Resistor dissipation vs. pulse current

For a pulsed-power application, that puts maybe 100A through a 5m-ohm resistor for several tens of milliseconds, how in heck do you figure out the resistor size?

I'm looking at data sheets, and I'm seeing power ratings but no sort of pulse energy rating. Is this because it just ain't done, because the resistors I'm looking at aren't hoity-toity enough, or because it's just something that you know?

At any rate, it's something like one Joule every ten seconds, so a 1/10th watt resistor should be just fine -- right? Or should I try to estimate the heat capacity of the part, and get myself all confused by thermodynamics?

--
www.wescottdesign.com
Reply to
Tim Wescott
Loading thread data ...

Yes ;-) ...Jim Thompson

--
                  [On the Road, in New York]

| James E.Thompson, CTO                            |    mens     |
 Click to see the full signature
Reply to
Jim Thompson

You're so @#$% helpful.

--
www.wescottdesign.com
Reply to
Tim Wescott

e

some have datasheets that include pulse/surge rating, something like this:

formatting link

looks like 100A in ~30ms, would need something like three resistors of that kind in 2512 package

-Lasse

Reply to
langwadt

I know O:-)

The only way I know to attack/assess your problem is to blow up some resistors.

It's going to depend not only on mass, but cooling paths on the board mounting, not exactly easy to estimate. ...Jim Thompson

--
                  [On the Road, in New York]

| James E.Thompson, CTO                            |    mens     |
 Click to see the full signature
Reply to
Jim Thompson

;D

Look at Ohmite's Pulseater resistors for an idea of pulse rated resistors. It's all in how thick the film is.

Cheers

Reply to
Martin Riddle

e

=A0 =A0 =A0...Jim Thompson

Wirewound ceramics can take mighty big pulses. I spec'd some recently, and there was pulse dissipation info from the mfr. Ohmite, IIRC.

-- Cheers, James Arthur

Reply to
dagmargoodboat

--
Regards, Joerg

http://www.analogconsultants.com/
Reply to
Joerg

The only proper way if this needs to go through any sort of serious certification testing is to get something from the manufacturer. In writing. Data such as this:

formatting link

Going by some thermal calcs alone is a risky path. Something in there could give up under enough current concentration and then *PHOOMP*

Unfortunately large pulse-power rated resistors are usually in the hoity-toity hi-$$$ class.

--
Regards, Joerg

http://www.analogconsultants.com/
Reply to
Joerg

e

We have had some problems in the past with voltage arcing across the resistor when we pulsed them but had met the average power rating. I think if you focus on average power and peak voltage ratings you will be fine.

Reply to
brent

That's only 50W.

1/10th watt means SMT? No, that's not fine.

Why not reduce the resistance? Going with 1 milliohm @ 100A gives you

100mV of signal, and cuts dissipation by a factor of 5.

You'll need about a 2W resistor, minimum. 4-lead connection if accuracy matters, otherwise I get a 20% error using a Ohmite #FCSL64R001JER (2-leads).

That part's rated for 125A x 10mS surges, or 225W absolute max instantaneous power, if that gives you any idea.

3W, or a physically larger unit may be a better choice (more margin). Depends on your surge.

-- Cheers, James Arthur

Reply to
dagmargoodboat

t
t

th

ate

Nice part. Multiple units, if used, should be paralleled, to minimize connection resistance error.

-- Cheers, James Arthur

Reply to
dagmargoodboat

e

Or you could ball park an estimate. Guess the mass of the resistive element. Calculate the power disipated during one pulse. Make a guess at what the resistive element is and look up the heat capacity. Now calculate how much temperature rise there would be. This neglects any heat transfer during the pulse, but there is not much time for any significant heat transfer. Now you can make an intelligent estimate of whether the resistor will survive the pulse. You should be able to say " this resistor has not got a chance " or " this resistor might work " or " this resistor should have no problem ".

And finally do a few tests to verify your estimate.

Dan

Reply to
dcaster

If thickness is the only requirement, Dimmie could stand an infinite pulse. ;-)

--
You can't have a sense of humor, if you have no sense.
Reply to
Michael A. Terrell

A _gross_ estimate would analyze the energy all going into the mass of the resistor. That would certainly be conservative ;-) ...Jim Thompson

--
                  [On the Road, in New York]

| James E.Thompson, CTO                            |    mens     |
 Click to see the full signature
Reply to
Jim Thompson

I just use the standard and go from there.

1 Watt = 1 Joule per Second;

Jamie

Reply to
Jamie

Data point: 65 watts will blow up a regular 1206 thickfilm resistor in roughly one millisecond.

John

Reply to
John Larkin

If it's a composite resistor, like metal film over ceramic, the thermal math would be very difficult.

A regular 1206 thickfilm resistor will explode at 65 watts in about 1 millisecond.

A solid manganin resistor, which some milliohm shunts are, is easier to calculate: joules into mass of known specific heat. But a 1206 size or so solid manganin resistor wouldn't have any trouble absorbing 50 watts for a few 10s of milliseconds. It's hard to vaporize solid manganin.

Big copper pours really help keep a resistor cool, but aren't as effective as pulse widths go down. The tau of a 1206 to big pours is in the 2 second ballpark.

If it were my problem, I'd buy a few various 5m resistors, solder onto a sorta-real pcb copper pattern, and blow them up, to see what sort of margin they would have. Fun!

You could just use a pcb trace as a shunt, if you can calibrate it and don't mind the biggish TCR.

John

Reply to
John Larkin

Only if the resistor will be isothermal, namely if it's a solid mass of a single material, like a chunk of manganin or carburundum or some such. A metal film on ceramic resistor is much harder to analyze.

John

Reply to
John Larkin

ut

RESISTANCE IS FUTILE.

James

Reply to
dagmargoodboat

ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.