Parallel Diodes

[clancy]

I see comment about parallel diodes, but not understand if it is joke or serious?

But it interest me so I do testings.

Found that if diodes (same type and manufacturer) are at same temperature then current shared very good. This happen if diodes not get hot, or, if they on same heatsink and get hot.

But if diodes not thermally connect, and diodes get hot (>40C > ambient), then current sharing bad, one conduct more and get hotter than other.

Theory says Io increase with increase temp in diode equation, but series resistance increase also. Net effect be less voltage for same current when diode get hotter.

[golfnut]

This is a popular conundrum. The doping in each diode will be different. So one diode WILL have more charge carriers available at a given temperature. One WILL conduct slightly more easily than the other. The One that goes first will hog the current and will get hotter. This acts to lower its resistance more - ensuring it gets the current. The closest you can get to 2 the same are special matched pairs fab'd on the same die.. S

[Paul99]

I see comment about parallel diodes, but not understand if it is joke or serious?

If you are referring to the comments pushed by Doug of the AEGPF forum, Doug uses denigration to make up for the fact he has a monumental inferiority complex. I am the bunt of his demented humour.

But it interest me so I do testings.

That's the way to go. Always prove things for yourself rather than accept what others say. As an example, the issue of parallel diodes arose because Bruce Candy used parallel diodes in the front end of the SD2000 metal detector. Moodz, a member of this forum, made the statement that the schematic was incorrect, and that one of the diodes needed to be reversed. Reversing one of the parallel pair diodes would have resulted in short circuiting of components with resultant catastrophic component failure.

Don't automatically assume that there are no applications where using diodes in parallel will give a benefit. It may be that the characteristics of 2 diodes in parallel give benefits that are not related to max current capability. An example of paralleling diodes to achieve design benefits not achievable with a single component can be seen in the paper "SIC technology, a way to improve aerospace inverter efficiency, VIEILLARD Sébastien, Hispano-Suiza, Safran Power Division, SAFRAN Group".... from that paper...


"To well understand and compare the technology, the switching results are compared with 2 other power modules that are almost identical (Package, layout, IGBT dies …), the only difference is that the first one has the Emcon diodes recommended with the IGBT die used and the second one has 2 SiC Schottky diodes in parallel."

"Indeed, two diodes are used in parallel to reduce the drop voltage that can degrade the overall efficiency of the inverter and to mitigate in the same way the cost of an expensive SiC die having a too big surface."


So, there you have it folks. Polarity aligned parallel diodes are not necessarily a no-no.

[PATCHES JUNIOR]

AZIZ, here is your chance to show what you and Doug think about parallel diodes use. I have read your comments on them being used in detectors before. But will let you say your view since you claim to know it all. I just wish I did.

[Carl-NC]

I've used parallel diodes many times in designing integrated circuits. I'll leave it to parallel diode skeptics to determine why I would do such a thing, and conclude by saying if the civility in this thread deteriorates any further there will be some denizens taking a vacation.

[Davor]

Too bad, the party just seemed to be started rollin'

[Aziz]

Load balancing on parallel diodes requires a series resistor for each diode. The series resistor is a negative feedback to the load balancing loop.
*LOL*
4212

[moodz]

...my original comment was nothing to do with incorrect use of diodes ...it was about the use of a diode and a slight modification to achieve ultrafast damping ...if the designer had only made the same "mistake" that I made in a similar cct they would have discovered the secret of fast coil damping .....at the time I could not say any more because the patent was not in place.

[Paul99]

...my original comment was nothing to do with incorrect use of diodes ...it was about the use of a diode and a slight modification to achieve ultrafast damping ...if the designer had only made the same "mistake" that I made in a similar cct they would have discovered the secret of fast coil damping .....at the time I could not say any more because the patent was not in place.

Your initial comment was specifically about the use of parallel diodes. Your comment was made to support Doug in his attempt to ridicule me. I had stated the SD2000 schematic displayed Candy using parallel diodes in the front end of the SD2000 and had asked Doug what explanation he could give for Candy doing so. You replied to my post stating the schematic was incorrect as one of the polarity aligned parallel diodes was oriented incorrectly. Simple as that Moodz. Mechanic (Mick) looked at the circuit and determined correctly that doing what you stated to be correct would result in a short through a fet anti-parallel diode. To take the egg off your face you later commented in a post that you said what you did to see if I would be stupid enough to reverse one of the parallel diodes in my detector and blow my detector up. If you would like your memory refreshed, refer your posts in the thread on Doug's AEGPF forum. Don't lie and do it at my expense, Moodz, show a bit of ethics even if it is going to make a fool if yourself.

[moodz]

...believe what you want ...this forum is for discussion about technical topics ... not your incorrect interpretations regarding comments made on other forums.

[Aziz]

Ok, a nice parallel diodes joke for you:

Why do you use parallel diodes?
Well, if one gets bzzzzzzt (smoke, kaput), you have another one in the cct! *LOL*

4212
(I hope, I don't get onto vacation now.)

[Paul99]

...believe what you want ...this forum is for discussion about technical topics ... not your incorrect interpretations regarding comments made on other forums.

We are not talking what I want to believe or how you want to deceive, Moodz. It is obvious to anyone reading the relevant thread that there was nothing incorrect in my interpretation of what you were saying. The facts are obvious. Would you like a couple of screen grabs to refresh your memory?

[clancy]

Load balancing on parallel diodes requires a series resistor for each diode. The series resistor is a negative feedback to the load balancing loop.
*LOL*
4212

This not true Aziz; yourself need to measure real life what happen.

I use DC current probe and thick short wires connecting diodes, so no extra series resistors, and I testings many many diodes. Always current sharing very good for same type and manufacturer. Only except if diodes get much hotter than ambient AND not on same heatsink; only then current sharing bad.

[Aziz]

This not true Aziz; yourself need to measure real life what happen.

I use DC current probe and thick short wires connecting diodes, so no extra series resistors, and I testings many many diodes. Always current sharing very good for same type and manufacturer. Only except if diodes get much hotter than ambient AND not on same heatsink; only then current sharing bad.

Do yourself a favour Clancy and look here:
https://www.google.de/#q=load+balancing+parallel+diodes

4212

[ivconic]

People you are not real, indeed!
Whole a lot topic you are pulling each others leg for complete trivial case!?
Even Carl stepped here with "my bigger than your's" spite attitude!?
Who gives a $hit about parallel diodes actually?!
What are you talking about here!?
Is such huge crisis and lack of ideas so you turned to childish spite!?
And i thought i was the forum's "denis" here!?