Google Groups no longer supports new Usenet posts or subscriptions. Historical content remains viewable.
Dismiss

LowNoise Transistor with known KF (SPICE)

12 views
Skip to first unread message

siliconluvr

unread,
Jan 4, 2006, 3:20:27 PM1/4/06
to
Do you know of a low-noise small-signal audio transistor with published
KF?

(KF is the flicker noise parameter in SPICE - transistor data sheets
seem to cop out of providing solid info on this)

Jim Thompson

unread,
Jan 4, 2006, 3:25:16 PM1/4/06
to

Characterize the device yourself. The manufacturers don't give a
hoot.

...Jim Thompson
--
| James E.Thompson, P.E. | mens |
| Analog Innovations, Inc. | et |
| Analog/Mixed-Signal ASIC's and Discrete Systems | manus |
| Phoenix, Arizona Voice:(480)460-2350 | |
| E-mail Address at Website Fax:(480)460-2142 | Brass Rat |
| http://www.analog-innovations.com | 1962 |

I love to cook with wine. Sometimes I even put it in the food.

siliconluvr

unread,
Jan 4, 2006, 3:43:31 PM1/4/06
to
>From a 43-year industry veteran, this is probably good advice. Thanks!

How would one even start to do this characterization?

Are there some books/papers/urls that describe the dance?

Pooh Bear

unread,
Jan 4, 2006, 6:01:54 PM1/4/06
to

siliconluvr wrote:

> Do you know of a low-noise small-signal audio transistor with published
> KF?

Short answer.

No.

Now what did you really want ? There's more to life than Spice you know !

Graham

Pooh Bear

unread,
Jan 4, 2006, 6:02:24 PM1/4/06
to

siliconluvr wrote:

Why would you want to bother ?

Graham


Jim Thompson

unread,
Jan 4, 2006, 8:24:45 PM1/4/06
to

No there's not ;-)

Jim Thompson

unread,
Jan 4, 2006, 9:30:12 PM1/4/06
to

It's an equipment issue.

I haven't done it in nearly 40 years, so I'm not even sure what it's
called... noise meter, spot noise meter.

An audio spectrum analyser would also work, probably close enough to
fit a model.

Kevin Aylward

unread,
Jan 5, 2006, 1:23:57 AM1/5/06
to

Either the manufactures data has a plot of noise over frequency, or it
doesn't. Low noise transistor often have such a plot. In which case,
setting Kf is trivial. The noise is set by Kf/F^AF. Its not like there
are a billions of varibles to try. Try a few values plot the results and
compare!

I have actually manually set KF for a few low noise transistor in
SuperSpice. Which ones? Well, you will have to download SS and find
out:-)

Kevin Aylward
431info...@anasoft.co.uk
http://www.anasoft.co.uk
SuperSpice, a very affordable Mixed-Mode
Windows Simulator with Schematic Capture,
Waveform Display, FFT's and Filter Design.


Pooh Bear

unread,
Jan 5, 2006, 3:08:45 AM1/5/06
to

Jim Thompson wrote:

> On Wed, 04 Jan 2006 23:01:54 +0000, Pooh Bear
> <rabbitsfriend...@hotmail.com> wrote:
>
> >
> >
> >siliconluvr wrote:
> >
> >> Do you know of a low-noise small-signal audio transistor with published
> >> KF?
> >
> >Short answer.
> >
> >No.
> >
> >Now what did you really want ? There's more to life than Spice you know !
> >
> >Graham
>
> No there's not ;-)
>
> ...Jim Thompson

Yes there is ! :-p

Seriously, I see ppl agonising over their simulations when simply
breadboarding something would tell far far more. Each in its place.

Graham

Jim Thompson

unread,
Jan 5, 2006, 9:53:32 AM1/5/06
to
On Thu, 05 Jan 2006 08:08:45 +0000, Pooh Bear
<rabbitsfriend...@hotmail.com> wrote:

Ever tried breadboarding a 5000 transistor design, operating at 5GHz?

