Temperature sweep analysis problem

[asandvand]

Hello

I am having problem getting even the simplest  temperature sweep analysis to work in multisim 10.

As an example, please see attached circuit and pictures.

The circuit contains a simple voltage divider where R1 has a TC1 of 1 ohm/C, and as shown in figure 1, the resulting voltage is 0.5V when running simulation at

27degree C with TNOM 27C.

Raising the Temperature to 28degree C results in a resulting voltage of 0.33V (figure 2) (indicating  a doubling of the R1 value because of the 1 degree temperature raise with a TC1 of 1ohm/C)

To my understanding, the R1 value should have changed with only 1 ohm.

Any hints on what might be the problem with my temperature analysis ??

Does anybody know if this is a known error with Multisim ??

Any advice/hint is appreciated !

[lacy]

You are correct in you observation. The value is effectively being doubled. This is apparently a software error that will have to be further addressed by NI. I am responding only verify your findings (2 or more people reporting the same results are always better than just one person reporting it).

I have no answer as to how to work around this. NI will have to read this and maybe they can come up with a temprary solution until a permanent one can be found.

[Øyvind Ø]

This was reported to R&D (#103149) for further investigation.

Thanks for the feedback!

Øyvind Ø

[OStep]

The problem is with the units label in the resistor dialog box, not with the result. tc1 actually has units [1/degC] and once multiplied by some temperature difference, acts only as a factor.

The formula for resistance is as follows:

R = Rnominal *(1+tc1*(T-Tnom) + tc2*(T-Tnom)^2), where in your case  tc1=1, tc2=0 and Tnom=27 and T=28

therefore R=2*Rnominal

This defect will be registered

Thanks,

[asandvand]

Thank's

Using the unit [1/degC] for the TC seems to give results as expected

Åsmund

[lacy]

Asmund, please forgive me for being stupid at the moment, but could you explain how you set this up to get the results you needed as I am still a little fuzzy on this. Thanks for any information you can give.

[asandvand]

OK

1 Ohm/degC at 5kohm nominal value is 1/5000 = 0,0002 [1/degC] (or 0,02% or 200ppm)

I then insert 0,0002 as TC1 and the temperature behaviour works as expected.

Of course, a temperature change of only 1 degree C results in almost 0 change, but that was only an example and my real simulations uses higher temperature differences.

Hopes this clarifies the behaviour.

Regards, Åsmund

[lacy]

Thank you. I was thinking in that direction, but wasn't sure I really appreciate your response to my inquiry. This is something in my opinion that Multisim should automatically calculate. I know NI is already aware of the situations with this, but I thought I would let my opinion be known.

Thanks again.