Vacuum Tube SPICE Models

rongon · 2015-12-18 6:15 pm

Ah, thank you. No worries, I'm glad you're generous enough to share your models with those of us who don't have the smarts to figure this stuff out on our own. Much appreciated!

--

sleiven · 2015-12-28 8:03 am

This work fine with LT spice
* LTSpice model
* Modified Koren model (8 parameters): mean fit error 0.26mA
* Traced by erik 25/12/2013 using Curve Captor v0.9.1
* from datasheet tracer
*
* 6gu7k8 LTSpice model
.subckt 6gu7k8 P G K
Bp P K I=(0.06497587648m)*uramp(V(P,K)*ln(1.0+(-0.05252127973)+exp((2.222713943)+(2.222713943)*((21.9586484)+(-218.6907395m)*V(G,K))*V(G,K)/sqrt((50.15327587)**2+(V(P,K)-(27.02078276))**2)))/(2.222713943))**(1.270480055)
Cgk G K 3.6p ; 0.2p added
Cpk P K 4p ; 0.5p added
Cgp G P 3.5p ; 0.5p added
Rpk P K 1G ; to avoid floating nodes
d3 G K dx1
.model dx1 d(is=1n rs=2k cjo=1pf N=1.5 tt=1n)
.ends

356980d1451289751-hva-er-det-som-gja-r-ra-r-lyden-capture.jpg

sleiven · 2015-12-28 8:25 am

Removed data above 3W. better fit for low currents
* 6gu7 LTSpice model
.subckt 6gu7 P G K
Bp P K I=(0.06926663426m)*uramp(V(P,K)*ln(1.0+(-0.04309057704)+exp((2.203642054)+(2.203642054)*((23.40399493)+(-214.2707618m)*V(G,K))*V(G,K)/sqrt((43.70540698)**2+(V(P,K)-(17.16813293))**2)))/(2.203642054))**(1.257403402)
.ends 6gu7

356981d1451291059-hva-er-det-som-gja-r-ra-r-lyden-capture.jpg

6GU7 is my secret tube, drives my KT120 SET connected with very good resuts ( low overall distortion due to harmonic cancellation)

dch53 · 2015-12-28 8:26 am

Thanks very much to cogsncogs and sleiven for the 6GU7 models!

blueglow · 2016-01-10 1:00 pm

Anyone have a model for 6B4G tube?
Thanks
Mark

jazbo8 · 2016-01-10 3:44 pm

blueglow said:
Anyone have a model for 6B4G tube?
Thanks
Mark

Isn't it equivalent to the 2A3?

dgta · 2016-01-20 12:40 am

6SN7 vs 6CG7 model

I just used a 6SN7 model and 6CG7 model on the same circuit and got different results. The DC analysis comes out almost identical, but not the AC. Huge difference in THD.

So, is is this a legit difference between the 2 types? And if not, can someone point me to models for the 2 types that have been verified (reasonably) accurate? Thanks.

jazbo8 · 2016-01-20 9:51 am

They should be the same, which models did you use?

dgta · 2016-01-20 3:51 pm

I can't remember for sure, I added them to my library a while back.

That's why I'm asking for "known good" models.

Edit: Forgot that they're labelled. One is Ayumi, the other is a modified Koren by Wayne Clay.

cogsncogs · 2016-01-20 6:58 pm

dgta said:
I can't remember for sure, I added them to my library a while back.

That's why I'm asking for "known good" models.

Edit: Forgot that they're labelled. One is Ayumi, the other is a modified Koren by Wayne Clay.

Here's a quick comparison of my 6CG7 model and Ayumi's 6SN7. Black curves = 6CG7, red curves = Ayumi's 6SN7 model. But first a screenshot of my 6CG7 model in CurveCaptor. The dots (circles) are the actual traced curves from the GE datasheet.

I've noticed that different timestep values in a simulation can give differing THD levels, especially from the Ayumi models... One reason I don't give much weight to THD simulation results. Add a little salt.

