Page 1 of 1
842 Triode Spice
Posted: Thu Sep 13, 2012 6:22 pm
by jarek
Hi all! I came across a very interestingly looking triode
from the 800-family: 842
*Moderate Ra @ 2.5k
*12W dissipation (or more if someone wants steroid KR tube version)
..anyone has a Spice model? The usual suspect Stephie
?
Thanks!
-j
Posted: Mon Sep 17, 2012 7:22 pm
by sbench
As I get an opportunity, I'll create a model and update the files... probably a few days. Unless, of course, someone wants to use that build model I posted and submit the resulting parms.
Stephie <3
Posted: Wed Sep 19, 2012 2:00 pm
by sbench
Adding 842 to the Triode drop down list. Enjoy,
Stephie <3
Posted: Thu Sep 20, 2012 10:27 pm
by jarek
Heey Stephie! Thank you very much!
Unfortunately trying to simulate I get a "Timestep too small"
error. Never had it with other models.
Wondering why...
PS Apparently the trouble is indicated with 6E5P tube, which so far has been working 100% ok with other models from triode_nh (801A, 2A3, 300B)
Posted: Fri Sep 21, 2012 4:34 pm
by jarek
OK, it has started to work. No idea what was wrong, but it works now
Thanks again Stephie!
Posted: Sat Sep 22, 2012 5:37 pm
by sbench
I find that when I do some very complex simulations, the matrix math seems to get complex enough internally to cause those kind of problems. It gets particularly frustrating when you're using LT built in op amps and whatnot.
Sometimes using GEAR analysis will relieve the problem. Also, try the .noopiter Spice directive.
I usually try all the devices once I've added them to make sure I didn't mess up somewhere when I entered the data.
At some point (yeah, my bucket list just keeps getting longer) I intend to take the models to a new level. With the current "simple" tube models, there are a lot of compromises as you simply can't model the "edges" very well. I feel that LTs built in BV and BI behavioural sources ought to be able to help, since they are really powerful.
Stephie <3
Posted: Sat Sep 22, 2012 11:21 pm
by jarek
Stephie, I've realized the tube was in some strange operating point as I entered it in the design. Once I've corrected the OP, it has started to work ok
Posted: Mon Nov 10, 2014 1:34 am
by lukaku1
I usually try all the devices once I've added them to make sure I didn't mess up somewhere when I entered the data.