gap-pad thermal conductivity

A few years ago, for best rsponse times I was embedding LM35's in allegedly thermally conductive epoxy.

I recall looking up the thermal conductivity of the epoxy, and realizing that while it was far better than air, it was terrible vs metal to metal.

-- A host is a host from coast to snipped-for-privacy@nrk.com & no one will talk to a host that's close.......................... Unless the host (that isn't close).........................pob 1433 is busy, hung or dead....................................20915-1433

The LM35 is available in a TO-220 package, which ought to offer really low thermal resistance.

doesn't provide all that much information, as you would expect from TI.

I was sensing temps of copper pipe. As I recall, the mounting tab was live, and needed isolation. I cast the LM35 in the expoy on the pipe.

The only way to get heat into an IC sensor quickly is via the leads.

Cheers

Phil Hobbs

On the TO-220 version of the LM35, the case/tab is ground.

LM35's love to latch up, and don't like capacitive loads, so they can be tricky, especially at the end of long wires.

Never run V+ over 5 volts, and don't pull the output negative if you don't absolutely need to. Bob Pease told me that he'd fix it, but he didn't.

RTDs or thermistors or thermocouples are safer.

The Diodes Inc ZNI1000 is cool.

low thermal resistance.

ve, and needed

Thermocouples are pretty horrible. You are measuring a low, constant DC vol tage, so offset voltages are always a problem.

Thermistors can be AC excited, so if you don't dissipate too much power in them you can measure really low temperature differences with them.

Temperature sensing thermistors are negative temperature coefficient device s, so if you dissipate too much power in them you could get hot channels fo rming (which I've never seen) but a lower power levels you can get weird ef fects as heat dissipation creates warmer channels inside the thermistor, wh ich move the lowest resistance channel around the device as it warms up, gi ving an unstable resistance. That came as shock when we ran into it.

Resistance thermometers are more stable than thermistors, and essentially l inear, so you can dissipate as much power as you like (if you keep track of the self-heating). They are a factor of ten less sensitive so you need AC excitation sooner, and do need to dissipate more power.

Semiconductor temperature sensors can get by with low power dissipation (th ough some of them don't) but do tend to be noisier and crankier than thermi stors or RTDs.

It's a nickel resistance thermometer, with a bit less than half again more sensitivity than platinum resistance sensors, which isn't much, and the dat a sheet doesn't go into why it's more sensitive, or that nickel is more rea ctive than platinum so resistive sensors made with it are correspondingly l ess reliable.

During WWII, pilots land a B-17 and mistakenly retract the gear, not the flaps, on roll out. The solution was a small wheel on the end of the gear handle, and a small flat knob on the flap handle.

Buy/keep bottles of 2-3 different sizes/shapes; pour new bottle of each product into said distinctive bottles.