Comparitor problem with Multisim

[Cali engineer]

I am using the Multisim power pro 10.0.1 evaluation software.

When checking my circuit with the oscilliscope I noticed the comparitors do not sink to near ground or allow pullup to reach VCC, It should reach from almost ground to the high rail (+5v) with a pullup resistor. Does the simulator have a bug???

I am getting a output of about (2v to 3.5v) I should be getting about (.1v to 5v)

I tried using the virtual and actual lM339 and still the same results.

Can someone tell me If this is a bug in Multisim or is there a setting wrong?

Thanks

Bob

[lacy]

Coulod you post your test circuit for us to examine? I may know of a solution, but seeing your circuit could confirm my hypothesis

[Cali engineer]

I first noticed this problem with Multisim 2001, It was the same with all comparitors that I tried and

then I tried the same circuit with Multisim power pro 10.0.1 and got the same results again using the LM339AN.

Now I have found that The circuit will work correctly on power pro if I use the virtual comparitor or the LM339D.

Attached are two simple circuits, one using the LM339AN and the other with the LM339D.

If you adjust the pot past the crossover point you will get an ouput on the LM339AN of 1.1V to 3.8V

However if you try the same circuit with the LM339D you will get from 50 millivolts to 5V (This is what I would expect).

I would greatly appreciate any help understanding this difference.

Thanks

[lacy]

This was my hypothesis confirmed. If you look at the model for the one that doesn't work you will find that it says in the model file that it is an "EWB Version 4 5 Terminal Comparator Model". For some reason none of these models work correctly even in the standard op-amp category. You can spot them in the parts selection dialog box by looking at the model/info box on the right hand side. Most of these models will be labled "IIT/LM339"  The one that works is a model from Texas Instruments. To avaiod a situation like this I usually opt for the manufacturer model over the "IIT" model.

These model problems have been reported to NI and hopefully they will be corrected in the next update (whenever that may be).

[Cali engineer]

That is good information to know.

Thanks !!!