Author Topic: Question: Why blocking diodes and not MOSFETs on solar panels?  (Read 1052 times)

0 Members and 1 Guest are viewing this topic.

Offline Dave92F1Topic starter

  • Contributor
  • Posts: 32
  • Country: us
    • Nerdfever
Question: Why blocking diodes and not MOSFETs on solar panels?
« on: January 25, 2025, 05:15:53 pm »
Forgive me if this has been asked before.

In my experience, solar panels are universally connected with blocking diodes, to prevent power loss from current flowing backward thru the panel when it's not fully illuminated.

Why a blocking diode, which loses power due to the forward voltage drop, and not a MOSFET circuit?

Using a MOSFET to stop the backward flow of current would seem to work just as well, and the loss would only be due to the RDSon, probably losing less than 20% as much power.

There must be a good reason. MOSFETs are cheap. Why?
 

Offline uer166

  • Super Contributor
  • ***
  • Posts: 1030
  • Country: us
Re: Question: Why blocking diodes and not MOSFETs on solar panels?
« Reply #1 on: January 25, 2025, 07:00:20 pm »
Since when do they have series blocking diodes? Citation required about the "universally have blocking diodes" part.

Almost all large solar panels that I've worked with have parallel bypass diodes for sections of it, to prevent partial shading from reducing string power to zero. I don't remember there being series diodes though.

If you do see series output diodes, it would actually make sense to not make them FETs due to complexity and low loss that diodes already have. Say for a 80V panel, a 0.5V schottkys will waste 0.6% of available power, why do you care to reduce that to 0.2% with a FET?
 

Offline Siwastaja

  • Super Contributor
  • ***
  • Posts: 9432
  • Country: fi
Re: Question: Why blocking diodes and not MOSFETs on solar panels?
« Reply #2 on: January 25, 2025, 07:38:32 pm »
Those parallel bypass diodes of course could be ideal MOSFET diodes but what OP is missing is control of the MOSFET. MOSFET itself isn't magically an ideal diode. Remember to consider leakage current and risk of accidental activation. And the cost of providing a power supply for the controller, which needs to supply gate voltage for the MOSFET.

Any gain would be completely marginal. It doesn't make sense to put extra money into extracting some 0.1% more power under partial shading condition, which should be <10% of the time so 0.01% in total.

Adding a few more of those bypass diodes would be a simpler and cheaper optimization. Like, for a 60-cell panel you can have three (one bypassing each 20 cells). Or you can add a fourth so that the three still bypass 20-cell segments and the fourth bypasses all 60 cells, so that if all three segments are shaded then the voltage loss over the panel is smaller (because the current from other panels only needs to pass one diode, not three). Maybe this fourth diode is worth it, but just barely so. Any more complicated than that and it won't make any sense.
« Last Edit: January 25, 2025, 07:47:00 pm by Siwastaja »
 
The following users thanked this post: 5U4GB

Online coppice

  • Super Contributor
  • ***
  • Posts: 10123
  • Country: gb
Re: Question: Why blocking diodes and not MOSFETs on solar panels?
« Reply #3 on: January 25, 2025, 07:43:27 pm »
There are special devices based on MOSFETs made specifically for that application. I think TI make some, but they are not alone. I've never really looked at how they work, but I assume they use more than 2 terminals, so get enough voltage to hard switch the FET.
 

Offline nctnico

  • Super Contributor
  • ***
  • Posts: 28492
  • Country: nl
    • NCT Developments
Re: Question: Why blocking diodes and not MOSFETs on solar panels?
« Reply #4 on: January 25, 2025, 10:30:58 pm »
A MOSFET is not going to work as it conducts in both directions when 'on'. So when a shading condition stops, the MOSFET will short the cells it is connected across. IOW: you'd need a very complicated control circuit which isn't going to work better or more reliable compared to a 'dumb' diode.
There are small lies, big lies and then there is what is on the screen of your oscilloscope.
 

Online coppice

  • Super Contributor
  • ***
  • Posts: 10123
  • Country: gb
Re: Question: Why blocking diodes and not MOSFETs on solar panels?
« Reply #5 on: January 25, 2025, 10:44:10 pm »
A MOSFET is not going to work as it conducts in both directions when 'on'. So when a shading condition stops, the MOSFET will short the cells it is connected across. IOW: you'd need a very complicated control circuit which isn't going to work better or more reliable compared to a 'dumb' diode.
Of course it requires more the a MOSFET. The whole point of using a switch is to switch it. The purpose it to cut the voltage drop well below what a simple diode, or even a Schottky diode, would drop, just as you do in synchronous rectifier applications.
 

