diy solar

diy solar

Bypass/Blocking Diode Alternative

KBWaldron

Solar Enthusiast
Joined
Nov 14, 2019
Messages
230
Location
Hamilton, Ontario, Canada
For those of you using Blocking Diodes you know that there is a constant power loss from them. Unlike Bypass Diodes, Blocking Diodes are under most circumstance forward biased, which means they introduce a .4 volt drop in each circuit. If you are pumping a lot of amps through your panels then this can represent a not insignificant power loss, and more importantly a heat source. In my case 32A x .4V = 12.8W or about .5% of my power. The half percent I probably wouldn't worry about but the heat is more of a concern.

Similarly, pretty much all panels manufactured today have bypass diodes in them. These are normally reverse biased and so don't normally impact power generation. However, when one or more cells in the string they are protecting on the panel are either in shade or damaged then the voltage on that string of panels becomes reversed and the diode begins to conduct. This is to protect the cells in the string as they would begin to generate heat otherwise. Over time, particularly in panels where the quality of these diodes is reduced they can eventually fail. The power loss when these diodes are conducting can be as much as a few percent and generates heat.

TI makes the SM74611, what they call a smart bypass diode. Based on the specification this looks like it could also be used in place of blocking diode, which is what I am more concerned with.
Have a look see if you are interested. These are parasitic devices which when forward bias consume a very small amount of power and reduce the voltage drop across themselves to virtually zero, through the use of a MOSFET.
Cost of these is about $3.50US each, so I'm really only interested in them in my junction box where they will be acting as blocking diodes.
 
Oh well, just looked at the data sheet more closely. Maximum reverse voltage is 30. Not going to fly for anyone using solar as we tend to like to get the voltage up, and Amps down for transfer to the SCC. My voltage is going to be in the 60volt range which clearly is more than this device is willing to block.
Fine for bypass diode replacement, but not blocking diode replacement.
 
For those of you using Blocking Diodes you know that there is a constant power loss from them. Unlike Bypass Diodes, Blocking Diodes are under most circumstance forward biased, which means they introduce a .4 volt drop in each circuit. If you are pumping a lot of amps through your panels then this can represent a not insignificant power loss, and more importantly a heat source. In my case 32A x .4V = 12.8W or about .5% of my power. The half percent I probably wouldn't worry about but the heat is more of a concern.

Similarly, pretty much all panels manufactured today have bypass diodes in them. These are normally reverse biased and so don't normally impact power generation. However, when one or more cells in the string they are protecting on the panel are either in shade or damaged then the voltage on that string of panels becomes reversed and the diode begins to conduct. This is to protect the cells in the string as they would begin to generate heat otherwise. Over time, particularly in panels where the quality of these diodes is reduced they can eventually fail. The power loss when these diodes are conducting can be as much as a few percent and generates heat.

TI makes the SM74611, what they call a smart bypass diode. Based on the specification this looks like it could also be used in place of blocking diode, which is what I am more concerned with.
Have a look see if you are interested. These are parasitic devices which when forward bias consume a very small amount of power and reduce the voltage drop across themselves to virtually zero, through the use of a MOSFET.
Cost of these is about $3.50US each, so I'm really only interested in them in my junction box where they will be acting as blocking diodes.
For those of you using Blocking Diodes you know that there is a constant power loss from them. Unlike Bypass Diodes, Blocking Diodes are under most circumstance forward biased, which means they introduce a .4 volt drop in each circuit. If you are pumping a lot of amps through your panels then this can represent a not insignificant power loss, and more importantly a heat source. In my case 32A x .4V = 12.8W or about .5% of my power. The half percent I probably wouldn't worry about but the heat is more of a concern.

Similarly, pretty much all panels manufactured today have bypass diodes in them. These are normally reverse biased and so don't normally impact power generation. However, when one or more cells in the string they are protecting on the panel are either in shade or damaged then the voltage on that string of panels becomes reversed and the diode begins to conduct. This is to protect the cells in the string as they would begin to generate heat otherwise. Over time, particularly in panels where the quality of these diodes is reduced they can eventually fail. The power loss when these diodes are conducting can be as much as a few percent and generates heat.

TI makes the SM74611, what they call a smart bypass diode. Based on the specification this looks like it could also be used in place of blocking diode, which is what I am more concerned with.
Have a look see if you are interested. These are parasitic devices which when forward bias consume a very small amount of power and reduce the voltage drop across themselves to virtually zero, through the use of a MOSFET.
Cost of these is about $3.50US each, so I'm really only interested in them in my junction box where they will be acting as blocking diodes.
 
I really enjoyed this article. In my application I am using shottky diodes to isolate LifPO4 batteries and my current draw can be as high as 100amps.
Could a person parallel 6 of these to increase my current demand?
 
I really enjoyed this article. In my application I am using shottky diodes to isolate LifPO4 batteries and my current draw can be as high as 100amps.
Could a person parallel 6 of these to increase my current demand?
Yes, you could parallel them. I’d leave at least a 20% allowance. Just make sure you don’t exceed the voltage rating. There is no way to combine them to get a higher voltage rating, at least that I can determine.
 
That sounds like a plan. I would really like to get rid of my voltage drop and heat produced. I have been using a SMC 208CMQ060. My voltage doesn't exceed 15 volts and the current is usually around 25 Amps. I would like to build in lots of headroom though. I will begin a board layout and share what discoveries I find.

 
Back
Top