the one I have is https://signaturesolar.com/hyperion-400w-bifacial-solar-panel-black-up-to-500w-with-bifacial-gain/ , model number HY-DH108P8 (without the B). The spec sheet also doesn't mention bypass diodes: https://signaturesolar.com/product_images/HY-DH108P8 390-410W-202110 (short cable) (2) (2).pdf
 
Ah, YouTube and the internet in general...

It never lies. But it sure generates clicks if you spell LIE in big capital letters to attract eyeballs...

If it wasn't so hot here (Phoenix) right now, I'd haul some panels out to the parking lot, connect them in series, and try shading one, now that you've got me wondering.

But it's just TOO DAMN HOT.
 
Carlos_Sol-Ark said:
Hey Axum, most panels do have bypass diodes, which will help individual panel production under partial shading. However, panels in series will all be affected by the shading of one panel as the bypass diodes do nothing to bypass current from one panel to another.

My recommendation would be to either utilize DC optimizers which should bypass the underperforming panel when it is shaded to prevent it from bringing your whole string down, or to move the problem panel out of the shade. (This exact issue is partially why DC optimizers were made)
Click to expand...
Actually, the drop from 3000W to 450W was when only a small part of the panel was shaded. It eventually dropped to 200W when the whole panel was shaded.
 
v_green57 said:
Ah, YouTube and the internet in general...

It never lies. But it sure generates clicks if you spell LIE in big capital letters to attract eyeballs...
Click to expand...
I understand, and I agree, but the schematics did make sense in this video. He also did a followup video demonstrating this by measuring current across the diode.
I'm not saying he is right, it's just my hunch that bypass diodes should work as expected when one panel in a string is shaded.
 
axum said:
I don't think that's the same panel. There's a model that ends in "B", which does list the diodes, but the one I have is https://signaturesolar.com/hyperion-400w-bifacial-solar-panel-black-up-to-500w-with-bifacial-gain/ , model number HY-DH108P8 (without the B). The spec sheet also doesn't mention bypass diodes: https://signaturesolar.com/product_images/HY-DH108P8 390-410W-202110 (short cable) (2) (2).pdf
Click to expand...

Ummm...it doesn't?

Page two, upper right, six lines down:

Mechanical Characteristics Solar Cell Mono PERC 182 mm
No. of Cells 108 (6 × 18)
Dimensions 1722±2 × 1134±2 × 30±1 mm
Weight 22.6kg (±5%)
Cable Cross Section Size 4mm2(IEC),12 AWG(UL)

Junction Box IP68 rated (3 bypass diodes)

Output Cables Portrait: (-)350 mm and (+)160 mm in length or customized length
Front/Back Glass 2.0mm AR Tempered glass 2.0mm Semi-tempered glass
Container 36 pcs/Pallet, 936 pcs/ 40’HQ

I don't think any so called "grid tie" panels don't have bypass diodes these days. Benefits of having them for slightly shaded panels are just too great.
 
v_green57 said:
Ummm...it doesn't?

Page two, upper right, six lines down:

Mechanical Characteristics Solar Cell Mono PERC 182 mm
No. of Cells 108 (6 × 18)
Dimensions 1722±2 × 1134±2 × 30±1 mm
Weight 22.6kg (±5%)
Cable Cross Section Size 4mm2(IEC),12 AWG(UL)

Junction Box IP68 rated (3 bypass diodes)

Output Cables Portrait: (-)350 mm and (+)160 mm in length or customized length
Front/Back Glass 2.0mm AR Tempered glass 2.0mm Semi-tempered glass
Container 36 pcs/Pallet, 936 pcs/ 40’HQ

I don't think any so called "grid tie" panels don't have bypass diodes these days. Benefits of having them for slightly shaded panels are just too great.
Click to expand...
You're right, it does mention the bypass diodes, but on the other hand the spec sheet shows (B) next to the model number so maybe there's a difference between HY-DH108P8 and HY-DH108P8B. I'm just guessing here.
I would be surprised if this panel didn't have bypass diodes. That wouldn't make sense.
 
My only though was maybe a panel being bypassed was putting you under MPPT range, but the math doesn't seem to agree with that.

Solark15k
Rated MPPT Operating Voltage Range 175 - 420V
MPPT Voltage Range 150 - 500V
Startup Voltage 125V

If you're putting out 3kWh with 8 panels that 375W per panel, closer to STC than NOMT.

Even at NOMT, 7 panels should be (28.8x7)=201V. Going down to 6 might start making issues, but 7 ought to be fine.

There do appear to be slight Operating Voltage/Amps depending on B/non-B panels, but both should be pretty similar performance.
 
Do you have a picture of the shaded and non-shaded array?

You said you tried with a different panel in that spot and same results. Have you tried completely removing that panel/location from the string and seeing what happens?

