Problem with implementing Spice model of Infinion BFP640

Hi

I have tried to implement the spice model of the mentioned transistor. When I make a simulation something strange happens. I get a lot of jitter or noise and the data points jumps up and down a lot. It seems that if you average the curve/points you get something reasonable.....but I cannot use it.

I have used a Gummelpoon NPN with substrate effect to make the model.

Anybody know what is going on?

Regards

Archive your design files with WinRAR and upload here so we can take a look.

Hi

Hope you can take a look.
The design is made en Designer.

I have also attached a S-parameter file which I use for comparison.
Thank you

I found this model in one of my libraries that might help you.

.MODEL BFP640 NPN(
+ IS = 0.61e-16 BF = 450 NF = 1.025
+ VAF = 1000 IKF = 0.47 ISE = 6.2e-14
+ NE = 2 BR = 42 NR = 1
+ VAR = 2 IKR = 18 ISC = 0.7e-17
+ NC = 1.8 RBM = 0.895 IRB = 4.548e-05
+ RB = 1.036 RE = 0.2 RC = 1.006
+ CJE = 682.5e-15 VJE = 0.8 MJE = 0.3
+ TF = 1.9e-12 XTF = 10 VTF = 1.5
+ ITF = 1.25e-05 PTF = 0 CJC = 2.046e-12
+ VJC = 0.6 MJC = 0.5 XCJC = 1
+ TR = 2e-09 CJS= 294.9e-15 VJS = 0.6
+ MJS = 0.27 XTB = -1.42 EG = 1.078
+ XTI = 3 FC = 0.75835)

Hello

Thanks for the reply.

Vfone I have tried your SPICE model but I seem to get the same problem as I started with. Which is a lot of jitter in my plots. Also the bias is very strange....Vbe have to very low, otherwise the current in the device is exremly high.
I do not know if the model works for you?

Regards

this model works fine using mwo

Hi vfone

If I use your SPICE model in Designer and make a S-parameter analysis using RF-chokes and DC-coupling capacitors. If I have a Vce of 3V and a Vbe of 0.7V to 0.8V, the current in the device is nearly 2 Amps. SO already there something is really wrong. If I decrease Vbe to get a more realistic Ic, it has to be as low as 0.1V.

I have put on the result of S11 just to show what my data looks like. If I implement the transistor using the model from infinions datasheet it looks better but still really wrong. The large jumps back and forth in the plot is also present in model I made from infineon, but here the DC chracteristics seems realistic. But if someone could take a look I would be very pleased.

Thank you

a resistor in series with the collector bias is missing. You've connected +3V directly through the choke to the collector...

add a resistor between the choke and Vcc. Set Vcc correspondingly higher to get +3V at collector.

Hi

First I would like to thank the one who replied my post.

I have found out what the problem was. It was the DC-block and RF-chokes I used. In Designer these are set to 10000H and 10000F, if these are set to a much lower value of 10H and 10F everything is working.