AetherWerx 11-17-2013 10:08 PM

Bipolar transistor gain

1 Attachment(s)
So I have a simple circuit in LTSpice that is supposed to plot Vce vs Ic for a single Ib.

I used the 2N222A which has a forward gain of 200 and the base current is set using a current source of 100u. So, the collector current should be 20mA except it is not. It is 19mA. This only gets worse if I increase the base current. For example, a base current of 1mA, should yield 200mA. According to the datasheet the maximum collector current is 800mA. However, the sim shows roughly 137mA.

What gives?

 MrChips 11-17-2013 10:19 PM

What is the resistance of R1?

A resistance of 1μ doesn't make sense. What are the units?

 MikeML 11-18-2013 12:02 AM

1 Attachment(s)
Look at this LTSpice sim. I am using the parameters for the 2N2222 that are supplied with the download.(I cant read yours).

Look at the red trace, which is a plot of the expression Ic(Q1)/Ib(Q1). Note it's dependence on the magnitude of collector current Ic(Q1). The left vertical axis is in Amps, the right vertical axis is the gain (dimensionless).

 AetherWerx 11-18-2013 12:53 AM

Quote:
 Originally Posted by MikeML (Post 667052) Look at this LTSpice sim. I am using the parameters for the 2N2222 that are supplied with the download.(I cant read yours). Look at the red trace, which is a plot of the expression Ic(Q1)/Ib(Q1). Note it's dependence on the magnitude of collector current Ic(Q1). The left vertical axis is in Amps, the right vertical axis is the gain (dimensionless).

I am also using the LT model of the 2N2222, but the only parameter I changed was VAF to ensure a flat collector curve (Ic vs. Vce).

So what parameter in a BJT is causing this roll off in beta with higher collector currents.

I still have the Ic = Beta * Ib as gospel, meaning beta is constant.

 MrChips 11-18-2013 01:15 AM

Who told you beta is constant?
Beta is never constant.

 AetherWerx 11-18-2013 02:22 AM

Quote:
 Originally Posted by MrChips (Post 667071) Who told you beta is constant? Beta is never constant.
The simulator is set to 27 degrees, so there is no variation against temperature causing a problem. Since I have VAF set to 10000 the Early effect should be eliminated for all practical purposes.

Or is it simply an issue with the details in the model?

 MrChips 11-18-2013 02:41 AM

I have never encountered a transistor that had a constant beta.

 ScottWang 11-18-2013 02:55 AM

1 Attachment(s)
You can measure the attached circuit and record it, set the Ib as 0.1mA~1mA,2mA,5mA,10mA, and to see how the Ic changing, also to see how the hfe changing.

 MikeML 11-18-2013 03:26 AM

Quote:
 Originally Posted by ScottWang (Post 667093) You can measure the attached circuit and record it, set the Ib as 0.1mA~1mA,2mA,5mA,10mA, and to see how the Ic changing, also to see how the hfe changing.
Scott, I suggest you look at the LTSpice sim I posted in #3, above. I did all you suggest and more. Look at how simple it is to do in LTSpice. Note how easy it is to plot "beta". Note how easy it is to span multiple decades of variables, plot different dependent variables vs an independent variable, make the plots log vs log, etc, etc...

 ScottWang 11-18-2013 04:20 AM

Quote:
 Originally Posted by MikeML (Post 667105) Scott, I suggest you look at the LTSpice sim I posted in #3, above. I did all you suggest and more. Look at how simple it is to do in LTSpice. Note how easy it is to plot "beta". Note how easy it is to span multiple decades of variables, plot different dependent variables vs an independent variable, make the plots log vs log, etc, etc...
I knew that, but if the people get involve too deep in the virtual world, it may lost many things in the real world, special to the beginner.

It just like I was too depends on google, so if there is no google and then I can't do many things.

So that's why I want the op to do that in the real practice, it will help a beginner to learn how a circuit working in a real electronics world.

All times are GMT. The time now is 01:22 PM.