op amp question

Thread Starter

Little Ghostman

Joined Jan 1, 2014
305
I thought I had it sorted, I want to take a small input voltage (10mV-300mV) and amplify it 100x so I get 100mV-3V.
Simples I thought, a non inverting amplifier and off we go. I have some op amp reference circuits and like a muppet, I thought I would take a look through and see if there is anyway to do it with a single supply. What I was looking for, was a circuit I collected a couple of years ago, it used a op amp to create a negative supply, i cant remember much about it and never used it, I was then going to use this to drive the supply for another op amp to up the small imput voltage.
Well I have now utterly confused myself! :confused::confused:
So any idea's how best to go about this? any hints would be great so I can untangle my brain again!
 

MrChips

Joined Oct 2, 2009
30,711
You only need 10x gain to go from 10mV to 100mV.
If your input and output signal are both positive then a single supply opamp will work. Something like LM324 will work except this is a quad opamp chip. LM358 is a dual opamp in an 8-pin package.

Here is what a basic opamp circuit looks like:



Make R1 = R3 = 1kΩ
R2 = R4 = 10kΩ

Connect your input signal to V2.
Connect V1 to GND.
 

Thread Starter

Little Ghostman

Joined Jan 1, 2014
305
Thanks for that. the 100X thing was a totaly brain malfunction! i admit my maths is poor, but almost not that poor :D, for some reason when I thought of 20mV I convinced myself that 100x that was 200mV!!!
And just when I thought I was making progress LOL :D

I forgot to mention, The sensor is a diode, I need to read temperature in a confined space, a diode would fit great, and the temp upper range is probably 150-170C, in other parts of the project I am using a DS18s20 (well several), but i couldnt even get a NTC thermistor to fit, but a nice small diode... ;)
 

ronv

Joined Nov 12, 2008
3,770
You could probably use a diode like a 1N914. It's temperature is good to 200C. You could then apply a small forward voltage and measure the change in voltage drop with temperature. You would need to calibrate it initially, but then I think it might work.
 

Thread Starter

Little Ghostman

Joined Jan 1, 2014
305
You could probably use a diode like a 1N914. It's temperature is good to 200C. You could then apply a small forward voltage and measure the change in voltage drop with temperature. You would need to calibrate it initially, but then I think it might work.
I will look at the datasheet, not sure of the diode I used to test size, but it only just fit between the heat sink fins, so space is very tight.
200C would be good if I can get it to fit! maybe I should use a constant voltage source? For this application accuracy isnt that important, I can live with 5C accuracy, its mainly to stop a element getting too hot, and also to make sure when switched on, the element is getting hot.


EDIT
AHHHHHhhhhhh
Now ive said that I am aiming for 1C accuracy!!! I knew I shouldnt have mentioned it :D
 

Thread Starter

Little Ghostman

Joined Jan 1, 2014
305
Thanks Ron, I will study, can ask what search term you used please? as that is one I didnt get on my google, so I like to know what others put in to get pages out
 

ronv

Joined Nov 12, 2008
3,770
I think I used diode voltage drop vs. temperature.

Here is the simulation. Note the voltage goes down as the temperature goes up and the current doesn't change much if the supply is high compared to the diode voltage drop.
 

Attachments

Thread Starter

Little Ghostman

Joined Jan 1, 2014
305
Apparently according to that link, LT spice is pretty accurate in the sim doing temperature! there was a small variation, but apparently easily with the range of experimental error!
That will be handy once I can do sims better in LT. Thats the next step! see if I can get LT spice to sim for the different diodes and compare
 

Thread Starter

Little Ghostman

Joined Jan 1, 2014
305
175C is max, I will shut the power off at around 140, the extra is overhead because the element continues to heat a bit after power off. Was temp is up to range, the diode will operate at around 40-50C. the higher temps are just for initial heat up.
Works perfectly on paper LOL, now watch the real world mess it up :D
 

GopherT

Joined Nov 23, 2012
8,009
Making the op amp amplifier, you can use the diode as sensor and the circuit on the first reply but add some potentiometers on the non-inverting input to allow you to adjust (while diode is in an ice bath) 2.7315V and one pot on the feedback to adjust gain to produce 3.7315 volt while in a boiling water pot. Repeat ice and boiling water until all is calibrated (1 degree Kelvin = 1.0 mV)

Your glass body diode should be linear to 200 degrees or so.
 

atferrari

Joined Jan 6, 2004
4,764
Many years ago I tested a huge bunch of 1N4148 diodes, measuring their forward voltage while changing the temperature between 0ºC and 100ºC (ice bath and boiling water temperatures respectively).

The extreme values seemed to more or less match throughout the batch.

Their response was not very linear and the curve was not even symmetric.

The highest separation from the ideal liner response was, IIRC, somewhere in the the top half of the curve. (around 60ºC maybe?)

Do not be surprised I say, "more or less" or "very linear" because I lost all my notes in a voyage abroad. After that, I never took again anything from my bench even if traveling for a long time.

It was previous to my real first design, many many years ago when I started with micros (16C57).
 
Top