0

I'm thinking to have 10 pcs 10amp rectifier Diode in parallel like in my drawing so I can have 100amp diode for 14.v 90amp DC input voltage. I tested the voltage drop is minimal from actual and simulation. Is this ok? please help

E B
  • 1
  • 3
  • 1
    This is very much like the typical "parallel LED" questions that show up here. The biggest difference is that you've got 100A for one of the diodes to hog, and then enthusiastically release. I think this will not work, but there may be some properties of larger diodes I'm not aware of that may make it workable. So, just a comment - but I'd be really wary of trying it. I'm no fan of exploding diodes. – JRE Aug 24 '16 at 11:43
  • Why do you think you need the diodes? What are you trying to keep away from your power source? – JRE Aug 24 '16 at 12:12
  • do you think a 10amp rectifier diode in series can handle the 90amp input that will pass thru it? – E B Aug 24 '16 at 12:22
  • A 10A diode will release the magic smoke if you push 90A through it. The question is really, how much current will the load draw? – JRE Aug 24 '16 at 12:24
  • Just because the charger can deliver 90A, that doesn't mean that your load will actually draw 90A. It may use only a few Amperes, but we can't tell because you haven't said what the load really is. – JRE Aug 24 '16 at 12:27
  • i added additional picts to what i'm trying to do – E B Aug 24 '16 at 12:41
  • 2000Watts at 12 volts is over 160A. Your battery charger can't deliver all that the inverter may try to draw. But, it still depends on the load. What is running off the inverter? A small radio, or a thousand Watts of lights? – JRE Aug 24 '16 at 12:45
  • the load is 950 watts 1 pc, 1 printer, 1 monitor, and 2 switches – E B Aug 24 '16 at 12:50

3 Answers3

0

It's not a great idea, there's nothing to guarantee each diode will do 1/10th of the work, and in fact that is quite unlikely. You could balance it better by adding a small resistor in series with each diode, but then you'd be dropping more voltage and making more heat.

Colin
  • 4,499
  • 2
  • 19
  • 33
  • hi Collin I can't afford to drop the voltage due to resistors in series. Do you think if I'll use T0-220 at 40amps and parallel 3 of them so I can have 120amps will work? – E B Aug 24 '16 at 11:52
  • That will suffer from the same problem. – Colin Aug 24 '16 at 11:54
  • how about if I will not parallel anymore additional diodes in 90amps 14.4volts instead series a single 10amps rectifier diode will it heat up? or destroy the diode? – E B Aug 24 '16 at 12:00
0

This is not a good idea, due to current hogging.

It is extremely unlikely (as both Colin and JRE have noted) that the forward voltage of all the diodes will be the same; that implies that one of them will have a slightly lower forward voltage than the rest.

Now we have a diode with a large amount of current and that is heating it up, but as it heats up the forward voltage will reduce, further increasing the amount of current through the device (it is now hogging current significantly), causing further heating and so forth until it gets so hot that it fails.

This nasty situation is known as thermal runaway and will very possibly make the victim part become spectacularly pyrotechnic; then the next lowest \$V_f\$ part will be subject to the same treatment until the remaining parts cannot handle the load and fail anyway.

Why do you need the diodes? To catch inductive kick or are you rectifying a source?

Peter Smith
  • 21,923
  • 1
  • 29
  • 64
  • hi Peter, I have this input from the battery charger of having 14.4v 90amp continues charging. I have a requirement that i need to series the diode in the positive of the charger before going to the positive line of the power inverter. I can't find in my local electronics shop a 120amp rectifier diode.Do you think i really need 120amp? what's available here is 10amp only. – E B Aug 24 '16 at 12:17
  • @EB: What power inverter? The diodes need to match the expected draw, not what the charger can deliver. If the inverter only needs 10A, then you don't need a 90A diode. – JRE Aug 24 '16 at 12:23
  • i added additional picts to what i'm trying to do – E B Aug 24 '16 at 12:41
0

Paralleling diodes: Definitevily not a great idea but can be done with good thermal coupling, wise wiring or pcb layout to help current share and, of course, allowing for some derate to single diode specification.

This is true since one-diode current positive temperaure coefficient leads to run-away only if overall loop again is above unity. Also positive diode series resistance temperature coefficient helps a while.

This AN from ST may help going deeper.

Let's put it down simple: take no math and go for some 20 off 10A diodes paralleled to carry 90A.

carloc
  • 3,174
  • 1
  • 16
  • 18