Vacuum Tube SPICE Models

[Hearinspace]

In this linked Bartola model for 864 / VT-24 , a long name and 4 nodes are given for the triode - Plate , Grid , and what appears to be two for cathode, K1 and K2. Does anyone here know how this should be modified to fit the LTspice symbol? or is it how the symbol needs to be modified to fit the model?

Thanks

 

[Ray Waters]

The 864 is a directly-heated triode and this SPICE model includes both of the filament leads instead of treating the entire filament as the "cathode." You need a different symbol which includes both of the filament pins. Attached is a symbol that is compatible with the SPICE model; rename it to TriodeDHT.asy before using it.

 

[Hearinspace]

Thanks Ray,

As the problem with the linked model seems to be more about the tube model than LTspice per se, I'll keep this next question here.

When put into the sim as a directive it gives the error :

extra closing curly brace in "}"

Any idea how it can be corrected?

Thanks again.

 

[Ray Waters]

I did a quick check of the SPICE model with my text editor but didn't see any unmatched curly braces, so there is no obvious error in the model itself. Can you post your LTspice .asc file so that we can take a look?

 

[Hearinspace]

Ah, you're too quick for my edit !

I was running in an earlier rev of LTspice . When put into the newer version it spelled out the problem with more detail.

Missing closing brace '}' in “{\rft1\ansi\ansicpg1252\cocoartf1561\cocoasubrtf610”

A search of the model doesn't find that line so maybe it's in the code for the symbol? Just asking. No idea myself.

 

[Ray Waters]

That line doesn't appear anywhere in Ale's model, and the symbol doesn't have any code at all. As long at the symbol's pin count and pin order matches the .SUBCKT definition (which it does) the symbol can be ruled out as the problem.

Without the schematic file that you are simulating, I'm just shooting in the dark. Post your schematic file and someone can probably help sort this out.

 

[Hearinspace]

I found the error in the model file as spice saved it , adding a bunch of text to the top of the model. I'll have to sort out how to add it without that happening. It's on a mac so much of the suggestions online doesn't apply. Thanks .

Edit, Got it to work by copying the file to the schematic. Good enough for now.

One question about the DHT symbol's having both ends of the filament available. How do you work with that ? Tie both together with cathode bias resistor connected to the pair?

 

[Ray Waters]

The .asc schematic file is just a text file so the Windows version of LTspice should open it without any problems. The same is true of symbol files -- they are just text files with a special file extension. So chances are that a Windows user is going to see the same behavior that you are seeing on your Mac as long as the source and symbol files are the same.

Good luck sorting this out.

 

[Ray Waters]

One question about the DHT symbol's having both ends of the filament available. How do you work with that ? Tie both together with cathode bias resistor connected to the pair?

Ale's "composite" model includes the DC resistance of the filament itself so it was probably used to simulate the filament circuit as well as the audio circuit. I suppose that you can just use one of the two filament leads as the cathode and leave the other unconnected, but below is a link to a circuit in which Ale used his composite model for a similar tube.

https://www.bartola.co.uk/valves/2022/01/16/aa-dht-spice-model/

Note that the circuit powers the filament with a current source, like in an actual circuit. I would have to examine the model to see if the tube's behavior is actually dependent on the filament current, but I'm not really inclined to go to that trouble. There isn't much point in using a composite model like this if you don't intend to model the filament circuit. A conventional triode model for the 864 will work just as well in that case. A Google search turned up this one. DISCLAIMER: I have not tested this model.

.subckt 864 1 6 3
+ params: mu=8.2 ex=1.372 kg1=9540 kp=165 kvb=2.84 rgi=6000 vct=.195
+ ccg=3.3p cgp=5.3p ccp=2.1p
e1 7 0 value=
+{v(1,3)/kp*log(1+exp(kp*(1/mu+v(2,3)/sqrt(kvb+v(1,3)*v(1,3)))))}
re1 7 0 1g
g1 1 3 value= {(pwr(v(7),ex)+pwrs(v(7),ex))/kg1}
rcp 1 3 1g
c1 2 3 {ccg}
c2 1 2 {cgp}
c3 1 3 {ccp}
r1 2 5 {rgi}
v1 5 6 {vct}
d3 6 3 dx
.model dx d(is=1n rs=1 cjo=1pf tt=1n)
.ends

 

[Ray Waters]

Well, even though I said that I wouldn't, I did examine Ale's 864 composite SPICE model and the model does include the effects of current through the filament. Here are the two lines of code that account for that.

E11 32 0 VALUE={V(1,31)/KP*LOG(1+EXP(KP*(1/MU+V(2,31)/SQRT(KVB+V(1,31)*V(1,31)))))}
E12 42 0 VALUE={V(1,41)/KP*LOG(1+EXP(KP*(1/MU+V(2,41)/SQRT(KVB+V(1,41)*V(1,41)))))}

So this model isn't going to give accurate results unless you include a realistic filament circuit in your schematic. As I said earlier, if you don't intend to do that (or don't have a good reason to) then you are better off using a conventional 3-pin model for the 864 triode.

 

[Hearinspace]

I didn't try putting a CCS on the more complex model filaments , just hung an LED string and bypass cap off one end, but trying the model you just posted, there was very little difference in operation or output gain.

BTW, Thanks for the follow-up.

 

[mz543578854]

Hi, I am looking for a model for the E83F. It is mentioned in this thread, saying it was in the Ayumi lib. But I cannot find it there.
Can someone please post it or point to it?
Thanks a lot.

EDIT: Of course... I had searched for it a while in this thread and now after posting, I searched again, this time for all "zip" in the thread and find it immediately in Ayumi's ZIP here: https://www.diyaudio.com/community/threads/vacuum-tube-spice-models.243950/post-7112532 (I likely had an earlier version before)

😝

 

[JPS64]

Hell, I´m searching for the 6SF5GT model, could anyone help?

Thank you,
JP

 

[mz543578854]

Hi @JPS64 , did you see https://www.diyaudio.com/community/threads/vacuum-tube-spice-models.243950/post-6243316 and the post above?

 

[JPS64]

Thank you!