What you describe really sounds like the Christmas lights effect, but as this entire thread has said, bypass diodes should stop that. I wouldn't 100% assume the panel is fine because you swapped one from the same batch and had the same results. If one is bad it's not unlikely that the same flaw would be in the others.
 
Could a different inverter with MPPT change the output values. Fronius and SMA come to mind.. I dont know of any MPPT shading tests between various inverters.
 
I did some more experiments and I am starting to be convinced that these panels don't have any bypass diodes at all.

I measured current and voltage of the string (while under load) and when any single panel is being shaded, the voltage barely drops. If there were bypass diodes, and a significant amount of current flows through the diodes, the voltage should drop significantly (by 15-30 volts) but it doesn't. The current drops by a whopping 90% though. Shading a second panel doesn't do anything at all. Voltage barely changes and neither does current.

I'm seriously considering trying to pry open one of the junction boxes to see if there's any diode(s) in it.
 
Update...

So as I said, when one panel gets shaded, output drops dramatically (from 3000W to 400W), but that's diffuse shading caused by trees.

Now, when I completely cover the shaded panel with cardboard, power immediately jumps to 2600W and voltage drops by 30V. So that indicates that the bypass diodes did kick in with 100% shading. It seems to be a true on/off switch though, so if shading is diffuse, the bypass diodes don't pass any current until a certain threshold is reached.

I wouldn't expect the bypass diodes to work this way.
 
axum said:
Update...

So as I said, when one panel gets shaded, output drops dramatically (from 3000W to 400W), but that's diffuse shading caused by trees.

Now, when I completely cover the shaded panel with cardboard, power immediately jumps to 2600W and voltage drops by 30V. So that indicates that the bypass diodes did kick in with 100% shading. It seems to be a true on/off switch though, so if shading is diffuse, the bypass diodes don't pass any current until a certain threshold is reached.

I wouldn't expect the bypass diodes to work this way.
Click to expand...

They work that way because there are only 3-5 diodes in a whole panel. Each diode is around a group of 20-30 individual solar cells. So it takes a lot of shading for that many cells to fall below -0.7 volts and cause the diode to be forward biased and pass current.

In the situation you have, with diffuse shading, bypass diodes are unlikely to help much. You would benefit a lot from something like Tigo Optimizers. And you only need them on panels that get the shading. And you don't really need the monitoring solution. Just the optimizer modules. Personally, I find the Tigo monitoring (CCA and TAP(s)) to be a PITA to get working correctly and not of that much value. But the optimizers are really great and not much more than a rapid shutdown module.

I know you aren't required to have shutdown modules on a ground mount. But you will have to have a rapid shutdown transmitter along with the Tigo Optimizers because there isn't a way to disable the RSD support in the optimizer. Unless your inverter has Tigo RSD support built in.
 
jpwhit said:
They work that way because there are only 3-5 diodes in a whole panel. Each diode is around a group of 20-30 individual solar cells. So it takes a lot of shading for that many cells to fall below -0.7 volts and cause the diode to be forward biased and pass current.
Click to expand...

But shading doesn’t cause the cells to drop in voltage that much. It causes the cells to reduce maximum current they can generate/pass through

MPPT should find the optimum current and voltage operating point for the string. Here, somehow the diodes are not present/failed open (which would be nuts because that exposes the cells to damage) or the MPPT is confused
 
There have been some theories on the forum that half cut panel’s different shading response behavior can confuse MPPTs. Also if you hard shade only half the panel the other panel can still pass 50%.

If it is diffuse shade across the whole panel I would expect the half cut to have a similar behavior to a full cut.
 
jpwhit said:
Then funny that this graph from that website shows what I'm talking about.......

View attachment 225088
Click to expand...

Thanks for clarifying that. I was looking at the left part of the graph and missed the diffuse shading callout. Doh.

Wouldn't the same problem happen if one cell was hard shaded and the other 19 cells (in a 60 cell panel with bypass diodes) were in full light?

Why isn't this issue declared more commonly in the forum/elsewhere when bypass diodes come up? Or maybe I just missed it all the time due to confirmation bias.
 
You must log in or register to reply here.

Similar threads

T
Disappointing output from Hyperion panels
Replies
15
Views
631
42OhmsPA
4
blrmaker
Bifacial Panel Questions
Replies
8
Views
363
sunshine_eggo
sunshine_eggo
plympton
Portland, OR: 4 Hyperion 400w bi-facial panels $120/ea
Replies
6
Views
997
chilly2
chilly2
V
Alrska Panel Output Drops 80% at 80F
Replies
0
Views
71
vvgogh
V
timblack1
Unexpected voltage drops in panel array
Replies
1
Views
170
MaikaiLife
MaikaiLife

diy solar

diy solar
Back
Top