connecting 2 rectifiers

[jamesjung21]

Hi,
I've got a question on rectifier connections...

I've got 25 pieces of BY229 Fast Rectifier Diodes and wanna use them for a big powersupply.

Each bridge (i.e. 4x BY229) needs to handle 6A , but I don't think their current capability is safe enough for this application.

So I wondered what if I could possibly connect 2 bridges (i.e. 8x BY229) in parallel, to increase (double?) their current capability.

Can anyone tell me what if this would work properly?

And as a reference, here's a brief data on BY229 Fast Rectifier Diodes:

Vr= 800V
If(AV)= 8A
Ifsm= < or = 60A
Package= TO220AC

Features-
*Low forward volt drop
*Fast switching
*Soft recovery characteristic
*High thermal cycling performance
*Low thermal resistance

 

[Hybrid fourdoor]

Haha...you checked "did you search" and you didn't!

I posted same message couple weeks ago...its actually still on the front page of this forum

My original thread

 

[jan.didden]

Hi,

You say you want a 6 amps supply. Does that mean continuous, 100% load of 6 amps all the time? Probably not (but I can be wrong). Since these diodes are 8 amp rated, a 6 amp supply with say 50% or (in case of a power amp) 10% duty cycle is no problem.

But to your question: if you parallel the bridges, you in fact replace each diode by two parallel diodes (draw it out). How do you know that in each pair they share the current? You don't, really. One diode may have a slightly higher forward voltage drop under load than its parallel mate, so it would probably carry 80% or more of the current. Now, if the diodes are from the same batch, forward drop MAY be similar and sharing could be reasonable.

An other way would be to use small resistors in series with each diode (.1 Ohms or so) to force equal forward drop and thus equal current sharing. But it gets rather involved for something I have a feeling is a non-problem.

Jan Didden

 

[millwood]

doesn't sound like you need to parallel those diodes. diodes are usually very rugged and yours is rated at 8amp continues (at 100 degrees to boot) and 100amp pulsed. Unless you have some unusal applications, they should handle 6amp.

If you really want to parrallel, you can put a small resistor to "degenerate" out any Vf differences between the diodes.

One diode may have a slightly higher forward voltage drop under load than its parallel mate, so it would probably carry 80% or more of the current.

Jan Didden

the diode with higher forward voltage drop will carry less current, not more.

 

[jan.didden]

doesn't sound like you need to parallel those diodes. diodes are usually very rugged and yours is rated at 8amp continues (at 100 degrees to boot) and 100amp pulsed. Unless you have some unusal applications, they should handle 6amp.

If you really want to parrallel, you can put a small resistor to "degenerate" out any Vf differences between the diodes.




the diode with higher forward voltage drop will carry less current, not more.

Yes, indeed, I meant the MATE would carry most of it. Sloppy grammar. Sorry.

And BTW, I DID say that a series resistor takes care of Vf delta's. Is it just a coincidence that you point out errors you believe I make, but almost verbatim repeat things I say that you agree with to make them look your statement??


Jan Didden

 

[jamesjung21]

Consider a failure condition

But, are they strong enough to handle an enormous current (let's say 15~20A) for a resonable time, when a failure occurs and strangely fuses don't blowout?

If they are strong enough to handle that much, they wouldn't break down, but get hot if it really occurs.

I need to tell you that 6A is a 'maximum RMS current' and 8A would be the 'peak current.'

 

[jamesjung21]

And as an addition, I don't really want the diodes to blow.

Each diode is quite expensive. Each is almost as expensive as a MJ15003/MJ15004, and I need 8 to make a twin mono block.

 

[millwood]

Re: Consider a failure condition

But, are they strong enough to handle an enormous current (let's say 15~20A) for a resonable time, when a failure occurs and strangely fuses don't blowout?

I think all your questions can be answered by the datasheet, specifically on the last page of the datasheet, .

It is just a few clicks away.

 

[Magura]

And as an addition, I don't really want the diodes to blow.

Each diode is quite expensive. Each is almost as expensive as a MJ15003/MJ15004, and I need 8 to make a twin mono block.

Its a very expensive approach to rectifiing, to build a bridge of single diodes, compared to buying a bridge.

I usually go extreeme overkill when i make a bridge of single diodes. I use the ON Semiconductor MBR6045WT schottky's for most applications. They work well in noise matters and run a little cooler than the regular diodes since theyre schottky's. They are around 8-10$ a piece, but in exchange they live pretty much for ever/- take any punishment you could accidentaly give them.

You may as well get used to the idea that in this game you dont go close to the limits unless you have to.



Magura

 

[millwood]

And BTW, I DID say that a series resistor takes care of Vf delta's. Is it just a coincidence that you point out errors you believe I make, but almost verbatim repeat things I say that you agree with to make them look your statement??


Jan Didden

because I read your very quickly and skipped the resistor part. If pointing our your errors and not repeating your correct part offends, it is too bad.

I really didn't have your feelings at the top of my priorities. trying to help this person asking the original question was. If that ****ed you off, too bad.

 

[JOE DIRT®]

Easy guys,there only bloody diodes

Bottom line is to choose a diode spec`ed out to your design requirements......running diodes in parallel is asking for trouble even with current sharing resistors...overall cost and board space will be reduced if you use the right part to begin with.


DIRT®

 

[jamesjung21]

Thanks everyone !