find replacement of the LM35DZ temprature sensor

Hello,
I am using LM35DZ temprature sensor for a project but due to long cable connected between the sensor and my device my temprature sensor is not able to give an exact value. So I just want to find the raplacement of my LM35DZ temprature analouge sensor..if anyone knows about this please help me.


Regards,
Abhishek Dixit

if your base problem is the long wire distance, then you must go to a digital sensor like DS18B20
the 1-wire protocol can go as far as some 100 metres
but the measuring circuit is very different from the lm35.
you have to use a microcontroller with the 18b20

You would find it very challenging to get 1-Wire to work over 100m !!

The accuracy of the LM35 should not be changed by the cable length so you really need to see what is causing the error. If it's the capcitance of the cable, either buffer the voltage at the LM35 end of the cable or change to a sensor like the MCP9700 which can drive large capacitive loads. If the problem is noise pick-up along the cable, consider a low pass filter at the measurement end. The output of the LM35 changes slowly so you could use a filter with a very low cut-off frequency, say < 5Hz. Do you have a supply decoupling capacitor at the LM35 end of the cable?

Brian.

betwixt said:

You would find it very challenging to get 1-Wire to work over 100m !!
Brian.

Click to expand...

Maxim claims to go to some 500 meters with that ?

Believe me I've done it!

Under ideal conditions, using the correct cable and with impedance matching and if not using parasite power it is possible to have that much cable in the system but only in certain topologies. For reliable use, especially using non-optimal cables I wouldn't recommend going longer than about 20m, especially if more han one device is on the cable. However, this isn't the issue in this thread when an analog voltage is being carried. That's why I'm trying to establish the cause of the error rather than find an alternative measurement method.

Incidentally, for very remote temperature measurements I use a DS18B20/DS18S20 driven by a SOIC PIC10F200 which converts the scratchpad bytes to low speed differential serial bytes. The differential serial signal can go possibly Km, I've never actually tested it's capability beyond about 100m. The sensor and PIC sit on a tiny PCB, barely wider than CAT-5 cable and the whole cable end and PCB is encased in a blob of resin. The systems have worked perfectly for many years.

Brian.

betwixt said:

You would find it very challenging to get 1-Wire to work over 100m !!

The accuracy of the LM35 should not be changed by the cable length so you really need to see what is causing the error.
Brian.

Click to expand...

Brian, that is Not Entirely True, the accuracy will be not change, at my place of work we used a plugin module with a LM35.
when ever one needed replacing or moved, customers reported wrong reading were being recorded.

When on a service call we had no way of changing the code in the main panel, it was one of the "we need to keep you paying to have our kit updated".
Panel was then out of service up to two weeks..

so a work round was to change the voltage to the module, this was also seen as a way to offset a device.

I'm not sure I understand you. With a digital sensor, using 1-wire the accuracy will not change if the cable length changes but at some point the timing will degrade to the condition that the digital information cannot be recovered.

Cable length to an LM35 should not influence the accuracy (at least for normal cable types) although it may allow the entry of interference or in extreme cases cause resistive drop. The LM35 is not particularly good at driving capacitive loads (the MCP9700 is far better) so its output might have some stability problems if a large capacitance appears across it's output. If you are seeing differences in outputs of different modules it is either because of production differences in LM35s or some other factor is introducing an error voltage.

Brian.

Hi
you can use shield wire for this...and then if it isn't true yet , compute the difference ...if different is fix always so you can change your formula with this difference...thereupon you have a good temperature sensor!!!

use two lm35 tem. sensor parallel instead of one .

akhilpaulv said:

use two lm35 tem. sensor parallel instead of one .

Click to expand...

Hi dear
what's the benefit of this metod ? can we use more longer wire ? would be correct our value?

- - - Updated - - -

akhilpaulv said:

use two lm35 tem. sensor parallel instead of one .

Click to expand...

Hi dear
what's the benefit of this metod ? can we use more longer wire ? would be correct our value?

greentree said:

Hi dear
what's the benefit of this metod ? can we use more longer wire ? would be correct our value?

- - - Updated - - -


Hi dear
what's the benefit of this metod ? can we use more longer wire ? would be correct our value?

Click to expand...

efficiency of sensing part will doubled

akhilpaulv said:

efficiency of sensing part will doubled

Click to expand...

but with this way error become double too!

greentree said:

but with this way error become double too!

Click to expand...

that we can controls by an opamp

first lm35 connect to V- and second to V+...yes?

nope.. just parallel. both connected to +v

abhishekdixit said:

Hello,
I am using LM35DZ temprature sensor for a project but due to long cable connected between sensor and my device my temprature sensor is not able to give exat value. So I just want to find the raplacement of my LM35DZ temprature analouge sensor..if anyone knows about this so please help me.


Regards,
Abhishek Dixit

Click to expand...

What is cable length ?

What cable did you use ?

OpAmp is used and where ?