OPAMP Peoblem

Thread Starter

sanketbarot

Joined Dec 12, 2006
14
Hello Friends,


I am using LM358 as differential amplifier..
Input to -IN is 103.6mv and +IN is 104.6.

gain of the differential amplifier is unity..

so my question is with this input accordigto the trnsfer function it should give me 1mv output..
But when I check with Multimeter it shows me -5.5mV..

Why this is happening? and even my VDD is +5V and Vss is ground then how can it shows me -ve voltage????

I kept 100Ohm fixed resistor inplace of RTD(PT100)

As you can see the circuit..

OPAMP.jpg

My voltage measurmnet is as follows..


Test Point Voltage
T0 103.6mV
T1 104.8mV
T2 83.3mV
T3 53.2mV
 

Distort10n

Joined Dec 25, 2006
429
Well...

First thing, you do not happen to have the leads switched on your meter do you? Everyone makes simple mistakes.

I ask because you say that you are using a 5V single supply and you expect to see 1mV on the output. Ideally this is true; however, practical limitations of the op-amp output stage will not exhibit the ideal parameters everyone learns during their academic career.

If you look in the datasheet for the LM358 under Low Level Output Voltage on pg. 5 you will see that the lowest voltage on a single supply spec'ed with 10k ohm load is 5 mV typical. This is also known as 'headroom.'

Check those first. If not, post a better picture of your schematic.
 

hgmjr

Joined Jan 28, 2005
9,027
I too could not obtain any useful information from the image you posted as it did not permit me to zoom in for greater detail.

Hopefully you can re-post an image that can provide the needed details.

hgmjr
 

Distort10n

Joined Dec 25, 2006
429
There are several things at work here to explain the non-linearities and they all come from pain of using a single supply.

I mentioned one already. You simply will not see 0V on the output of a single supply op-amp. This is impossible because no amplifier from any manufacturer has literal rail-to-rail performance. Some get very close; i.e., 10mV from the rail.

Your rails here are 5V and 0V (GND). The LM358 has a typical head room of 5mV. That means even if you where to assume all other golden rules of op-amps to be ideal (infinite input impedance, infinite bandwith, zero output impedance etc.) you will see no less than 5mV typically. Your 1mV expected output is 'buried in the headroom.' So 5.5mV on the output is not suprising although a negative voltage is. Either you have the leads on your meter switched or something else is going on.

This also happens with offset voltages with single supply op-amps as well. Since head room is larger in magnitude than offset, you will not see the intial offset on the output because it is also buried in the headroom. You can still see the affects, but that is another topic.

Another thing that could be happening is the op-amp cannot servo the summing junction to equal the voltage at the non-inverting node. I have not come across this much but since it is single supply the input headroom is just as critical as the output headroom previously mentioned. The op-amp cannot drive the summing junction further down than the input headroom specification. I am not exactly sure if this is the mechanism at work here so I will ask around and get back to you. The LM348 says that the input common mode range includes 0V but I have my suspicions that it is not exactly 0V.

I have attached a spice model showing how the voltages at the summing junction and non-inverting nodes behave on single and dual supply. The output voltage is a little high but since this is Spice and the model is old, I am not too concerned. This is where testing an actual device would serve better.

In short, use a dual supply and I bet you the issue *POOFS* into thin air.
 

Attachments

Thread Starter

sanketbarot

Joined Dec 12, 2006
14
Thank you friends,


I will work on it and I will get back to you in Few Days..

*************Wish You all Happy New Year***********


Regrds,
Sanket Barot
 

Thread Starter

sanketbarot

Joined Dec 12, 2006
14
Hello Friends,


I was using Hitachi's HA17358 instead of LM358. But when I connect the LM358 of ST make then it is working good. Still it is giving me some offset voltage when both input is grounded. but I will try to overcome this problem by my software.

so this may be the problem because of the make.. and even for good result as you all guided me that analog ground and digital groung sholud be separatly routed. so I am working on it an after few days I will get the PCB in my hand. and I will get back to u again..


Regards,
Sanket Barot
 
Top