# Problem with opamp and instrumentation amplifier

Discussion in 'The Projects Forum' started by Dejan, Aug 8, 2009.

1. ### Dejan Thread Starter New Member

Apr 14, 2009
7
0
Hi,

I'm trying to power up the signal with instrumentation amplifier, that comes from operation amplifier. Problem - gain loss! When a source of stabilized voltage is connected with instr. amp. instead of LEM transducer, everything works fine. What's the problem?

File size:
29.3 KB
Views:
82
2. ### beenthere Retired Moderator

Apr 20, 2004
15,815
283
Is the transducer output AC in nature? If so, none of th eop amps can swing negative and follow the signal.

The signal appears to be single-ended, too. Why bother with the instrumentation amp, when gain is all you need (no difference voltage or common mode to reject)?

3. ### Dejan Thread Starter New Member

Apr 14, 2009
7
0
Point is that the transducer gives 2.524V at 0A. I have to do some limitations, and it would be much easier when it would be 0V at 0A, and of course 5V at current maximum (3A), which is not satisfied, because 4,627V i get whet I=20A, so i can only operate with about 0.212V... If using simple voltage subtractor (diferential amplifier) the CMRR gets high, so at 0A have 0,6V.

4. ### beenthere Retired Moderator

Apr 20, 2004
15,815
283
Got a number for that transducer?

5. ### Dejan Thread Starter New Member

Apr 14, 2009
7
0
Lem cas 6-np, here is transducer's spec...
2.5V on instrumentation amp is supplied from voltage devider, 10V devided by 4700Ohm and resistance from potentiometer (about 17KOhm)...

Now i have a new problem which is same type - when trying to gain up the signal from voltage devider 21V/0.1V and then 0.1V to 5V with ordinary opamp gain is ok and is 10. In second case, when the voltage from shunt resistor (0.05 Ohm * 3 A = 0.15 V) is being gained to 5V, cant get more than 3V??????
Does this have something to do with matching source (shunt res as voltage source) and load (opamp) impedance? If does, how can it be achived?

File size:
213.6 KB
Views:
35
• ###### LEM_spec2.jpg
File size:
196.7 KB
Views:
26
Last edited: Aug 9, 2009
6. ### beenthere Retired Moderator

Apr 20, 2004
15,815
283
Part of you r problem may be in the LM324. It is an old op amp and can't swing the output to the supply rail. That 3 volt output with a 5 volt supply is probably as good as it gets.

If I read the spec sheet correctly, the transducer output is going to be 1/2 Vcc at 0 amps. + 6 amps will give 4.675 volts out, - 6 amps will give 375 mv out. That makes it possible to use a single supply op amp, but even a rail-to-rail op amp can't have any gain with the transducer signal being that close to 5 volts.

What is the circuit feeding? What range of voltage do you need?

7. ### Dejan Thread Starter New Member

Apr 14, 2009
7
0
You mismatched couple of rows - for +20A sensor gives 4.627V and for -20A, 0.375V. For zero current it's 2,5V... LM324 has a 10V supply...

8. ### nicsky New Member

Aug 1, 2009
8
0
I dont see why you need int amp.
if the transducer outputs 0-5 (0.3-4.5) and you want it to go to a PIC (which has a mx analog in of 3V) gain needs to be 5 to 3V- 0 to 0, 2.5 to 3v.
use a 10/10k voltage div followed by Gain of 1.2 amp (10 -14K)
The voltage div on iput meens that opamp only sees 2.5V max and output is 3V max which is 2V below 5V VCC which should be OK for most opamp and fine for PIC micro AD input

9. ### Dejan Thread Starter New Member

Apr 14, 2009
7
0
Problem solved , thank's guys for all suggestion, and comments. Mistake i made was that inputs weren't connected to supply voltage, so measured voltage was floating, thus nothing, (or something ) gained up. When I did a simulation, both ends of shunt resistor, and trasducer were connected, (across resistance 10K) with voltage supply and ground and it worked fine. Cheers!

Last edited: Aug 10, 2009