Schottky Diode Datasheet Question?

[codex653]

ok so i found two diodes packages today on a circuit board with one being a B82M and the other a C82M...now i was looking at the datasheets for them, which i will post in just a second, and i noticed that there are 2 diodes in one package for each of them.
the rated current for the B82m is 5A, but is that 5A per diode or is it 5A if the diodes are paralleled?

here are the datasheets:
B82M: ESAB82M-006 datasheet pdf datenblatt - Fuji Electric - SCHOTTKY BARRIER DIODE ::: ALLDATASHEET :::

C82M: ESAC82M-004 datasheet pdf datenblatt - Fuji Electric - SCHOTTKY BARRIER DIODE ::: ALLDATASHEET :::

 

[MikeMl]

Good question. That is a crappy data sheet. Since this curve only goes 2.5A, I'm guessing that it is 2.5A per diode or 5A for both

 

[codex653]

hmmm yah you are probably right. i'm gonna parallel them since it wouldn't hurt to have a "10A" theoretical current rating if you happened to be wrong...however if you are right, which i believe you are, some catastrophic things could happen if i don't parallel them for my application! :/

 

[MrAl]

Here's the deal, with packages like that the rationale is as follows...

We take one diode and put it in a package with surface area S and allowable power dissipation P, and we take another diode the same in another package also the same. How much power can we get from the two diodes? Well, the surface area is S for both diodes and allowable power P for both and since they are separate packages we can get a total power of 2*P out of both of them.

Now take those same two diodes and cram them into only one package with surface area S. How much power can they dissipate now? Well, the surface area is the same, so we can only dissipate a total power of P watts again, even though there are two diodes in there. The power dissipation depends strongly on the surface area, not how many diodes are inside. Interestingly though, since we now have two diodes instead of one we have more options then we did before. It doesnt matter which diode we have running either, if we have D1 and D2 in there we can run either D1 or D2, or run D1 for 50 percent of the time and D2 for the other 50 percent of the time, as long as we stay within the limit of P watts for the entire package.

Now out there there are plenty of applications where we want to full wave rectify a waveform to get DC from AC or from a square wave. We can use two diodes connecting say the two cathodes together, and since each diode will only be on for 50 percent of the time we can use diodes that are smaller than if one single diode had to handle the entire amount of current.

Now enter in the application specific part: the dual diode. Knowing beforehand that designers are going to want to full wave rectify waves to get DC, we want to manufacture a part that has two diodes built in and the two cathodes are already connected together. Knowing the current requirement we can select two diodes that are the same and can handle the entire current, but since they only have to be on for 50 percent of the time each we can stuff them into one package with the same surface area required for the total power dissipation. In other words, one diode with surface area S handling power P or two diodes with surface area S handling power P where only one diode is on at any given time. That's the part we are talking about here.

Now for the rating...how do we rate this thing? Again, knowing beforehand what it will be used for, we can rate it for the end application itself rather than either diode alone. That's one way to do it knowing designers will be looking for this kind of thing anyway. Ok, so we found that we can achieve 5 amps average rectified current when the duty cycle is 50 percent for each diode, so we rate it at 5 amps. But then the question might come up about just what that means for each diode. The answer again is that since there are two diodes sharing one unit surface area, we can run either diode up to 5 amps each, but because of the package restriction we can not run both of them at the same time unless we derate based on the package size. Derating requires consideration of the actual voltage at reduced current and just how a package heats up with internal power being dissipated, but a good estimate with some safety factor is one half for each diode.

To sum up:
We can run EITHER diode up to 5 amps continuous, or, we can run BOTH diodes up to 5 amps if we limit the time they are on to a total of 100 percent of the period divided between both diodes (ie 50 percent for one and 50 percent for the other for example). If we run both diodes continuously though, we have to derate them to approximately 1/2 the total current which is 2.5 amps in this case, which means both diodes are conducting all the time but only 2.5 amps each. There are a host of other possibilities, but the 50/50 scenario is the most common.

What does not work:
Paralleling both diodes does not allow more than 5 amps total because if both diodes are conducting then we have to default to the above where we derate them each to 2.5 amps. That means that one diode conducting 5 amps is the same as both diodes conducting 2.5 amps each, except if we only run one diode we dont have the problem of current sharing to consider. However, considering that if the current is limited to 5 amps through the entire package (both diodes) then that means that even if they dont share the current very well (one diode conducts 4 amps and the other 1 amp) as long as we know the total can only be 5 amps and we keep it like that everything should work out just fine.

In closing, i have to say that there is also the chance that they cheaped out on the diodes and used under rated dies.

 

[codex653]

WOW that is a loooonnnggg answer! but thank you for clarifying! i understand better now

 

[MrAl]

Hello again,

He he, that was an interesting subject so i couldnt resist adding comments. Sort of like trying to run two resistors in a one resistor package. Two resistors in two packages gives us twice the power capability as one resistor in one package, but two resistors in a one resistor package still only gives us the one resistor power rating.

 

[codex653]

yah i completely get it now lol hey but thanks again for posting that much! i know that must have taken some time, so i appreciate you answering me like that thank you!

 

[MrAl]

yah i completely get it now lol hey but thanks again for posting that much! i know that must have taken some time, so i appreciate you answering me like that thank you!

Oh hey you are very welcome, and have a Merry Christmas too.

 

[codex653]

thanks! you have a Merry Christmas also!!