Hello,



For those who need it, here are two (or four, depending on your view) 3f4
spice models for the 71A direct heated triode. One version without filament
and one with filament. The filament version has an extra pin for the
filament (pin f), the other end of the filament is pin k (I was too lazy to
call them f1 and f2). I have used these models in Circuitmaker 2000 and
especially the ones based on Norman Koren's model are good.



I hope that they are any useful.



With kind regards,

Simon van de Laak



*71A direct heated triode 5V 0.25A version without filament pkg:UX4 1,2,3
..SUBCKT 71A a g k
* Data from the RCA datasheet for use with 3f4 spice
* Made by Simon van de Laak, august 2005
* Although made with care, this model is provided "as is" and
* I give no warranty for its use in any application.
* Please feel free to use it and improve it.
*best model:
Bout a k i=(1/5699.167193)*((v(a,
k)/35.3634889)*ln(1+exp(35.3634889*((1/3.060930456)+v(g, k)/v(a,
k)))))^1.635475676
*simple model:
*Bout a k i = ((uramp((v(a, k)/3.72062)+v(g, k)))^1.5)/4582.72
Cgk g k 3.7e-12
Cga g a 7.4e-12
Cak a k 2.1e-12
..ends 71A


*71A direct heated triode 5V 0.25A version with filament pkg:UX4 1,2,3,4
..SUBCKT 71Adht a g k f
* Data from the RCA datasheet for use with 3f4 spice
* Made by Simon van de Laak, august 2005
* Although made with care, this model is provided "as is" and
* I give no warranty for its use in any application.
* Please feel free to use it and improve it.
*best model:
Bout a kp i=(1/5699.167193)*((v(a,
kp)/35.3634889)*ln(1+exp(35.3634889*((1/3.060930456)+v(g, k)/v(a,
kp)))))^1.635475676
*simple model:
*Bout a kp i = ((uramp((v(a, kp)/3.72062)+v(g, kp)))^1.5)/4582.72
Cgk g kp 3.7e-12
Cga g a 7.4e-12
Cak a kp 2.1e-12
Bheat1 f kp i=(v(f, kp))/10
Bheat2 kp k i=(v(kp, k))/10
..ends 71Adht

Hmm,

By one or another reason, some returns turn up in my posting on some
critical points in the model, especially on the places where a comma shows
up.....

Simon


"Simon" schreef in bericht
l...
Hello,



For those who need it, here are two (or four, depending on your view) 3f4
spice models for the 71A direct heated triode. One version without
filament
and one with filament. The filament version has an extra pin for the
filament (pin f), the other end of the filament is pin k (I was too lazy
to
call them f1 and f2). I have used these models in Circuitmaker 2000 and
especially the ones based on Norman Koren's model are good.



I hope that they are any useful.



With kind regards,

Simon van de Laak



*71A direct heated triode 5V 0.25A version without filament pkg:UX4 1,2,3
.SUBCKT 71A a g k
* Data from the RCA datasheet for use with 3f4 spice
* Made by Simon van de Laak, august 2005
* Although made with care, this model is provided "as is" and
* I give no warranty for its use in any application.
* Please feel free to use it and improve it.
*best model:
Bout a k i=(1/5699.167193)*((v(a,
k)/35.3634889)*ln(1+exp(35.3634889*((1/3.060930456)+v(g, k)/v(a,
k)))))^1.635475676
*simple model:
*Bout a k i = ((uramp((v(a, k)/3.72062)+v(g, k)))^1.5)/4582.72
Cgk g k 3.7e-12
Cga g a 7.4e-12
Cak a k 2.1e-12
.ends 71A


*71A direct heated triode 5V 0.25A version with filament pkg:UX4 1,2,3,4
.SUBCKT 71Adht a g k f
* Data from the RCA datasheet for use with 3f4 spice
* Made by Simon van de Laak, august 2005
* Although made with care, this model is provided "as is" and
* I give no warranty for its use in any application.
* Please feel free to use it and improve it.
*best model:
Bout a kp i=(1/5699.167193)*((v(a,
kp)/35.3634889)*ln(1+exp(35.3634889*((1/3.060930456)+v(g, k)/v(a,
kp)))))^1.635475676
*simple model:
*Bout a kp i = ((uramp((v(a, kp)/3.72062)+v(g, kp)))^1.5)/4582.72
Cgk g kp 3.7e-12
Cga g a 7.4e-12
Cak a kp 2.1e-12
Bheat1 f kp i=(v(f, kp))/10
Bheat2 kp k i=(v(kp, k))/10
.ends 71Adht

"Simon" wrote

For those who need it, here are two (or four, depending on your
view) 3f4
spice models for the 71A direct heated triode. One version without
filament
and one with filament. The filament version has an extra pin for
the
filament (pin f), the other end of the filament is pin k (I was
too lazy to
call them f1 and f2). I have used these models in Circuitmaker
2000 and
especially the ones based on Norman Koren's model are good.

I hope that they are any useful.

Thanks Simon!
How did you match the model to the data? Trial and error by eye, or
multiple regression?

cheers, Ian

Hi,

I used an excel spreadsheet with some macro's to capture the data points.
The solver in excel calculates the needed parameters by optimizing to the
lowest difference in respect with the calculated value and the captured
value. It is also possible to use this spreadsheet with diodes, and there is
one special spreadsheet for penthodes. I downloaded this spreadsheet a long
time ago, but the site where it used to be is gone :-(

I can send you to excel files if you like.

With kind regards,

Simon van de Laak


"Ian Iveson" schreef in bericht
. uk...
"Simon" wrote

For those who need it, here are two (or four, depending on your
view) 3f4
spice models for the 71A direct heated triode. One version without
filament
and one with filament. The filament version has an extra pin for
the
filament (pin f), the other end of the filament is pin k (I was
too lazy to
call them f1 and f2). I have used these models in Circuitmaker
2000 and
especially the ones based on Norman Koren's model are good.

I hope that they are any useful.

Thanks Simon!
How did you match the model to the data? Trial and error by eye, or
multiple regression?

cheers, Ian

"Simon" wrote

I used an excel spreadsheet with some macro's to capture the data
points.
The solver in excel calculates the needed parameters by optimizing
to the
lowest difference in respect with the calculated value and the
captured
value. It is also possible to use this spreadsheet with diodes,
and there is
one special spreadsheet for penthodes. I downloaded this
spreadsheet a long
time ago, but the site where it used to be is gone :-(



Ahah. I forget about solver because I haven't got it and don't
understand why not. I have used demo copies of regression tools but
they tend to have limited data capacity and the full versions are
expensive.

I can send you to excel files if you like.

Thanks, but no use without solver!

Have you tried to find or make a good saturable inductor core yet?
That is a key contribution someone should make. I was nearly there
then lost my thread and need inspiration to get a fresh grip.

Thanks for the model.

cheers, Ian