temp measure with diode 1N4148 and differential amplifier

[ericgibbs]

Okay.. take your time and take your rest, Eric.

hi Viz,
Found one of my reference books, this is a PWL [piece wise] approximation.

 

[atferrari]

Convenience of using it nowadays?

Found one of my reference books, this is a PWL [piece wise] approximation.

A (very) cheap sensor followed by a costly linearization circuit.

With so many linear sensors available, the LM335 among them, what could be the reason to use that approach? I cannot see the advantage.

 

[SimonTHK]

A (very) cheap sensor followed by a costly linearization circuit.

With so many linear sensors available, the LM335 among them, what could be the reason to use that approach? I cannot see the advantage.

The advantage is to learn it awesome and I want to build that thing that can liniarize a bowed line with electronic parts. Try to imagine what else I could liniarize xD like everything

 

[Sceadwian]

atferrari, you have your idea of what is proper Eric has his, so does everyone else around here. Make no mistake that your idea of what is right is not what is right for everyone else; because you do not see any advantage doesn't mean there aren't any.

 

[bychon]

I think the best way to see an advantage is to get in a Wayback machine and go back about 40 years, to the time when LM35 chips were not invented yet. Under those conditions, it suddenly makes sense.

 

[Vizier87]

hi Viz,
Found one of my reference books, this is a PWL [piece wise] approximation.

I see. Looks like an employment of numerical methods.. Thank you very much Eric.

A (very) cheap sensor followed by a costly linearization circuit.With so many linear sensors available, the LM335 among them, what could be the reason to use that approach? I cannot see the advantage.

the cost is not that much by the way, I calculated the cost of all the components and it came around to RM5.00 or approx eqv to USD 1.50. That is if we use exact-valued resistors according to the diagram. The power consumption is another thing... perhaps two MCP62414 quad op-amps might serve the purpose (50uA supply current). So if you're still using the network then it would still be very useful.

double sided PCB+surface mount parts will further put this one down to just as big as a dollar coin.

 

[SimonTHK]

Can anyone tell me if this is possible?

I cant firugre it out. The current must be constant over the diode?
I would really like an explanation on how this circuit would react.
Due to this circuit, it should be possible. https://www.capgo.com/Resources/Temperature/Semiconductor/DiodeTemp.GIF

 

[ericgibbs]

The diode is wired as shown in my post #4, the post also shows the result.????

 

[SimonTHK]

I would say that also yes. Even when you say it again ... I will add a inv. Opamp aswell. Thx alot.

 

[SPDCHK]

One member is using this type of temperature control for the roof solar heated panels and his swimming pool, been working ok for over 2 years.

Yip, and that member would be me. Eric helped me with the circuit.

As reference, I also added a PT100 sensor right next to the 1N4148, and calibrated both sensors for 0 .. 100°C.
I’m amazed at how accurate the Si diode is. Yes, I admit, it’s not 100% linear, (at worst about 2°C out) but for my application it works like a charm.
Have a look at this post (https://www.electro-tech-online.com/threads/pic16f88-freezing-in-operation.39061/) for the circuit and code.