Optical-tester project

Thread Starter

Al.ro92

Joined Sep 7, 2016
31
"Vout= -R2/R1 (V2-V1)"
The output is the ratio of the resistors multiplied by the bit in brackets. This is the difference between V2 and V1. So 1V and 1.!V will give the same output as 3.3V and 3.4V. That is the magic of that circuit.
Oh ok, that's nice! What I need than is a resistor ratio of 10 or more. I thought the inputs also influenced the gain.



You might need to null any intrinsic offset of at least the first stage opamp.
Thanks, i will look that up. I didn't know i had to take that into consideration as well.
 

Alec_t

Joined Sep 17, 2013
14,280
The (typical) input offset voltage (as specified in the opamp datasheet) gets added to (or subtracted from) the 10mV to 50mV voltage difference you are trying to measure. It gets multiplied by the opamp gain, so it's best to null it at an early stage. Lower offset opamps are generally more expensive than higher offset types.
 

Thread Starter

Al.ro92

Joined Sep 7, 2016
31
The (typical) input offset voltage (as specified in the opamp datasheet) gets added to (or subtracted from) the 10mV to 50mV voltage difference you are trying to measure. It gets multiplied by the opamp gain, so it's best to null it at an early stage. Lower offset opamps are generally more expensive than higher offset types.
Thanks for the info, I can either null the intrinsic offset at the first stage or get a low offset opamp. I think the second option is better, paying a few euros extra is not a problem.

I found this one: http://www.produktinfo.conrad.com/d...-IC_OPAMP_CHOP_1__TLC2652AC_8D_SOIC_8_TID.pdf

It has an input offset voltage of 2.35 μV, and a max differential input voltage of +-16V. So I think this one could work. What is not clear is if it needs a negative voltage supply or if you could ground it.

I now finished the circuit i was working on yesterday:
comparator5.PNG

I use a gain of 50 at the first stage. So the vref voltages are higher. For the window comparator in the middle, for the orange light, I have a LOW Vref of 2.5v which is the same as the HIGH Vref on the first one for the green light. Could that be ok or should i change the LOW Vref to about 2.51v?
 

AlbertHall

Joined Jun 4, 2014
12,343
Without the '4049 the window comparator works the other way round. So the green LED will be lit for voltages outside the window.
You could fix that by using a transistor (NPN, common emitter, base connected by a resistor (10k perhaps) to the diodes) between the the diodes and the LED to do the inversion.
 

Thread Starter

Al.ro92

Joined Sep 7, 2016
31
Without the '4049 the window comparator works the other way round. So the green LED will be lit for voltages outside the window.
You could fix that by using a transistor (NPN, common emitter, base connected by a resistor (10k perhaps) to the diodes) between the the diodes and the LED to do the inversion.
Ok so that's the reason for the inverter, do you mean something like this?
inverter.PNG

But I could also use the 4049 and get the 5v supply from the 12v rail.
 

AlbertHall

Joined Jun 4, 2014
12,343
The circuit is OK, but you don't need the resistor in series with the LED. The current setting resistor for the LED is the one to the supply.
You could use the '4049 - the choice is yours.
 

Alec_t

Joined Sep 17, 2013
14,280
What is not clear is if it needs a negative voltage supply or if you could ground it.
You can ground it, so a single polarity supply is fine. The clue is this bit of the datasheet : "Common-Mode Input Voltage Range Includes the Negative Rail".
 

Thread Starter

Al.ro92

Joined Sep 7, 2016
31
Hey guys,

After a few busy days I started working on the project again.

I was able to built the schematic on a breadboard and get the correct values finally, so now I can continue with the building part.:)

As mentioned earlier I want to replace the multimeter with LED's for the FC measurements, I simulated the schematic I made and it works too:tinacircuit.PNG

This is what it does:
Green LED burns when FC2 is between 0.01-0.05v bigger than FC1.
Orange/Yellow LED burns when FC2 is between 0.051-0.1v bigger than FC1.
Red LED burns when FC2 0.11v or greater than FC1.
Red LED burns when FC1 is greater than FC2.

I used a gain of 50, so 0.01-0.05v = 500mv-2.5v
0.051-0.1v = 2.55v-5v
0.11v= 5,5v


After I got that right i started with the complete schematic of the new box design, this is what i got so far:
completeschema1.jpg

As you can see you have to connect the sensor first to do the measurements. The sensors we make are all basically the same, but they are for different models of the machine they are made for, so they have different connector types:
connectors.JPG

In my drawing i have one place where you can connect the sensor to, what i actually want is to have 3 or 4 of these for the different connector types.

I could already draw that part of the schematic, but what has got me thinking right now is:
Say I have 2 or more sensors connected to the box, then if i meausre, I would be measuring the values of 2 different sensors which is not what i want.

How could i do this the correct way? So when I connect sensors to the box i can choose which one i want to do the measurements on.

Thanks!


 
Top