Adhesive heatsink ceramic resistor

Check out the temperature limits of the adhesive. Ceramic resistors are rated for much higher core temperatures then semi-conductors. There's a good chance that the heat-sink would fall off if the ceramic resistor was disspitating at its rated limit.

-- Bill Sloman, Nijmegen

Agree. The adhesives are terrible heat conductors and the contact area will be small, so it will likely cook the adhesive.

These are cool,

ftp://jjlarkin.lmi.net/Welwyn.JPG

thick-film on porcelain on a spring steel base. The base is curved so it snugs down on a heatsink nicely. Power density can be very high.

John

Nice! Are those available from Newark or... ? ...Jim Thompson

The base (claimed to be stainless steel in the datasheet) is extremely ferromagnetic, if that matters to anyone other than me. 8-(

It would matter to me in some situations too. I bet a current step will produce radical voltage waveforms because of eddy-current coupling into the baseplate.

Turns out it's a non-trivial problem to heatsink current shunts without having nasty eddy-current effects. Not to mention thermal transients, thermoelectrics, and hum pickup.

John

I bet those are nice and fast, too--good for the fast control loop in a TEC driver.

Cheers

Phil Hobbs

Probably, but see my other comment about eddy currents. I've seen roughly 50% voltage overshoots and hundreds-of-microsecond complex taus from a planar current shunt epoxied to a heat sink, and similar response from a Vishay precision 4-lead resistor in a TO-3 can. The MIL-type metal-case resistors can do stuff like this too, depending on how they're wound.

Interestingly, I don't see this if the resistor is glued to mu-metal or equivalent, just if it's on a nonmagnetic substrate, aluminum or copper or stainless. Maybe somebody can explain that to me.

Our NMR pulsed-gradient drivers have to settle to a few PPM in 100 us, so all sorts of higher-order nasties pop up, the current shunts being a major annoyance.

John

The skin depth formula has a factor of 1/mu in it, so the resistance of the surface layer you're reaching with a given transient is thousands of times higher.

Cheers

Phil Hobbs

Ah, but Dee I. Mond is an extraordinary man, such a great conductor, he leads the London Philharmonic.

Maj. Gen. Ization knows reaching conclusions quickly is a bad idea. ;-)

Tim

Firstly, Sir Ramic is not associated with trains; he is an extremely poor conductor... So dabbing any glop on one surface will not help a whole heck of a lot. The heatsink product should wind up having a squarish hole in the center to just accommodate those resistors, matching the length. Then silicone grease the resistor to help conduct heat to the heatsink. That way, the (aluminum?) box surrounds the resistor and can "soak up" radiated heat.

OOoooo....