Doubt about NMOS Transistor in Microcap: Varunag

[varunag]

I was making a model of a NMOS transistor in Microcap. It was part of an assignment. I was supposed to make graph with changing value of V_{BS}. But after a certain value of V_{BS}, I got negative current. It came to my mind it is not possible. But, I couldn't think of a reason instantly, as to why current can't flow in the opposite direction.
Here V_D = 2.5V, V_G = 1.5V, source is grounded and V_B = 4V. I hope that a reasoning could be given without other parameters of the transistor, yet if other parameters are required I would give them.

I myself will look about this and share it with everyone.

-varunag

 

[Phrak]

It's fairly unclear what your test circuit looks like, but, yeah, a MOSFET will conduct equally well between drain and source, in either direction--according to bias.

 

[varunag]

I found the problem with the design. When we give positive supply to the body, we are making the body-source junction reverse bias.(I have not mentioned in my query that source is grounded, but I suppose that could be expected by the way I had given the values)
And on further increasing the body voltage, we make the body-drain junction reverse bias as well. Making these junctions reverse bias, we are changing the way the transistor works. I could only figure out this.

I hope someone could give a reason more suitable.

-varunag

 

[Phrak]

Not all mosfets are four terminal. Most commercial decretes are three terminal.
I don't know what "negative current" means.

 

[varunag]

Not all mosfets are four terminal. Most commercial decretes are three terminal.
I don't know what "negative current" means.

Well we know 4 terminal mosfets. And we usually short the body with the source. only this time we tried to give some voltage to the body, to check the theoretical relations we have. It seems the first order model used by "Microcap" software, is inadequate to deal with this, and we need to check with higher models.

 

[maverick280857]

Well we know 4 terminal mosfets. And we usually short the body with the source. only this time we tried to give some voltage to the body, to check the theoretical relations we have. It seems the first order model used by "Microcap" software, is inadequate to deal with this, and we need to check with higher models.

Not if MicroCap can be used for a circuit simulation with body effect considered. And that is described very briefly here: http://www.zenex.fi/microcap/files/9rm.pdf , on page 484. So there's probably a small glitch with your model.