Blocking diode with parallel solar panels

[tanglewood]

I'm looking for a sanity check here. The situation is multiple parallel solar panels (4, to be exact) feeding a single MPPT charge controller. The panels are 60V Vmp, 6A Isc, and battery voltage 24V.


Historically blocking diodes would have been used to prevent nighttime battery drain back through the panels, but all modern controllers now have this blocking capability built in, so nobody seems to use them anymore.



The question, however, has to do with operation when one or more of the parallel panels is shaded. If one is shaded, it will become a power sink and drag down the voltage of the unshaded panels. Correct?


So wouldn't the best shade immunity be achieved by installing a blocking diode at the output of each panel? That way a shaded panel would not produce any current, but it would also not become a power sink and drag down the voltage of the whole array.


I supposed it depends on exactly how much shading there is, and what the reduced voltage and net current is with the panels interconnected without blocking diodes. How much net output drops will depend on the exact shading, and also what the bypass diode arrangement is in the panel itself.


With blocking diodes, I think power will diminish in increments of a whole panel. I think the more parallel panels you have, the more loss there will be by dragging down the voltage of the whole array vs dropping out a single panel, assuming a single panel is shaded. But again, it depends on the specific situation.



There also is the issue of having enough parallel panels such that unshaded panels can overload a shaded panel and pop it's fuse or burn it out. In the particular system I'm working on, there are four panels, 6A Isc each. And each panel has a max fusing requirement of 15A. If three unshaded panels are dumping their load into a single shaded panel, that's 18A into a panel that can only handle 15A. That will blow the fuse, or damage the panel if not fused.


Does this sound right? I don't see, read, or hear anyone using blocking diodes for parallel panels, and it seems to me that they are needed.


Thoughts?

[TeddyDiver]

I'm no expert in this but if there's a blocking diode allready it fixes your problem too. No need for another. Besides every blocking diode drops 0.5v so not useful to have more than is needed.

[tanglewood]

Originally Posted by TeddyDiver

I'm no expert in this but if there's a blocking diode allready it fixes your problem too. No need for another. Besides every blocking diode drops 0.5v so not useful to have more than is needed.

I think it depends on what problem you are trying to solve, and where the blocking diode is inserted. To prevent battery drain into the panels at night, the blocking diode in the controller is sufficient.


But a blocking diode in the controller won't prevent a producing panel from draining into a parallel non-producing panel. That's the problem I'm trying to solve, assuming it's actually a problem, which I think it is.

[GordMay]

Ideally, you'd want individual charge controllers (for each panel), each having an integral blocking diode.
I think, you're proposing a BYPASS diode, which is another method of dealing with a shaded panel, in a series string.

[TeddyDiver]

Originally Posted by tanglewood

I think it depends on what problem you are trying to solve, and where the blocking diode is inserted. To prevent battery drain into the panels at night, the blocking diode in the controller is sufficient.


But a blocking diode in the controller won't prevent a producing panel from draining into a parallel non-producing panel. That's the problem I'm trying to solve, assuming it's actually a problem, which I think it is.

Dunno about all, but every solar panel I've seen had a blocking diode installed..

Gord! He was talking about paralled panels so no bypass diodes..

[GordMay]

Originally Posted by TeddyDiver

... Gord! He was talking about paralled panels so no bypass diodes..

Indeed. Edited accordingly.

[tanglewood]

Originally Posted by TeddyDiver

Dunno about all, but every solar panel I've seen had a blocking diode installed..

Gord! He was talking about paralled panels so no bypass diodes..

Are you sure those aren't bypass diodes in the panels?


What's typical today is that panels have bypass diodes included, and charge controller have blocking diodes included. So nobody pays any attention to it today, and pretty much nobody (including me) really understands the difference and the needs.


So I'm brushing up on it, and the fundamental question I think is whether an unshaded panel will backfeed into a shaded panel when they are connected in parallel. If a battery will backfeed a shaded panel without a blocking diode, then wouldn't a lighted panel also back feed a shaded panel?

[Martkimwat]

Blocking diodes on each panel will protect the panels from reverse current so no need to worry about fuse rating so long as the fuse is between the blocking diode and the panel, however as others have said, you will lose 0.5v on each panel. Whether that is a real issue with a MPPT controller is debatable. The other option is to have a separate controller for each panel, that is what I have (sort of )done. I have three small 18w panels fed into one controller, then my other two panels (38w and 80w) each have their own controllers. Because the controllers are controlling a smaller amount of power each, they are less expensive. I have two Genasun GV4 (4 amps each) controllers for the 3x18w panels and the 38w panel and an Epever Tracer 10Amp controller for the 80w panel. Seems to work fine. Because the 18w panels are all pointing in different directions (on each on the coachroof sides and one inside the fore hatch) I do have Schottky diodes on all three of those.

[Martkimwat]

Originally Posted by tanglewood

Are you sure those aren't bypass diodes in the panels?


What's typical today is that panels have bypass diodes included, and charge controller have blocking diodes included. So nobody pays any attention to it today, and pretty much nobody (including me) really understands the difference and the needs.


So I'm brushing up on it, and the fundamental question I think is whether an unshaded panel will backfeed into a shaded panel when they are connected in parallel. If a battery will backfeed a shaded panel without a blocking diode, then wouldn't a lighted panel also back feed a shaded panel?

In short, yes they will unless they already have blocking diodes integral to the panels or you fit ones on each panel Even today not every panel ,even expensive ones have integral blocking diodes) With your size of panels I would go with three separate controllers, especially if they are different brands or ratings as each manufactures panels have slightly different characteristics which would be looked after by a dedicated controller fo each.

