Bob Cordell's Power amplifier book

www.hifisonix.com
Joined 2003
Paid Member
I've been to Croatia. About 4 years ago, I got an email at work:-

"you will make yourself available at 9:30am at Munich (IIRC) railway station on this (date). Wait at the diner located on the LHS of the plaza."

I was based in Japan and knew it was some leadership related thing for work. So I booked my flights and got there, meeting a whole lot of work colleagues at the diner and getting a nice coffee as well. We took trains/flights etc to Split and then hiked, drove and took trains around Croatia for about 4 days. The last part of the journey we went down to the coast and got picked up by a ferry. We landed on this big island off the coast and spent two days in yachts sailing around the Adriatic (I kid you not). The captain of our yacht (there were 6 of us per yacht) told us lots of amusing stories about the Italians - seems they have a thing going between them a bit like the Irish and the English. After that, we flew back to Munich and went our separate ways again.

Croatia is stunningly beautiful - awesome trip!
 
Member
Joined 2011
Paid Member
spice EG parameter

Dear Bob,

is it possible, that the EG parameter of the model 2SC3503C may be wrong? I have some thermal troubles using a 2SC3503C in a Bias generator compared from simulation to real life (using the base/emitter diode as temperature sensor). If using the BD139C model the difference between simulation and real life is smaller.

Thanks!

BR, Toni


* 2SC3503C - created March 11, 2011 copyright Cordell Audio
.MODEL 2SC3503C npn
+IS=40e-15 BF=170 VAF=769
+IKF=0.08 ISE=200e-15 NE=1.5 NF=1.0
+RB=75 RC=1.5 RE=0.1
+CJE=95e-12 MJE=0.35 VJE=0.75
+CJC=7e-12 MJC=0.35 VJC=0.75 FC=0.5
+TF=585e-12 XTF=10000 VTF=35 ITF=20
+TR=10e-9 BR=0.6 IKR=0.05
+EG=0.75 XTB=1.5 XTI=3 NC=1.5
+ISC=7e-15 NR=1.0 VAR=100 IRB=3e-6
+RBM=0.035 XCJC=1.0 mfg=CA031111

* BD139C - created April 13, 2011 copyright Cordell Audio
.MODEL BD139C npn
+IS=150e-15 BF=260 VAF=99
+IKF=1.2 ISE=70e-15 NE=1.2 NF=1.0
+RB=5 RC=0.01 RE=0.08
+CJE=293e-12 MJE=0.33 VJE=0.67
+CJC=49e-12 MJC=0.39 VJC=0.52 FC=0.5
+TF=585e-12 XTF=10000 VTF=35 ITF=20
+TR=10e-9 BR=78 IKR=0.14
+EG=1.21 XTB=1.14 XTI=3 NC=1.45
+ISC=19e-12 NR=1.0 VAR=7.5 IRB=0.03
+RBM=0.001 XCJC=0.53 mfg=CA041311
 

Attachments

  • spice_temp_parameter_range.png
    spice_temp_parameter_range.png
    29.8 KB · Views: 424
Dear Bob,

is it possible, that the EG parameter of the model 2SC3503C may be wrong? I have some thermal troubles using a 2SC3503C in a Bias generator compared from simulation to real life (using the base/emitter diode as temperature sensor). If using the BD139C model the difference between simulation and real life is smaller.

Thanks!

BR, Toni


* 2SC3503C - created March 11, 2011 copyright Cordell Audio
.MODEL 2SC3503C npn
+IS=40e-15 BF=170 VAF=769
+IKF=0.08 ISE=200e-15 NE=1.5 NF=1.0
+RB=75 RC=1.5 RE=0.1
+CJE=95e-12 MJE=0.35 VJE=0.75
+CJC=7e-12 MJC=0.35 VJC=0.75 FC=0.5
+TF=585e-12 XTF=10000 VTF=35 ITF=20
+TR=10e-9 BR=0.6 IKR=0.05
+EG=0.75 XTB=1.5 XTI=3 NC=1.5
+ISC=7e-15 NR=1.0 VAR=100 IRB=3e-6
+RBM=0.035 XCJC=1.0 mfg=CA031111

* BD139C - created April 13, 2011 copyright Cordell Audio
.MODEL BD139C npn
+IS=150e-15 BF=260 VAF=99
+IKF=1.2 ISE=70e-15 NE=1.2 NF=1.0
+RB=5 RC=0.01 RE=0.08
+CJE=293e-12 MJE=0.33 VJE=0.67
+CJC=49e-12 MJC=0.39 VJC=0.52 FC=0.5
+TF=585e-12 XTF=10000 VTF=35 ITF=20
+TR=10e-9 BR=78 IKR=0.14
+EG=1.21 XTB=1.14 XTI=3 NC=1.45
+ISC=19e-12 NR=1.0 VAR=7.5 IRB=0.03
+RBM=0.001 XCJC=0.53 mfg=CA041311

In my SPICE modeling I did not adjust EG from the manufacturer's models. It does look like the value here of 0.75 is on the low side, as EG for silicon is most often in the range of 1.1. Try changing it to 1.1 and see what you get. However, if I am not mistaken, that may interact with IS in such a way that simulated Vbe at room temperature may change.

Cheers,
Bob
 
Member
Joined 2011
Paid Member
In my SPICE modeling I did not adjust EG from the manufacturer's models. It does look like the value here of 0.75 is on the low side, as EG for silicon is most often in the range of 1.1. Try changing it to 1.1 and see what you get. However, if I am not mistaken, that may interact with IS in such a way that simulated Vbe at room temperature may change.

Cheers,
Bob

Dear Bob,

thanks for this information - this explains the problem very well.
FYI: Just tested 2SC3503C model with EG=1.11 .
This seems to fix the differences. Now the simulated bias change from temperature 22 - to 50 degrees is nearly exactly the same as the real life measured one.

BR, Toni
 
Hi
during reading Bob's book I found a possibly mistake,
am I right the circuit should be like amended in attachment?

Hi Pawel,

Thanks for pointing this out. I always appreciate feedback on the book, be it in regard to errors, or sections that need better explanation, or technical assertions I have made that one does not agree with. Many people on this forum have been very helpful in this regard, and I had been previously made aware of this error. I have been keeping a list of all such errors so that I can make corrections in the second edition.

Best regards,
Bob