Pspice Modeling of BJT curve Ib vs Vbe

[viren]

vbe on voltage data pspice

Hi,
I need to model the BJT again now to get the IB vs Vbe curve which I got by experimenting with the real BJT using a curve tracer. Now when i used the curve tracer i had kept the collector open. I gave a step current to Ib for 200mA . So can anybody have any idea how i can setup the circuit in pspice schematics for that.

 

[flatulent]

gummel plot in pspice

Since PSpice does not like open terminals, put a 1G ohm resistor from collector to ground.

 

[viren]

gummel plot pspice how to make

Now I have setup my circuit as follows. There is a current source at the base with step curent of 200mA and then a resistor of 1 G ohm is connected . So how do i measure for getting Ib vs Vbe curve . I am getting a linear curve having a slope of 1 when i run the probe for Ib vs V: Q1(b) (ie ) voltage across Base and emitter.
Since the Base emitter junction is acting like a diode it should give me a flat curve for voltage less than threshold voltage and then at the voltage when the diode turns on it should give a vertical straight line high current at that threshold voltage.
So can u advice where i am doing wrong or can u refer to someone who could help me out.

Added after 1 hours 48 minutes:

I am now able to model and get the Ib vs Vbe curve but the knee voltage of Vbe of the measured BJT model is 2.7 V and I am getting .8 V for my pspice model. What parameters should I change to get the knee voltage of Vbe to 2.7 V ?

 

[flatulent]

vbe bjt

There are several things involved. One is the extrinsic base resitance. This will control the slope at high currents. The other is the base-emitter junction area. This will control the voltage drop at a given low current.

Try some simulation experiments with these two parameters. You can scale the area by putting a number after the line describing your transistor. The extrinsic base resistance will be set by one of the model parameters. RB is the model parameter name.

 

[krashkealoha]

measuring spice parameters for bjts

Viren,

A measured 2.7V knee, for Ib vs. Vbe doesn't sound correct. It is about 2x too high.

The way you are getting an Ib vs. Vbe data from the curve tracer with the collector open is not a good way. I know you have a curve tracer to use. Do you have in your lab a HP or Agilent 4145, or 4155, or 4156? These tools are semiconductor parameter analyzers and they are more suited to get the data you are looking for.

Let me know and we can come up with a measurement method that best suits your available equipment, even if it is only a curve tracer.

Krash

 

[viren]

spice modelling of bipolar vbe

Hi,
I only have a curve tracer to do the BJT measurements.

 

[krashkealoha]

pspice vbe

Viren,

The best way to get your Ic & Ib vs. Vbe data is to set up your transistor test this way. Use a voltage source to apply 0V to the collector, and put a ammeter in series with it to measure the collector current. Do the same with the base, where you will apply 0V to the base and measure the base current with a ammeter. On the emitter apply a negative voltage from -0.4V and sweep it to -1.2V at 0.1V increments. For each increment of Vbe record the Ic & Ib. Once you are completed, you can plot your data, log Ic, log Ib vs. Vbe, which will be you gummel curve.

To optimize where the collector current, where it starts curving or deviating from being linear, the parameters which affect it are RB, RE, & IKF.

Good luck.

Krash

 

[viren]

vbe measurement

Hi,
Thank you very much for your suggestion. I have changed the values of ISE and ISC to a value between 0 and 1. So now I have to get the curve down flat between 0 V to 2.7 V and then it should begin to rise upto 2 V . So my question is how to get the slope down to zero between 0 and 2.7 V .
Let me know if u have a suggestion.
Viren

Added after 2 minutes:

I have attahced the graph in this message

 

[krashkealoha]

pspice ib

Viren,

This is your simulated data.

Could you post your measured Ic & Ib data versus Vbe. I can show you how to measure the critical SPICE parameters from that data. Please post both IC & IB measured data.

Thanks,
Krash

 

[viren]

transistor parameters vbe pspice

Hi ,
I have attached the IC vs Vce plot . I did not measure the Ic vs Vbe for the device. I also when I try to convert the curve tracer plot to excel file format it does not show me the plot but the the Ib vs Vbe plot is similar to one I am going to attach but its slope is not zero between 0 V and 2.7 V as compared to my measured Ib vs Vbe plot.

 

[krashkealoha]

spice model vbe

Viren,

If you are not going to measure the Ic vs. Vbe curve, then you are on your own. You cannot get the measured vs. simulated to match unless you measure Ic vs. Vbe. Please read the gummel-poon modelling document that I posted, which you downloaded. It has all the answers you are looking for.

 

[viren]

bjt characteristic curve excel

The Test setup you have mentioned is not using curve tracer. I can use only a curve tracer here. So if u have a suggestion for setting up a test using a curve tracer for the Ic vs Vbe measurement let me know.
thanks

Added after 2 hours 21 minutes:

Hi,
I have to model the Ic vs Vce and Ib vs Vbe plot of a BJT device using the same parameters . I had to change the values of IS and ISE to model the Ib vs Vbe plot . So how do I calculate the parameters which can give me both the plots .

