T&L Publications

I have designed a cct. for a temperature sensor. Can it work?<p>[img=http://img371.imageshack.us/img371/1403 ... ept4cm.png]

?????????

Wierd schematic- it looks like a yellow frog with a bunch of words.

well I give another try!<p>

Sure it'll work. You probably won't get any significant self-heating of the diode at 800µA. But you are going to have one heck of a look-up table or other conversion routine if your intent is to derive an actual measurement in degrees or any linear change in temperature, for a simple circuit like that will have zero linearity except for very small temperature changes. If all you need is to turn a device on or off at a certain set point, then it works great. But feeding an ADC ... I doubt that's where you're going.<p>Dean

I base my circuit on this.<p>http://img244.echo.cx/my.php?image=sensor6vs.png<p>The silicon diode voltage changes linearly from about 50K tp 400K.

Usually you need a “absolute” current source, to keep it linear.

<blockquote><font size="1" face="Verdana, Helvetica, sans-serif">quote:</font><hr> I base my circuit on this.<p>http://img244.echo.cx/my.php?image=sensor6vs.png<p>The silicon diode voltage changes linearly from about 50K tp 400K.<hr></blockquote><p>Well, unless you are four years old I'd say you did not really "design" anything. Not a flame, just an observation.

The battery voltage in that circuit will change over time/usage so not only will there be error but the size of the error will change over time. You could measure the voltage with your hammer, er DAQ card, and compensate but I think a better way to go is use a temp sensor like the LM34 or the Centigrade version of it that I can never remember the number of... At least the errors are well characterized.<p>On a somewhat cranky note, it seems like you are using this board as a design shop. Toss a random design at us and we'll fix it. I'd like to see more care go into your work before you ask for my help. I'd prefer you not ask questions that could be obtained from a 5 minute reading of the manual or from the top 2 or 3 google hits. I suspect others feel the same. I'm not saying this to chase you away but if you used the resources available, you could stand more on your own two feet.

The reason, I want to use the battery rather than the main supply is because I believe that the battery current would be a constant. On the other hand there would be a ripple from the mains supply (normally ac converted to dc). This might effect my readings.

The battery current is not constant, and this is why they use “current regulators” to set the current, and keep it constant. Any page out of the National Semi” application notes will show you how its done.

Can I use the current/voltage regulator with a battery? If I use the mains with the current/voltage regulator there might still be an "ac" ripple?

Labview
At what stage of electronic theory are you at? This is not meant to be insulting, just curios. Most of your posts could be solved with simple Ohms law calculations. If not up to speed on this, then brush up on it as it will answer a lot of your questions,enabling you to proceed at a faster rate with your designs.
A 120VAC derived power supply thru a proper regulator should have virtually no ripple, at least none that you could measure. These chips can even be wired as constant current sources.,as per Chris's suggestion.

http://www.national.com/pf/LM/LM34.html<p>These are so cheap it's ridiculous. Try allelectronics.com<p>Bear in mind that a good design chief is well aware of the relative costs of having a functional circuit designed in-house versus the cost of an off-the shelf gizmo that does the same thing.<p>You want temperature readout? Almost anything can be used as a transducer. But this LM34 series takes in just about any power supply up to 35V at 75 uA, and puts out linear temperature readings in tens of millivolts per degree from -50dF to +300dF. A variant is available for degrees Centigrade.<p>I am sometimes disappointed at all the gyrations and gymnastics that people go through to do something electronical, when off-the-shelf items are usually available to perform the function, and usually for cheap. I am well aware of the challenge of "doing it yourself" (DIY), and I am well aware of the instructional aspects of DIYing it, but sometimes I see some riciculous "solutions" to problems which can be solved with a simple search for an off-the-shelf product.<p>Used to be you could walk in an e-store and state the function, and knowledgable salespersons could at least point you in a direction toward a solution. Them days is pretty much gone, but nowadays a web search on a given function will yield an abundance of possible solutions. For example, the search term, "temperature sensor" gives you 10,800,000 hits. Filter it down a little, and you can readily find what you need.<p>And in this case, the LM34 costs less than three whole US dollars, interfaces directly to DACs, and is independent of power supply variations, to boot. <p>Snarl.<p>Growl.<p>Growf!<p>(Message edited for substitution of various unsuitable Biblical and Middle-English words.)<p>[ September 09, 2005: Message edited by: terri ]</p>

I want to measure the temperature of liquid nitrogen which is about 77K. LM34 can't do that. I think! However I need only about +-5K accuracy.