I grew up on breadboards, when an OpAmp was 20 transistors.

Now it is rare for a bias network to be as little as 20 transistors

Gerhard Hoffmann

unread,
Jan 5, 2006, 10:57:52 AM1/5/06
to
On 4 Jan 2006 12:20:27 -0800, "siliconluvr" <jvn...@gmail.com> wrote:


These are models of the Infineon (formerly Siemens) BC850 and BC860
AF transistors. I think there are newer models on their web site.

regards, Gerhard

(my first post with Agent)


*****************************************************************
* SPICE2G6 MODEL OF THE NPN BIPOLAR TRANSISTOR BC850 (SOT-23) *
* REV: 98.1 DANALYSE GMBH BERLIN (27.07.1998) *
*****************************************************************
.SUBCKT BC850C 1 2 3
Q 6 5 7 BC850 1.000
LC 1 6 0.350N
L1 2 4 0.400N
LB 4 5 0.500N
L2 3 8 0.400N
LE 8 7 0.600N
CGBC 4 6 70.00F
CGBE 4 8 0.150P
CGCE 6 8 15.00F
.ENDS
.MODEL BC850 NPN (IS=45.000F NF=1.010 BF=516.544 VAF=74.000 IKF=0.708
+ ISE=55.668F NE=2.567 NR=1.015 BR=7.745 VAR=14.000 IKR=1.000
+ ISC=1.084P NC=4.063
+ RB=9.000 IRB=0.100M RBM=4.500
+ RE=0.350 RC=1.445
+ CJE=13.050P VJE=0.690 MJE=0.375 FC=0.750
+ CJC=4.100P VJC=0.750 MJC=0.420 XCJC=0.650
+ TF=0.620N TR=2.5N PTF=1.000
+ XTF=68.000 VTF=1.000 ITF=0.720
+ XTB=1.400 EG=1.110 XTI=3.200
+ KF=10.000F AF=1.000)

*****************************************************************
* SPICE2G6 MODEL OF THE PNP BIPOLAR TRANSISTOR BC860 (SOT-23) *
* REV: 98.1 DANALYSE GMBH BERLIN (27.07.1998) *
*****************************************************************
.SUBCKT BC860C 1 2 3
Q 6 5 7 BC860 1.000
LC 1 6 0.350N
L1 2 4 0.400N
LB 4 5 0.500N
L2 3 8 0.400N
LE 8 7 0.600N
CGBC 4 6 70.00F
CGBE 4 8 0.150P
CGCE 6 8 15.00F
.ENDS
.MODEL BC860 PNP (IS=28.000F NF=1.000 BF=284.436 VAF=43.000 IKF=0.380
+ ISE=24.903F NE=2.234 NR=1.005 BR=4.800 VAR=6.960 IKR=0.932
+ ISC=0.125P NC=2.074
+ RB=2.200 IRB=0.100M RBM=1.500
+ RE=0.300 RC=2.251
+ CJE=11.800P VJE=1.000 MJE=0.435 FC=0.750
+ CJC=8.700P VJC=0.900 MJC=0.600 XCJC=0.650
+ TF=0.600N TR=2.604N PTF=1.000
+ XTF=6.500 VTF=2.000 ITF=0.314
+ XTB=1.600 EG=1.110 XTI=3.300
+ KF=5.000F AF=1.000)

Kevin Aylward

unread,
Jan 5, 2006, 3:20:51 PM1/5/06
to


Seriously, I see ppl agonising over their hand calculations, when simply
running a simulation...

I hardly ever do ohms law on paper nowadays. I can't be bothered. I just
want the answer. Gives me more time to get on with my music.

siliconluvr

unread,
Jan 5, 2006, 7:45:27 PM1/5/06
to
Gerhard, vielen Dank! Trust the German Engineers to dot the i's and
cross the t's.

0 new messages