 

[krashkealoha]

bjt curve tracer in spice

So you are telling me you don't have a voltage source and an ammeter in your lab???

On the curve tracer, plot an Ic vs. Vce curve for constant Vbe values. Your set up should have the base and collector at the same potential (but not shorted) when you step Vbe. Since Vcb=0V => Vce = Vbe on the Ic vs. Vce curve (constant Vbe steps). Record Ic for 0.1V Vbe(Vce) increments. This will be your Ic vs. Vbe data. You can then use your Ic vs. Vce curve (constant Ib steps) to get the Ib vs. Vbe data.

 

[viren]

change vbe voltage pspice

How do I keep the potential of base and collector at same potential ? . I have a 370B tektronix curve tracer . What connections or what configuration should I use on this curve tracer to set it up for my measurements.
I connected one wire each to base ,collector,emitter. then used the horizontal div knob and adjusted the collector volt to 2 volt . and using the step gen knob I adjusted the base to 2 V using 10 steps each of 200mA . then I varied the collector supply knob and it did show any graph.
Let me know if I am doing wrong or refer to someone.

Added after 2 hours 36 minutes:




How do I keep the potential of base and collector at same potential ? . I have a 370B tektronix curve tracer . What connections or what configuration should I use on this curve tracer to set it up for my measurements.
I connected one wire each to base ,collector,emitter. then used the horizontal div knob and adjusted the collector volt to 2 volt . and using the step gen knob I adjusted the base to 2 V using 10 steps each of 200mA . then I varied the collector supply knob and it did NOT show any graph.
Let me know if I am doing wrong or refer to someone.

Added after 2 hours 17 minutes:

I have the voltage sources but the BJT device which I am using is made of silicon and the transistor turns on when the Vbe is greater than 2.7V . You gave me the increments in -.4 V to -1.2V which doesn't look right. And also how do I know what current should I apply at the base and collector so that the transistor turns on.
so let me know if u have any idea or if i am wrong.

 

[viren]

vbe of bjt

hi,
I have a circuit to calculate the gummel plot Ln(Ic) Ln(Ib) vs Vbe . I connected the base and collector together. I connected a 10 ohmresistor to the base . Then I connected a current source to the resistor and emitter was grounded. I am measuring the collector current and base current. Now when I applied a current of 3 A to he base the resistor got burned . So I need to know how to calculate the value of a resistor that I should apply to the base .
If u have any idea or suggestion let me know.
Viren

 

[viren]

how to plot characteristic curves in pspice

hi,
I have a logarithmic plot of Ib ,Ie vs Vb attached with the message. Please let me know if u know how to extract parameters from the plot.
viren

 

[krashkealoha]

ib x vbe curve

Since the turn on voltage is 0.7 - 0.8V, stepping Vbe by 1 volt increments is too large. You need to step Vbe by 0.1V, which is the maximum I would recommend. If you can, step it by 50mV increments.

You might want to look at the output characteristics of your part you are testing to make sure you are still getting the proper transistor characteristics. A Vbe=6V might damage the device.

 

[viren]

plot bjt curves

hi,
I have attached the excel file of gummel plot in the message. But when I measure Vbe for negative voltage from 2V to 3.7V I get a message that negative values cannot be plotted in log scale and then after i ignore it it shows an empty plot.
Let me know if I am wrong or my measurements are wrong as i have them in the table. Sometimes the base current is larger than the collector curent and sometimes it is not.
If u have any suggestion let me know.
thanks
viren

 

[krashkealoha]

bjt curves spice

I believe you are measuring at too high of a Vbe value. Remember the emitter-base junction is a diode so it should turn on in the 700-800mV range. Therefore, your Vbe of interest should be from 400mV to 1200mV. Going beyond 1200mV the transistor will start-going into high injection and start saturating. I'm speculating that the reason you are seeing a higher Ib than Ic at the large Vbe is that the transistor is in saturation due to the high collector current dropped across the internal collector resistance of the bipolar.

You gummel plot should have log IC & IB on the y-axis and linear Vbe on the x-axis.

 

[viren]

graph of ib versus vbe

hi,
I have tried to measure at the range of 700mV-1200mV but could not get any current . It is because the SiC BJT has threshold voltage at 2.7 V and I have already measured in that range and also did not get good data . So now I would like to know is it possible to do it some other way other than using the voltage source and curve tracers.
Is it possible to get gummel poon plot by testing the BJT device on 4140 HP Semiconductor tester. And if it is possible I need to know how to measure it using that machine.
If u have any idea let me know urgently.
Thanks
viren

Added after 52 minutes:

Hi,
I know the one link which u sent earlier in beginning is for bipolar modeling. But if there is another better than that then let me know.
Thanks
Viren