• WARNING: Tube/Valve amplifiers use potentially LETHAL HIGH VOLTAGES.
    Building, troubleshooting and testing of these amplifiers should only be
    performed by someone who is thoroughly familiar with
    the safety precautions around high voltages.

Vacuum Tube SPICE Models

Now that this has come up, just out of curiosity...

What is the difference between 'triode-strapping' a pentode model or using a triode model of a particular pentode?

If I put a 6L6 pentode (Ayumi) model with its g2 tied to anode into a particular circuit, I get different results if I insert a 6L6-Triode (Ayumi) model instead. Is that because of differences in the programming of the models themselves, or is there some actual difference between triode-strapping a pentode model or using an actual triode model?

--
 
That makes complete sense!

That's a good way to evaluate a pentode model, then. If you know a particular pentode-in-triode model is very good, substitute the straight pentode model and strap g2 to anode. Run the simulation and see how different the results are. If they're really far off, the pentode model might not be very good.
 
Ex-Moderator
Joined 2011
Hi rongon
It just depends on fit quality. Pentode of high fit quality will deliver correct results when used in triode mode.
That’s exactly right, also bear in mind that Ayumi’s pentode models are either built with the triode curves from the datasheets or from the curve tracers, so there are bound to be some fitting errors. In addition, for certain pentodes, their triode-connected curves in the datasheets were in-correct to begin with, i.e., if you add up Ip and Ig2 from the pentode curves, you do not get the Ip shown on the triode curve. So I wouldn’t really worry about the small discrepancies in these models...
 
Is there a model for type 49?

Casting around for an LTspice model for type 49 tetrode. Is there one out there somewhere? Or maybe I can impose on one of you expert curve tracers to whip up a model? If so, many, many thanks.

Data sheet:
https://frank.pocnet.net/sheets/021/4/49.pdf

I've attached tetrode and triode curves as well.
 

Attachments

  • 49-tetrode_curves.png
    49-tetrode_curves.png
    227.2 KB · Views: 272
  • 49-triode_curves.png
    49-triode_curves.png
    307.6 KB · Views: 296
Disabled Account
Joined 2013
The 1st curve above is not a true tetrode plot as it has g1 and g2 connected together as one input and tube is biased or driven up to +30v. The correct way I think is to generate a true tetrode wi model from the triode curve (2nd curve) with bias start from 0 volt and then connect g1 and g2 and plot the curve to see it matched 1st curve. And also the screen current in true terode model can be re-adjusted so it matched with screen current of the 1st curve. Ayumi tool could be easier then paint tool. The other curve shown is type 47 not type 49 not sure what the screen g2 voltage reference is in use.
 
Thanks Koonw.

Both curves say '49' on them. Which one looks like 47 to you?

I made a mistaken assumption. Type 49 is not a tetrode at all, it's a low power 'Dual-Grid Power Amplifier' designed for either Class B use with g1 and g2 tied together (held at 0V, I think) or Class A triode use with g2 tied to the plate and g1 used as the control grid.

I think the 49 looks most interesting as a standard class A triode (second curve, g2 tied to plate). Low mu of only 4.7 could be used as a line stage, and the triode curves look quite linear. Gm is pretty low, but maybe with a source follower on the output? Internal rp of about 4.2k ohms isn't so bad.
 
Disabled Account
Joined 2013
https://frank.pocnet.net/sheets/021/4/49.pdf

In the last curve of the above it's for type 47, which is a pentode of 3 grid, the other 2 is for type 49 not sure why they appeared on same sheet.

I still think one needs to generate first a model similar to a tetrode even though it's not a true tetrode but it behaves like one, as seem in triode mode g2 is same as screen (g2) of conventional tetrode, connected to plate HT , while g1 is used as input, not the other way round.
 
Last edited:
I hadn't noticed that the 49 data sheet includes a set of 47 pentode curves. That's odd. I'm not sure how that would work out, since even if you connect the 49 as a tetrode, it should have a kink in its curves due to the lack of a suppressor grid (g3).

Is it not possible to generate a triode model based on the curves from the 49 data sheet? I doubt anyone would be interested in Class B operation anyway, but I'll bet there are people out there who'd be interested in Class A triode use.
 
I made a mistaken assumption. Type 49 is not a tetrode at all, [...]

Hi rongon
Of course the 49 is a tetrode, as a tetrode is defined by containing 2 grids. The 49 is just intended by it's designers to be driven in not so common modes. ;)

Like the 12K5, another tetrode intended for unusual modes, where g1 is on a fixed (positive) voltage and the modulation is done with g2 by applying negative voltages. Tetrodes for this special mode are known as space charge tubes.

all the best, Adrian
 
Disabled Account
Joined 2013
....
Is it not possible to generate a triode model based on the curves from the 49 data sheet? I doubt anyone would be interested in Class B operation anyway, but I'll bet there are people out there who'd be interested in Class A triode use.
Yes, you can separately generate a triode model using the 49 triode curve with CurveCaptor, but the model does not measure or detect dissipation of g2, for instance since it's presented as triode only, user has to make aware of this..

Paint tool can deal with or insert kink, knee and other odd things..these are lacking in other tools.
 
Last edited:
Hi rongon

I have a freshly baked 49_i2 tetrode model for you! :D
It is by the way the premiere for my generic tetrode model, and I'm quite happy how god it fits. :rolleyes:

As my model consist of 94 lines of code, it seems to me better not to insert it on this forum... So, please download it here:
http://adrianimmler.simplesite.com/440956786

all the best, Adrian
 

Attachments

  • fit_49_classB-Mode.png
    fit_49_classB-Mode.png
    781.2 KB · Views: 420
  • fit_49_TriodeConnected.png
    fit_49_TriodeConnected.png
    769.9 KB · Views: 414
It strikes me we have an excellent resource for storing all of the datasheets and their curves in the form of frank.pocnet.net but haven't yet organised the ever growing number of spice models derived from the datasheets or self-traced curves .

Would it make sense for someone to work with the admin on frank.pocnet.net and simply add all of the spice models onto that database in some way? I appreciate there'd be swaps and interpretations but at least they'd all be in one place.

kind regards
Marek