Glitch with 6550 tetrode model

stephie's place

Moderator: sbench

Post Reply
Paul Fawcett
Posts: 6
Joined: Thu Sep 13, 2012 5:54 pm

Glitch with 6550 tetrode model

Post by Paul Fawcett »

Hi Stephie,
Happy New Year, I hope you've been well! Thanks also for helping me out last year by working your magic on the 12BH7 triode model.
There's an active thread (or more accurately, one that has sprung back to life) right now over at AX84 where some folks are comparing and evaluating some models for the 6550. In the course of that discussion it came to light that there may be a serious glitch with the 6550 model in your library..not sure if its one that you tweaked personally or it was someone else's, but it in any case it seems to have an issue. If you're so inclined perhaps you could take a peek?

Cheers,
Paul
Im evidently not allowed to post urls here yet - the thread # is:
thread=301369
Paul Fawcett
Posts: 6
Joined: Thu Sep 13, 2012 5:54 pm

Filler, please ignore

Post by Paul Fawcett »

Ignore, just padding post count to get to 5 for url posting
Paul Fawcett
Posts: 6
Joined: Thu Sep 13, 2012 5:54 pm

AX84 URL

Post by Paul Fawcett »

sbench
Posts: 296
Joined: Fri Jan 28, 2005 3:45 pm

Post by sbench »

I am not particularly happy with ANY of the models in the tetrode file. Koren's basic models are pretty good for specific applications, but you'll notice they don't change screen current as a function of plate voltage, which is a flaw. I had tweaked some of those models to get around that limitation, but none of the current ones particularly well model things during grid conduction, or cover secondary emission (kink) effects etc. I had started late last year developing a model that does handle those, and am slowly converging on a methodology that handles all those effects so that you CAN try unusual things with the tetrode and achieve meaningful results. The "quick fix" for the 6550 in the "tetrode.txt" is to alter two values in that file...

KG1 was 890. Change it to 1018.
In the g2 equation, change the value 20 to the value 22.5.

These two changes hit the "published" operating points
vp=250, vg2=250, vg1=-14 producing ip=140 and ig2=12
and
vp=400, vg2=225, vg1=-16.5 producing ip=87 and ig2=4

These tweaks do not fix the kink region, but they do hit the published operating points.

Hope that helps.

Stephie <3
Post Reply