Diodes in parallel to increase maximum current ?

Yes for making fireworks. They will pop one by one.

Thats what i feared.. :-)

Thanks for the very quick reply..

--
email hihihi wanadoo nl

Well, you could put them in parallel, but it wouldn't serve your purpose. The reason is that each diode will likely have a different Vf. The diode with the lowest Vf will conduct first, possibly not allowing the others to turn on. The diode with the lowest Vf will hog current, possibly to the point of destruction, probably shorting, thus creating havoc with whatever lies downline. If it fails open, then the diode with the next lowest Vf conducts and repeats the cycle. A losing situation either way. Solution is to get the correct diode.

Cheers!!!

--
Dave M
MasonDG44 at comcast dot net  (Just substitute the appropriate characters in 
the address)

Never take a laxative and a sleeping pill at the same time!!

hmm, not really, i assume your referring to things like supply line mains etc. if you get perfectly matched diodes you could do it how ever, the chance of getting 4 diodes to match the current curve is slim with in reason.

--
Real Programmers Do things like this.
http://webpages.charter.net/jamie_5

"hihihi"

** Yes - it will likely work out just fine.

The ***f****it doomsayers** here have simply never tried it.

** To get reasonable current sharing, the diodes should all be from the same batch and mounted in thermal contact.

This means installing them snugly side by side, with short leads into large heat-sinking pads on a PCB.

Or else - twisting the four leads together at each end.

Long as the diodes are similar and the temps the same, close sharing is assured.

However, four 3 amp diodes will not give you a 12 amp average current rectifier.

Derating will be needed - to maybe 7 or 8 amps.

........ Phil

If there is something else in each diode's circuit to ensure that they share the current approximately equally, then it is sometimes possible.

You can use a resistor if the power loss isn't going to be a problem, or you might be able to make use of some existing loss, such as the winding resistance of a transformer.

If you were using a mains transformer with four identical parallel secondary windings, keeping the windings separate with one diode on each winding would probably work - but you would have to test it thoroughly and de-rate the diodes sufficiently to allow for any imbalance. Similarly, if you are using a choke-input smoothing arrangement, a quadrifilar choke could be designed to do the job.

I have used something like this in reverse, to distribute output transistor current in a large audio amplifier with a multi-primary output transformer.

In all probability, none of these solutions will be possible for your particular problem - so the answer will be "no".

--
~ Adrian Tuddenham ~
(Remove the ".invalid"s and add ".co.uk" to reply)
www.poppyrecords.co.uk

"Adrian Tuddenham"

" In all probability, none of these solutions will be possible for your particular problem - so the answer will be "no". "

** Fuck off - you bloody idiot.

......... Phil

Well, Yes, No, and maybe.

You can't put them directly in parallel, as the one with the lowest voltage drop will hog the current, causing it to heat up. Unfortunately diodes have a several millivolt negative TC, so when it warms up it will hog more current, leading to more heat, more current,.... I think you see where this is going.

You can definitely get this to work IF you add enough series resistance to EACH diode individually to compensate for the negative TC.

"Enough" resistance is beyond my ability to compute, but a wild guess would be enough resistance to equal the diode voltage drop.

Now this isnt too bad if we're talking about a high voltage power supply where another 0.7 volts won't matter. It is bad for a under 10 volt PS where anotehr 0.6 volts is a big deal.

--- "Likely" isn't the same as an unequivocal "yes", which is what I think the OP was looking for.

In any case, you've given him bad advice.

Connecting diodes in parallel is, at best, very poor practice, and what you should have advocated was that he simply buy diodes specified to carry the current that he needs.

-- John Fields Professional Circuit Designer

"John Fields" schreef in bericht news: snipped-for-privacy@4ax.com...

But did you try it?

--
Thanks, Frank.
(remove \'q\' and \'.invalid\' when replying by email)

--
Of course not. 

Neither have I jumped off a cliff to prove that it\'s fatal.

"John Fields" schreef in bericht news: snipped-for-privacy@4ax.com...

Depends how high it actually is, and what's at the bottom.

It says "No thermal runaway".

I never tried putting diodes in parallel, because I was once told not to do so. OTH, keeping them close together, and thermally coupled by wide copper tracks, it may well be okay to do so.

But you have a large wall full of raaco drawers. I bet you have a couple of these 1N5408 in one of those. If you try it, then I'll donate $10 in the relay pot.

--
Thanks, Frank.
(remove \'q\' and \'.invalid\' when replying by email)

So, according to your "derating", to get a 4x current fold, you propose the OP to use at least 7 diodes from the same batch in parallel (7*(7/12)=4.0833>=4x) and keep all 7 in thermal contact.

How is all that easier, better or even cheaper than getting the correct diode? Also, if the OP has problems in getting the correct diode, imagine the problems he'll have in telling whether the 7 diodes are from the same batch :-)

[snip]

We all know who the "f****it" is, don't we ?:-)

...Jim Thompson

-- | James E.Thompson, P.E. | mens | | Analog Innovations, Inc. | et | | Analog/Mixed-Signal ASIC's and Discrete Systems | manus | | Phoenix, Arizona Voice:(480)460-2350 | | | E-mail Address at Website Fax:(480)460-2142 | Brass Rat | |

| 1962 | It's what you learn, after you know it all, that counts.

"Jim Thompson" schreef in bericht news: snipped-for-privacy@4ax.com...

I have no idea. However, I do know who the pile of shit is.

--
Thanks, Frank.
(remove \'q\' and \'.invalid\' when replying by email)

--
I don\'t have any, but even if I did, I wouldn\'t try it.  To do it
right I\'d have to make sure they were all in good thermal contact
with each other and measure the currents through each of them with
several different currents in the circuit to tell what was going on.
A huge PITA at this point, and just not worth it, IMO, but thanks
for the offer.  Maybe you\'ve got some and want to try it???

Yep, thats the way i am going to solve the problem.

--
email hihihi wanadoo nl

"John Fields" schreef in bericht news: snipped-for-privacy@4ax.com...

I'll be honest with you - all of a sudden I've run out of 1N5408 as well! Perhaps Winfield has some, he also has all the proper gear to do the measurements ;)

--
Thanks, Frank.
(remove \'q\' and \'.invalid\' when replying by email)

Actually (being pedandic) - paralleling with no resistance is possible using Silicon Carbide diodes (CSD04060 from Cree). They will settle with current sharing by nature

Regards

Klaus