Simple Temp Sensor Circuit using TMP36 and LM358

Thread Starter

Lucky-Luka

Joined Mar 28, 2019
165
Hi all
I have to monitor the temperature of a surface for several hours per day. The circuit will be battery powered.
I already have a LM358 op-amp and a TMP36 sensor at home.
I thought 4x AA NiMh batteries (1.2V) are ok.
I just want an LED ON when the temp is over 50 degrees Celsius and another LED ON when the batteries are low (I thought 0.9V*4= 3.6V = low voltage).
Do you see anything wrong?
Any improvements or suggestions?
Thanks
 

Attachments

ericgibbs

Joined Jan 29, 2010
15,332
hi LL,
The change in battery voltage will affect the temp level comparator input, you need a voltage reference for the RV1 pot set voltage.
E

++++

Consider a tap off D1, [ decoupled] for RV1 set voltage.
 

Ian0

Joined Aug 7, 2020
4,821
I imagine you are sensing temperatures between 0°C and perhaps 70°C. That is an output of 500mV to 1.2V from the TMP36.
Your pot can vary the voltage from 0V to 4.8V, which is -50°C to 480°C (although the TMP36 would melt first), so the region you are interested in is a really tiny adjustment on the pot.
A 1.2V reference would be ideal as would set the upper level to 70°C. Then you might want a resistor in series with the pot to make the lower limit more sensible
 

Ian0

Joined Aug 7, 2020
4,821
You can probably use the same 3.6v reference voltage from D1.
Or perhaps not, as Fairchild is not forthcoming with the temperature coefficient of the zener, and you may find it changes with temperature by more than the output of theTMP36!
 

ericgibbs

Joined Jan 29, 2010
15,332
hi agu,
I don't see a tiny picture.? It fills the PC screen.
I agree the components are further apart than needed.

E
The line squared graph paper makes diagrams difficult to read.
 

Thread Starter

Lucky-Luka

Joined Mar 28, 2019
165
hi agu,
I don't see a tiny picture.? It fills the PC screen.
I agree the components are further apart than needed.

E
The line squared graph paper makes diagrams difficult to read.
He is the only one complaining about my schematics in different threads and even in different forums.
It does not bother me tho.
The grid is due to Proteus simulator.
 

Audioguru again

Joined Oct 21, 2019
4,523
The zener diode you selected needs a much higher current (69mA) than you feed it (5.4mA). Many 3.6V zener diodes are available that operate at 5mA.
But a low voltage zener diode makes a poor voltage regulator and its voltage drops as it warms up.
 

Ian0

Joined Aug 7, 2020
4,821
But a low voltage zener diode makes a poor voltage regulator and its voltage drops as it warms up.
Agreed. It needs to be a bandgap reference with almost zero temperature coefficient.
Even if the sensor is located remotely (post #7) the circuit is no good if the trip point varies all over the place with the local temperature. 1.2V makes a good reference as it corresponds to 70°C on a TMP36
 

Thread Starter

Lucky-Luka

Joined Mar 28, 2019
165
Agreed. It needs to be a bandgap reference with almost zero temperature coefficient.
Even if the sensor is located remotely (post #7) the circuit is no good if the trip point varies all over the place with the local temperature. 1.2V makes a good reference as it corresponds to 70°C on a TMP36
The circuit is far from the heat source.
Can I obtain that 1.2V reference using another TL431?
So the circuit will see 2 TL431 and a lm358, right?
 

Ian0

Joined Aug 7, 2020
4,821
The circuit is far from the heat source.
Can I obtain that 1.2V reference using another TL431?
So the circuit will see 2 TL431 and a lm358, right?
TLV431 perhaps, but there are plenty of 1.2V references out there.
You can use a single reference for both parts of the circuit.
 
Top