These are the errors.
1)I cannot understand WHY this happens i coping the circuit from the datasheet but the result is wrong
2)Without diode the result is better (in the pic it is 2.63 and not 2.5 or at least near and i do not know how to trim this)
3)Can i delete the *10 scalling factor ?
If it has c2>c2 (for example 10 *2/1) the *10 makes the Vout=20 and the op amp gives error,How to solve this ?
Help ?!
Thanks

 

1) Don't know why you simulation is showing such a high error. My LTspice simulation shows only a small error.
Perhaps there's a problem with your model.

2) What's the supposed purpose of the diode?
I would expect that to give you a large error due to the diode voltage drop.

3) You could change the scaling factor by changing the relative value of R2 and R5.
Edit: For example, increasing the value of R5 to 100k ohm will reduce the scaling factor to 1.
But doing that may increase the output error even more.

You can provide some adjustment of the output level if you replace R2 with an 9k resistor in series with a 2k pot as shown below:

 

I tried your suggestions and thank you,it was good.
Multisim throws an error as the picture discribes ,any idea why,Is this my fault ?

 

Multisim throws an error as the picture discribes ,any idea why,Is this my fault ?

What does it do when you click on "Yes"?
Convergence problems are common in Spice simulations and the cause is often obscure.
Sometimes just changing a component value will solve the problem.
Other times changing some of the simulation parameters can help.