0603 smd resistor - 0.1W

Hello Victor,

Don't go over the data sheet limits. Check the RP1608S-470-F from Susumu (Japan). 47ohms 0603 rated at 200mW, but verify with the data sheet. Digikey has them in stock.

You could also try to stack two 0603 of double value but production won't like that at all. Best to find one that works.

Don't worry about it too much. It will work, but it is highly recommended to change it as soon as you go into mass-production with it.

Victor schreef:

The important issue is the power rating. 0603 is usually rated at 100mW which (I believe) is the power dissipation at which the core reaches some elevated temperature, but I couldn't offhand tell you that temperature. The relevant spec is CECC 40401-802/EIA-575. (It's specified as power dissipation at Ta = 70C).

In your case, much depends on the accuracy required. If it's not too critical, simply place 2 x 0603 91 ohm (1 on the pads, one fitted above it - most build houses will do that provided you pay :). That yields, of course, a 200mW maximum power rating on 45.5 ohm which is a little lower, leading to more current perhaps (power rating somewhat derated because they are in thermal contact), but might work for you. Using two x 100 might be a better solution - slightly lower current and more easily available value.

Cheers

PeteS

Deponds.... chances are that if you have a good data sheet from the minufacturer then they will give you the PCB layout and conditions (inviromentally) at whoch their resistor can disspate 100mW plus they migt tell you how ling it will lost.

In ollden dais resistors all oprated at 3million C but itz got more complic now.

Anyway, put that one down to experience and go find some proper ditta. That might involve talking to application engineers (sales droids).

Last tim I luked at elektrolitik capacitors the hudline figur said 3A RMS max ripply and my ripply wuz 10A RMS.

Howev, the minuficturer had a thingy that let me put the ripply harminics into it and it said the liftim wus 89,000 hours insted ov 3 minits to explose. OV corse before I do it I would check wiv the minuficturer, liok wot they say you shud.

Bonus!!!!!!

DNA

It'll probably be OK, but longterm the value may drift a bit more than if it ran cooler.

One fix is to add copper pours to conduct heat out the end caps; thermal vias allow the pours to be on the bottom if the top is too crowded, or you can sometimes via the heat into a ground or power plane. Even fat traces help a lot. We're running some 0603's at about half a watt each, dumping the heat into pours. Turns out that 1206,

John

Do you have trouble soldering to such a heat sink?

I'd take a guess that the oven profile manages thing quite nicely.....

DNA

No problem at all with reflow oven... everything gets hot. Hand soldering is fine with a good iron, like a Metcal with a medium-hunky tip... if you can see the damned things.

John

How do you measure the hotspot temps? I typically use a small thermocouple with a tiny dab of thermal grease on it, but you have to think that for components 0603 size it must be conducting enough heat away to affect the results. Do you have a non-contact method?

We recently got a spiffy FLIR thermal imager. You can almost rub the lens up against a part and still focus, and that gives a bunch of pixels on an 0603. It's neat... spots studs in walls and such. I'm planning to try it on flowers next.

John

Hello John,

Hey, this could be cool to find where that darned hornet's nest is. Some of those burrow into the wood and unless you find the entry hole you just don't know where to look. We had one under an outside staircase and it took three cans of chem warfare to fight them, but still not sure if we got the hornets really good.

On Thu, 21 Sep 2006 15:56:47 -0700, John Larkin Gave us:

One can parallel two double value resistors (or series two half value)and get better thermal drift characteristics as well, but one should still observe your mentioned heat dissipation methods even in parallel or series set-ups. Same power and heat, but spreading it out yields a bit better transfer to both air and the thermally conductive trace/pcb elements (vias, etc.).

On Thu, 21 Sep 2006 21:46:19 -0700, John Larkin Gave us:

What brand did you buy?

FLIR!

John