[sailcrazy]

Following. We have 4 rigid panels up on our Radar Arch. We installed a blocking diode on the (+) wire from each panel just before it connects to a common (+) busbar, with a single heavy wire that is then runs to the single Victron controller, and then to the 12v house bank. The (-) from ea panel goes to a separate busbar that also connects to that controller. Seems to work well, and we've noted no significant reduction in power when some shading happens.

[rha9]

I just asked a very similar question of marine solar power supplier, and this is the response that I got -

Bypass diodes are in fact in the panels, you are correct for when individual cells are shaded the panel can still output power, rather than going into open circuit state. If you chose to then parallel the panels into the same controller rather than each panel having its own controller you then need the blocking diode. It prevents backfeeding when say one panel is shaded and the other is in sun. Since you've created one circuit by paralleling wires together into the same port on controller, the shaded panel becomes a load and the sunny panel then dumps its energy into the shaded panel rather than your batteries.

These diodes rob ~1V from the power output off the top so we always recommend 1 controller per panel to maximize the highest efficiency systems.
We sell Blocking diodes that go inline with with MC4's. To test what I have stated above get an amp clamp and measure the outputs down stream of controller when both are in sun. Then cover one panel and see what that output is...

[Suliwar]

Blocking diodes serve also to solve a safety problem. Usually solar PV modules are not fused since able to generate a short-circuit current only slightly higher (+20%) than their rated current.
Instead common electrical installations are able to generate very high short-circuit current values far beyond their rated current, which is why protection by fuses is required.
However in case that several unfused PV modules are connected in parallel, in case a short circuit occurs in one of the PV module branches, without a blocking diode, all the other PV modules connected in parallel will discharge together the sum of all their short-circuit currents into this one single shorted connection. Since related cabling is usually sized for the short-circuit current of only one single PV module, this provokes an overload conditions which might result in a fire hazard.

[skipgundlach]

When I was in the process of planning my solar upgrade, I asked the tech rep of the adequately-sized controller I already had what I needed.

While we were still within potential parameters to run them in series, it COULD have gotten very close; he recommended parallel.

He then followed up with essentially the same comments as seen a couple of spaces up, so we ordered a pair and installed them.

Just because I'm nervous about needing, maybe, to get into that power line at some point, I installed a breaker in the + line after our swamping killed the previous controller (threatened to catch on fire! - forcing me to disconnect under load, which I was very reticent to do).

[tanglewood]

I ran a little test today and confirmed that with two or more panels in parallel, shaded panels will become a load on the unshaded panels. So if you expect uneven shading across multiple parallel panels i.e. you own a boat, blocking diodes will be a big benefit.


The specific test I ran was to apply a variable voltage to a shaded panel using an external power supply, the power supply simulating an unshaded panel. As I increased the voltage, there was very little current flow until it reached about 80% of the rated voltage, then it started to flow. Beyond that, current increased as voltage increased. So a panel in the sun producing rated voltage will indeed backfeed into a shaded panel, turning the shaded panel into a wasteful power load.


I'm really surprised that this isn't more of a topic of conversation with cruisers since unevenly shaded panels is common, if not a certainty on a boat. With land based systems I suspect that uneven shading is infrequent enough that nobody cares. Plus the trend with land systems is towards series wired panels for much higher voltages.


Of course a charge controller dedicated to each panel (or string) is probably the optimum approach, but it won't always be possible. This particular boat I'm working on has the 4 panels wired in parallel, and it would be a significant change to alter that.


Now as many know, there is a down side to blocking diodes since they consume power all the time when the panel is producing. With older, low voltage panels, the loss is a larger percentage of the panel's total output. But with higher voltage panels, the loss is a very small fraction. I'm working with 360W panels, and expect the diode loss under full power to be 3W, so less than 1%. It's well worth it when a shaded panel can otherwise consume the full output of an unshaded panel.

[Martkimwat]

Originally Posted by tanglewood

I ran a little test today and confirmed that with two or more panels in parallel, shaded panels will become a load on the unshaded panels. So if you expect uneven shading across multiple parallel panels i.e. you own a boat, blocking diodes will be a big benefit.


The specific test I ran was to apply a variable voltage to a shaded panel using an external power supply, the power supply simulating an unshaded panel. As I increased the voltage, there was very little current flow until it reached about 80% of the rated voltage, then it started to flow. Beyond that, current increased as voltage increased. So a panel in the sun producing rated voltage will indeed backfeed into a shaded panel, turning the shaded panel into a wasteful power load.


I'm really surprised that this isn't more of a topic of conversation with cruisers since unevenly shaded panels is common, if not a certainty on a boat. With land based systems I suspect that uneven shading is infrequent enough that nobody cares. Plus the trend with land systems is towards series wired panels for much higher voltages.


Of course a charge controller dedicated to each panel (or string) is probably the optimum approach, but it won't always be possible. This particular boat I'm working on has the 4 panels wired in parallel, and it would be a significant change to alter that.


Now as many know, there is a down side to blocking diodes since they consume power all the time when the panel is producing. With older, low voltage panels, the loss is a larger percentage of the panel's total output. But with higher voltage panels, the loss is a very small fraction. I'm working with 360W panels, and expect the diode loss under full power to be 3W, so less than 1%. It's well worth it when a shaded panel can otherwise consume the full output of an unshaded panel.

Modern Schottky diodes only drop the voltage about 0.5 volts but they don't affect the current so as you correctly say, modern panels are less likely to be so much affected, I believe you are incorrect in saying larger panels will be less affected since the diode does not affect the current which is what a larger panel will give you more of.