Online langwadt

  • Super Contributor
  • ***
  • Posts: 4882
  • Country: dk
Re: Question: Why blocking diodes and not MOSFETs on solar panels?
« Reply #6 on: January 25, 2025, 11:38:35 pm »
A MOSFET is not going to work as it conducts in both directions when 'on'. So when a shading condition stops, the MOSFET will short the cells it is connected across. IOW: you'd need a very complicated control circuit which isn't going to work better or more reliable compared to a 'dumb' diode.

https://robotics.ong.id.au/2014/07/30/raspberry-pi-b-power-protection-circuit/ ;)
 

Online bdunham7

  • Super Contributor
  • ***
  • Posts: 8315
  • Country: us
Re: Question: Why blocking diodes and not MOSFETs on solar panels?
« Reply #7 on: January 25, 2025, 11:52:48 pm »
In my experience, solar panels are universally connected with blocking diodes

Only multiple strings in parallel would need such blocking diodes and then you'd be looking at a Vf of 0.7-1.2V in a string of 300-1000V.  That's pretty small and shouldn't really happen anyway during high-power production unless the system is badly designed.  Such diodes are cheap, simple and very durable. 

The bypass diodes internal to the panel are different, but again you'd be talking about a pretty small improvement and only during shading or malfunctions. 
A 3.5 digit 4.5 digit 5 digit 5.5 digit 6.5 digit 7.5 digit DMM is good enough for most people.
 

Offline nctnico

  • Super Contributor
  • ***
  • Posts: 28492
  • Country: nl
    • NCT Developments
Re: Question: Why blocking diodes and not MOSFETs on solar panels?
« Reply #8 on: January 26, 2025, 12:37:41 am »
A MOSFET is not going to work as it conducts in both directions when 'on'. So when a shading condition stops, the MOSFET will short the cells it is connected across. IOW: you'd need a very complicated control circuit which isn't going to work better or more reliable compared to a 'dumb' diode.

https://robotics.ong.id.au/2014/07/30/raspberry-pi-b-power-protection-circuit/ ;)
That circuit only works reliable (in volume production, over a wide temperature range and with room for external interference) with a large amount of RDSon in the MOSFET, a high enough current flow and extremely tight matching of the transistors making the comparator. IOW: it doesn't really work in a practicle situation. It is a waste of components as D5 and the fuse already provide a reverse polarity protection.
« Last Edit: January 26, 2025, 12:52:05 am by nctnico »
There are small lies, big lies and then there is what is on the screen of your oscilloscope.
 

Online langwadt

  • Super Contributor
  • ***
  • Posts: 4882
  • Country: dk
Re: Question: Why blocking diodes and not MOSFETs on solar panels?
« Reply #9 on: January 26, 2025, 01:27:33 am »
A MOSFET is not going to work as it conducts in both directions when 'on'. So when a shading condition stops, the MOSFET will short the cells it is connected across. IOW: you'd need a very complicated control circuit which isn't going to work better or more reliable compared to a 'dumb' diode.

https://robotics.ong.id.au/2014/07/30/raspberry-pi-b-power-protection-circuit/ ;)
That circuit only works reliable (in volume production, over a wide temperature range and with room for external interference) with a large amount of RDSon in the MOSFET, a high enough current flow and extremely tight matching of the transistors making the comparator. IOW: it doesn't really work in a practicle situation. It is a waste of components as D5 and the fuse already provide a reverse polarity protection.

it prevents back feeding the USB if you you power the 5V from somewhere else
 

Offline Siwastaja

  • Super Contributor
  • ***
  • Posts: 9432
  • Country: fi
Re: Question: Why blocking diodes and not MOSFETs on solar panels?
« Reply #10 on: January 26, 2025, 06:45:46 am »
A MOSFET is not going to work as it conducts in both directions when 'on'.

Ideal diode circuit of a MOSFET wires the MOSFET such that the body diode direction matches the desired current direction. Then a control circuit monitors the voltage across the MOSFET to emulate a diode, somehow measuring the current to keep the MOSFET on (bypassing the Vf=0.7V body diode) during forward current, but quickly turn the MOSFET off if the direction of current changes - just like a diode would do. Most obvious way to measure the direction of current is to measure Vds - to maintain a small voltage drop over the MOSFET - just like a diode has Vf. The more accurate the control circuit is, the lower this drop can be, but drop is still needed.

So now we have a complex, more expensive MOSFET circuit, which needs external supply voltage, and which still has Vf drop - maybe just 0.1V instead of 0.3V that of a schottky diode.

This drive complexity, requirement of external power supply, and minuscule gain to be had, are the reasons why this is not being done on solar panels.
« Last Edit: January 26, 2025, 06:50:32 am by Siwastaja »
 

Offline max_torque

  • Super Contributor
  • ***
  • Posts: 1331
  • Country: gb
    • bitdynamics
Re: Question: Why blocking diodes and not MOSFETs on solar panels?
« Reply #11 on: February 01, 2025, 12:12:09 pm »
Because "perfect" is the enemy of "good enough"!!!   :-DD

Today, with our modern tools, processes, simulation and manufacturing we really can just about do anything, so in most cases the limiting factor, or more accurately the level of "optimisation" required for any design is actually set primarily by economics and not physics.

So yes, you could make a solar panel with slightly lower loss, but would the customer benefit, or again, more accurately speaking, are they willing to pay for that benefit?
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf