r2spice function use and calculation of alpha, XIg and Glim
Dear Mr. Nakabayashi,
is there a demo somewhere, with which version of RStudio your r2spice program works?
Parameters a, b and c seem to get calculated using alpha. However it is not exactly described where
alpha comes from - https://ayumi.cava.jp/audio/pctube/node43.html. I assume its discussed more in detail in Yoshinori Ichiki book, however its in japanese and hard to come by.
Also for calculationg xig and Glim its written that xig = Ig/Ik when Vg=Va. So cathode and grid current ratio when plate and grid bias voltage are (positive) and the same. There is no calculation inside of r2spice of these values, yet they are as numbers inside of for example 12AX7 subcircuit.
Regards
Hi DSU,
Sorry, I don't use RStudio. Instead, I use R 4.4.2 command line interface.
Please try to run following code on your RStudio.
> getOption("encoding")
[1] "cp932"
> source("pctube.r")
> Ip.cal("12AU7.csv", "t12AU7", Cgp=1.5e-12, Ci=1.6e-12, Co=0.4e-12)
Read 153 items
code=0 err=0.02626518
> r2spice(t12AU7, "t12AU7", "t12AU7sample")
Writing spice/tubemodel/t12AU7sample.inc
Writing simetrix/tubemodel/t12AU7sample.mod
Writing tina/tubemodel/t12AU7sample.inc
If this code doesn't work on RStudio, try to run it on R 4.4.2.
As you mentioned, a, b and c is derived from alpha.
alpha is the main parameter and calibrated by the plate current characteristics curves.
xg (Ig.ratio in function r2spice) and Glim (G.lim in r2spice) are
parameters given when you run Ip.cal or if they are missing,
their estimates are calculated inside r2spice.
They themselves are not used directly in spice model,
but they are multiplied by another value.
This multiplication makes spice simulation slower,
so the multiplication is done in r2spice in advance,
then the result is written in spice model file.
