Surge Capable Chip Resistors

1206's aren't bad. Some of the TO247 mosfets we destructicated made some serious shrapnel. Some of the "300 watt" parts exploded in 100 milliseconds at 300 watts, bolted to a copper block. We didn't use them.

John

My projects don't blow up anymore... I must be doing something wrong :) That or, I'm getting good at electronics... D from BC

No. You're just not testing them hard enough.

--
Gibbo

This email address isn\'t real.

I once saw a board with about a dozen TO-3s, and every one had a little crater where the die had been. :-)

Cheers! Rich

just use an MMA0204.

JL's right (he always is with thermal stuff) a 1206 should do, but an MMA0204 will give you bags of margin, and they fit in the same space as a 1206.

plenty mfgs make them.

Cheers Terry

I took a look at that on

MELF resistors!! ...learn something new everyday... I might have to balance my work bench so they don't roll away..

Also tripped on this while I was there.... Resistor pulse loading paper on:

Thanks... Always good to know alternatives... D from BC

I've seen TO-3's with holes blown open in the top of the can. Cool.

I wonder what it's like to bring up a megawatt hf transmitter or a big radar.

John

Yeah, a melf should have a lot more resistor element and a lot more mass to soak up joules. Good idea.

You can make a pad with a little slit in it to keep the tiny buggers from rolling.

John

Apparently, kinda dangerous. I used to hear stories, like the antenna guys would bring along a steel wool pad to throw in front of the antenna to see if it was safe to climb up on the structure. If the steel wool hits the ground, it's probably safe.

Another guy told me about a high-power installation where there was some rigid coax about 3" in diameter, with an elbow bolted into place. some guy said that the guy who had installed the elbow didn't torque the bolts down right (or left one uninstalled or something), and when they fired it up, the elbow disappeared.

Cheers! Rich

I agree, much better. There's the safety factor we'd like to see.

Nice idea!

I have both footprints, but I have never had a problem with hem rolling off the board, and i've used LOTS of them. They are just ideal for gate drive and current sense resistors, among other things.

the IRC ones are even more impressive, but hard to get and $$$

still not sufficient for a 3R3 damping resistor in series with a hefty cap in a 3-phase 400V EMI filter though :(

Cheers Terry

Marginal at best for an 0805; we shouldn't run the parts right up to the limit. I agree the 1206 data looks strange, no better than 0805. It'd be wise to play it safe and double up at least.

stumbled across this

maybe of some help

martin

We knew this mode as a "soft" failure. ;-) Moons ago there was an issue with phase-controlled regulators (couple thousand amps) where they'd let the magic smoke out at first power on. A "hard" failure took the entire TO-3 top off and around the room. Eventually they came up with an armored tub to test the regulators before installation in systems. The text fixtures all looked like they'd been through a war.

Poorly specified devices can get one in a lot of trouble.

Fingers in ears?

--
  Keith

I found it! In the big blue (former Tektronix) binder labeled THERMAL. It's full of all sorts of arcane stuff.

It's data for a 100 ohm 1206 thinfilm platinum RTD, which should be close to that of a standard 1206 resistor. The nice thing about the RTD is that you can apply power to it and measure the actual resistive element temperature in real time. I took this data to evaluate self-heating errors when this RTD was used as the cold junction ref for a thermocouple system.

In free air, with just tiny leads frugally soldered on to make connections, I got theta = 276 K/w and tau = 6 seconds. The temp-versus-time curve looks, visually at least, classicly exponential.

On a pc board that had modest thermal antennas, the actual application board, 69 K/W and 1.5 seconds.

On a 1" square hunk of copperclad (each side of the rtd getting 0.5 sq inch of copper), 35 K/W and roughly 500 milliseconds. Data is not so good here, as I was getting only a 6K temp rise in my setup. There's also, no doubt, a longer slow tail associated with the entire pcb warming up.

So with no heatsinking at all, a 20 millisecond, 1.3 watt zot will heat the resistance element to roughly 7 K above ambient!

John

On Tue, 20 Feb 2007 09:41:45 -0800, John Larkin wrote: [snip]

It's just amazing how a little chip resistor can get complicated.. Ohms law and P = I^2 just wasn't enough to know...

Looks like thumbs up for the 1206 again !! :) (When it's thumbs down, I'm checking how hot the part is and could burn my thumb. :) D from BC

Your power supply is under speced. :(

--
Service to my country? Been there, Done that, and I\'ve got my DD214 to
prove it.
Member of DAV #85.

Michael A. Terrell
Central Florida

That's a wise idea. You have no need to pass out riot gear to the test techs. ;-)

--
Service to my country? Been there, Done that, and I\'ve got my DD214 to
prove it.
Member of DAV #85.

Michael A. Terrell
Central Florida

Then you are not properly using the "Amtal rule", test it to destruction, preferably in several different ways. Then you know just how far you can afford to respect your warrantee.

--
 JosephKK
 Gegen dummheit kampfen die Gotter Selbst, vergebens.  
  --Schiller