Page 3 of 4

#31

Posted: Sat Jan 30, 2010 10:35 pm
by Mike H
Panel Meter Component

Any purpose.

Basically a low value inductor with a small resistance. Edit SpiceLine to change values:

Ind, inductance; Ipk, peak current; Rser, series resistance.

.

#32

Posted: Sun Jan 31, 2010 12:14 am
by Mike H
Mistake In Triodes ZIP Again!


Capacitance parameter CPG spelt wrong (is not CGP). Mostly affects bottom half of list.


.

#33

Posted: Wed Feb 10, 2010 6:25 pm
by Mike H
There is something weird going on ......


Just discovered in the tetrodes models, screen current is most often ZERO if this calculation line is used (the 'new' version):

GIG2 S K Value={(EXP(EX*(LOG((V(S,K)/MU)+V(G,K)))))/KG2}


Is all OK if use the 'old' version:

GIG2 S K Value={pwr((V(G,K)+V(S,K)/mu),1.5)/kG2}


Don't understand what's going on but so far I've only established EF86 works with both OK, with the same result of current. It seems to depend on whether param KG1 is < 1,000 or not, if it is, screen current is zero. Image

Duhhh... Image


(NB GIG2 is a current source representing the screen current)

At the mo resorted to the old version which at least produces something sensible; anyway on his web page Koren says that that part basically hasn't changed :?:

.

#34

Posted: Wed Feb 10, 2010 9:31 pm
by Mike H
OK I've found out what's happening ~ if there's no signal, screen current is zero! Only in some cases, e.g. KT66

Soon as there's an AC signal, however small, the screen current appears as if by magic.

Bizarre ~

Had to revert to new version of the calc cos the old version was causing "time step too small" bomb outs where screens are connected to anodes

#35

Posted: Thu Feb 11, 2010 2:59 pm
by Mike H
Mistake in New Tetrodes.zip, wrong name for grid anode capacitance parameter (but didn't raise errors ~ or did it?). List is updated.

-------------------------------------------------------------------------------

I now know what's going on re previous ~ it's so bleedin' obvious yet at the same time, not ~ the tetrodes can only be used as tetrodes, i.e. with a fixed screen Voltage. Wiring the screen to anode does not make it work like a triode.

To make it work as a triode you must use the proper triode model! If for the only reason that the curves are very different.

.

#36

Posted: Fri Feb 12, 2010 12:31 pm
by Mike H
All Valves Zip ~ The Big Digest

All in one big lump so it's easier to maintain.

new_triodes.txt
Using Norman Koren's improved vacuum tube models formulas as from here:

http://www.normankoren.com/Audio/Tubemo ... ticle.html

new_tetrodes.txt
Ditto Norman Koren's 'new formulas'.

pentodes.lib
True pentode symbol where has external suppressor (g3) terminal. A rough approximation will alter anode and screen currents in opposite directions by a small ratio where g3 Volts is different to cathode Volts.

Basically anode current increases and screen current decreases with an increase of g3 Volts relative to cathode, and vice versa.

If g3 connected to cathode then merely reverts to basic tetrode model operation.

New Implementation
Suffix '_at' = 'as triode'

For both new_tetrodes.txt and pentodes.lib some device types may be suffixed '_at', e.g. 'EF86_at', this uses a triode model on the assumption that screen grid g2 is connected to anode. In this case a real triode model must be used, doesn't work just by connecting g2 to A! Further g2 and g3 connections are ignored, but included so that schematic diagram is complete showing all possible connections for that device type.

Not too many triode versions yet, I could make more but it literally takes hours....

vrectifier.txt
Full-wave rectifiers (specifically, with a common cathode).

vdiodes.txt
Other single or double valve rectifiers / diodes that are NOT specifically full-wave configuration (i.e. separate cathodes).

ZIP updated: 13th Feb 2010

ZIP updated: 17th May 2010 (6EM7 added)

Updated: 1st Aug 2010 (simplified diode for grid)

ZIP updated again: 02 Sept 2010


.

#37

Posted: Fri Feb 12, 2010 2:00 pm
by Mike H
CCS Devices

List as currently stocked by RS.

So-called 'Constant Current Diodes' (CCD), aka 'Current Regulating Diodes' (CRD), actually a type of JugFET device in the typical form:

Image

.

#38

Posted: Fri Feb 12, 2010 9:09 pm
by Mike H
New Improved Fuse Component

This one actually blows like a real fuse provided heating time is reached in sim if current exceeded, for these ones set at 0.5 sec. (changeable)

From a working demo by Helmut Sennewald. I turned it into a proper lib with a list of values ranging 50mA to 10A. (Rapid 20x5mm glass F fuse web page listing.)


Image

.

#39

Posted: Sat Feb 13, 2010 10:49 pm
by Mike H
Nothing is ever straightforward is it? Naturally ~ spent most of today going nuts trying to find out what's causing fatal errors.

.

#40

Posted: Mon May 17, 2010 3:15 pm
by Mike H
LM385 Adjustable Voltage Reference


RS Stock No. 534-2854

Edit ~ model updated to make it simpler (just a V element)

NB the 1.24V is always across R3 in the example diagram for when you come the calc your own resistor divider.

 

#41

Posted: Mon May 17, 2010 7:44 pm
by Mike H
All valves.zip above updated; mistake in ECL82, plus 6EM7 added.

 

#42

Posted: Mon May 17, 2010 7:55 pm
by Mike H
IXCP10M45S

 

#43

Posted: Fri Jun 18, 2010 10:29 pm
by Mike H
Drain current vs gate Voltage graph for the above:

#44

Posted: Fri Jun 18, 2010 10:32 pm
by Mike H
Updates:

Power Darlingtons (TIP147/142), filament lamp, bridge rectifier and LM317 download ZIPs have been updated.

 

#45

Posted: Sun Jun 20, 2010 11:55 pm
by Mike H
Rechargeable Lead-Acid Battery

Converted from PSpice model by S. C. Hageman 09/01/1993

But I added the rechargeable bit.


ZIP includes test circuit and a schematic ASC version of the model (but won't run, only used to make a library file)