Idiot's Guide to LM35

[Vizier87]

Here's my schematic:
https://www.electro-tech-online.com/attachments/amp-jpg.34774/
so the gain is 1+ (R1/R2), and R1=50 K trimpot, and R2 is a 10K carbon resistor.
Problems:
1. The LM35 heats up easily, so do I need to put a resistor in series into Vin?
2. The amplification didn't work! What's the problem? I've seen numerous schematics from the net, but I don't know which one should be used, so I chose this one for starters. So we can work a way slowly from here.

I use LM358 for the op amps.
Cheers.

 

[ericgibbs]

Here's my schematic:
https://www.electro-tech-online.com/attachments/amp-jpg.34774/
so the gain is 1+ (R1/R2), and R1=50 K trimpot, and R2 is a 10K carbon resistor.
Problems:
1. The LM35 heats up easily, so do I need to put a resistor in series into Vin?
2. The amplification didn't work! What's the problem? I've seen numerous schematics from the net, but I don't know which one should be used, so I chose this one for starters. So we can work a way slowly from here.

I use LM358 for the op amps.
Cheers.

hi,
The LM35 should not heat up when using a 5Vdc supply.

As drawn your circuit should work, it suggests that you have wired the LM35 or LM358 incorrectly.

What voltage do you measure on the output pin of the LM35.?

NOTE: LM35 pin view.

 

[AllVol]

2. The amplification didn't work! What's the problem? I use LM358 for the op amps.
Cheers.

Don't see why you want to amplify the output of the LM35 anyway. Its output is a voltage that directly corresponds to the ambient air temperature. If you connect a digital multimeter set to read voltage across the output and ground of the LM35, you will see a direct temperature readout by simply moving the decimal two places right. Example: .268 represents a temperature of 26.8 degrees C.

No amplification needed.

 

[MikeMl]

If the LM35 is self-heating, you have it miswired! (or it has been damaged).

Your amplifer should work, except that the LM358 is not rail-to-rail on its output, and with its supply pin tied to +5, it will only pull-up to ~3V, so if the gain is set too high, even with the ambient temperature at 25degC, the output may saturate.

If you have a higher positive supply available, like 12V, power just the opamp on that. That way, the output can pull much higher...

Another way; get a genuine rail-to-rail input and output opamp.

Yet another way; add an offset circuit to the inverting input of the opamp so that the output swing is shifted downward.

 

[Vizier87]

Don't see why you want to amplify the output of the LM35 anyway. Its output is a voltage that directly corresponds to the ambient air temperature. If you connect a digital multimeter set to read voltage across the output and ground of the LM35, you will see a direct temperature readout by simply moving the decimal two places right. Example: .268 represents a temperature of 26.8 degrees C.

No amplification needed.

Well, I'm going to feed its output to a PIC, and the output is too small since the Vref of the PIC needs to be small enough to detect small changes of the LM35 output. This might mess things up when I have a other types of sensors to be used in this system, I suppose.

Anyway, there should be quite some changes in the output if I touch the plastic casing of the LM35, right?

 

[blueroomelectronics]

Any temperature sensor should have extremely low self heating. It would be a very bad temperature sensor if it heated up!

 

[AllVol]

Anyway, there should be quite some changes in the output if I touch the plastic casing of the LM35, right?

Yes, there should be a very definate change in output. If you attach your dmm as I described, you can observe this change for yourself. I use the LM34 myself, (same chip except it reads in God's temperature, not man's /jk) and have observed some very subtle changes in output, coniciding with other electronic thermometers.

There was another poster here a couple of days ago experiencing grief over his LM34. Turned out he had it wired completely backwards.

 

[Vizier87]

Another way; get a genuine rail-to-rail input and output opamp.

Yet another way; add an offset circuit to the inverting input of the opamp so that the output swing is shifted downward.

Well...
1. How do I get a genuine one?
2. Please elaborate on the offset thingy. I heard that some ICs do the job producing negative outputs to be fed into the inv. input, so is that what you meant?

 

[MikeMl]

Well...
1. How do I get a genuine one?
2. Please elaborate on the offset thingy. I heard that some ICs do the job producing negative outputs to be fed into the inv. input, so is that what you meant?

Go to TI's/National/Fairchild/DigiKey/Mouser web sites. Use their part search to locate "low voltage, rail-to-rail input, rail-to-rail output" opamps. There are hundreds of them nowadays. These are the ones that are suitable to amplify a sensor output so that it can be applied to the A/D input on a PIC operating on 3.3 or 5.0V Vdd.

Look up TLC2272/2274 for example.

 

[Vizier87]

While I'm scrounging for a TLC2272 in our department and killing time , is there a way to improvise using the LM358 for the 5V apps only?

 

[Vizier87]

Yesssss.... It worked.. my mistake is to see the LM35 view as 'top' view rather than 'bottom' view.. hence the heating and dysfunction..

I tell ya... I've been trying to amplify signals over a year... today it worked.. whew. Lesson the hard way I guess.

 

[agwtpcbox]

you should see the datasheet carefully