LM335 precision?

Apr 26, 2015 25 Replies

One zone and a forgiving customer you're probably right. I think

most applications.

--Spehro Pefhany

Our new house has three zone controllers, and their accuracy is as

...Jim Thompson

| James E.Thompson | mens | | Analog Innovations | et | | Analog/Mixed-Signal ASIC's and Discrete Systems | manus | | San Tan Valley, AZ 85142 Skype: skypeanalog | | | Voice:(480)460-2350 Fax: Available upon request | Brass Rat | | E-mail Icon at http://www.analog-innovations.com | 1962 | I love to cook with wine. Sometimes I even put it in the food.

*NO* assumptions. Work with an automated semiconductor testing line. Also learn simple manufacturing economics. For every five cents added to an auto part, the price must go up by five dollars. A simple analog opamp will cost maybe a millicent (do you know her?) to fab, but it can cost up to a cent or so to package,and one CANNOT afford to toss too much money into the testing rathole because the total cost must be multiplied by a *MINIMUM* factor of five just to break even (economic FACT).

that is?) not only to move the part from the "room temperature" environment to the test chamber, but even MORE TIME to GUARANTEE its temperature has settled to the required temperature; and THEN AND ONLY THEN tested. Ever hear of five (or more) time constants?

Did i leave something out?

Common sense? Market knowledge?

You appear to be assuming that a *thermal* sensor would be allowed to stabilize be and calibrated one at a time.. no way is that practical in any volume with high accuracy.

We know it's possible, because non-semiconductor sensors interchangable to high accuracy are available for literally pennies (including calibration, packaging, etc.), even in moderate volume.

Also, since there are multiple makers in different countries, we know it's really not all that difficult in 2015 (it was a bit harder 20 years ago).

In fact, I'm 100% sure I (and a number of other people here) could design such a setup that would work straightaway. I'm also pretty sure the off-the-shelf semiconductor testing equipment designed for non-thermal testing won't work well enough, or they would be using it. So, some capex is required to address the market. The alternatives are very cheap and are established (because the other guys were down low and the semi guys were too slow) so ROI may not be sufficiently attractive.

BTW, EBIT in automotive parts supplier companies (plenty in this area) is in the 5% range- they would laugh at the idea of seeing anthing more than a marginal price increase for mere tightening of specs. Their customers generally nail them solidly to the wall.

--Spehro Pefhany

Why not do a single point calibration at "room temperature"?

Or have a temperature controlled "room" containing 10k of the parts to calibrate.

Or a "conveyor" with length of 1000 parts equivalent and a temperature controlled "tunnel".

Or...

John Devereux

There is a middle eastern souk/supermarket around the corner from my office- they've got an automated pita facility that takes bulk dough, slices it off, takes it through proofing flattening dusting baking and cooling (perhaps 45 minutes to an hour total)- they're packed in bags for about 1/10 the price of an LM335 each. Long time per piece, lots of pieces per hour for the machinery. Absolutely delicious fresh, though they seem to go dry very quickly.

formatting link

I've seen similar things for tortillas.

Of course commercial bread baking facilities use similar, but much bigger, setups- but this is practical even at a relatively small scale.

Best regard,

--Spehro Pefhany

Join the Discussion

Have something to add? Share your thoughts — no account required.

Didn't find your answer?

Ask the community — no account required