Real tubes will never align perfectly with published curves and will measure differently in actual circuits. All we can hope for is reasonably accurate models.

dgta · 2016-01-21 12:53 am

Cogs, if you're Wayne, I have it the other way around, I have a CG7 by Ayumi and SN7 by Wayne Clay. Regardless, doesn't matter, thanks for posting that info, I'd like to use the Wayne version. Can you verify that this is correct? Thanks much

* 6SN7_GE LTSpice model
* Modified Koren model (8 parameters): mean fit error 0.0771164mA
* Traced by Wayne Clay on 10/11/2013 using Curve Captor v0.9.1
* from General Elctric data sheet
.subckt 6SN7_GE P G K
Bp P K I=
+ (0.01996138472m)*uramp(V(P,K)*ln(1.0+(-0.06003194624)+exp((4.721146098)+
+ (4.721146098)*((20.78010878)+(-111.9910206m)*V(G,K))*V(G,K)/sqrt((29.06231184)**2+
+ (V(P,K)-(6.821890538))**2)))/(4.721146098))**(1.378722311)
Cgk G K 2.8p ; 0.2p added
Cpk P K 2.2p ; 0.5p added
Cgp G P 4.5p ; 0.5p added
Rpk P K 1G ; to avoid floating nodes
d3 G K dx1
.model dx1 d(is=1n rs=2k cjo=1pf N=1.5 tt=1n)
.ends

cogsncogs · 2016-01-21 7:11 pm

dgta said:
Cogs, if you're Wayne, I have it the other way around, I have a CG7 by Ayumi and SN7 by Wayne Clay. Regardless, doesn't matter, thanks for posting that info, I'd like to use the Wayne version. Can you verify that this is correct? Thanks much

6SN7_GE:

Code:

* 6SN7_GE LTSpice model 
* Modified Koren model (8 parameters): mean fit error 0.0771164mA 
* Traced by Wayne Clay on 10/11/2013 using Curve Captor v0.9.1 
* from General Elctric data sheet 
.subckt 6SN7_GE  P G K 
    Bp  P K  I= 
+ (0.01996138472m)*uramp(V(P,K)*ln(1.0+(-0.06003194624)+exp((4.721146098)+ 
+ (4.721146098)*((20.78010878)+(-111.9910206m)*V(G,K))*V(G,K)/sqrt((29.06231184)**2+ 
+ (V(P,K)-(6.821890538))**2)))/(4.721146098))**(1.378722311) 
  Cgk G K 2.8p ; 0.2p added 
  Cpk P K 2.2p ; 0.5p added 
  Cgp G P 4.5p ; 0.5p added 
  Rpk P K 1G ; to avoid floating nodes 
  d3  G K dx1 
.model dx1 d(is=1n rs=2k cjo=1pf N=1.5 tt=1n) 
.ends 6SN7_GE

6CG7_GE:

Code:

* 6CG7_GE LTSpice model 
* Modified Koren model (8 parameters): mean fit error 0.0749904mA 
* Traced by Wayne Clay on 10/03/2013 using Curve Captor v0.9.1 
* from General Electric data sheet 
.subckt 6CG7_GE  P G K 
    Bp  P K  I= 
+ (0.02053902262m)*uramp(V(P,K)*ln(1.0+(-0.06147138054)+exp((4.780357089)+ 
+ (4.780357089)*((20.90066891)+(-105.6185609m)*V(G,K))*V(G,K)/sqrt((23.56820043)**2+ 
+ (V(P,K)-(4.933537229))**2)))/(4.780357089))**(1.37367599) 
  Cgk G K 3.0p ; 0.7p added 
  Cpk P K 2.7p ; 0.5p added 
  Cgp G P 4.7p ; 0.7p added 
  Rpk P K 1G   ; to avoid floating nodes 
  d3  G K dx1 
.model dx1 d(is=1n rs=2k cjo=1pf N=1.5 tt=1n) 
.ends 6CG7_GE

They are slightly different from each other as I traced them from their respective datasheet curves.

Cheers from Wayne in Wayne County. :drink:

dgta · 2016-01-21 7:45 pm

Thanks Wayne.

General question: If I understand it correctly, there are specialized tube "testers" out there capable of running curves and then generating a model from those. Is this true? If so, does anyone have a link to manufacturers, etc. or even just an article so I can read up on this?
Also, does anyone offer this as a service?

dgta · 2016-01-21 8:40 pm

cogsncogs said:
Real tubes will never align perfectly with published curves and will measure differently in actual circuits. All we can hope for is reasonably accurate models.

Agreed. I'm not worried about small differences, but some differences are very large and that kind of makes the whole simulation concept worthless. I've had some pretty large differences in AC signal levels as well as THD. I'm thinking it would be a good idea for authors to put in the notes inside the model what was used for curves: what tube manufacturer, actual tube measurement or published curves, etc. Would narrow down the variability to manageable levels, hopefully

Ufudu · 2016-01-22 5:21 am

dgta said:
Thanks Wayne.

General question: If I understand it correctly, there are specialized tube "testers" out there capable of running curves and then generating a model from those. Is this true? If so, does anyone have a link to manufacturers, etc. or even just an article so I can read up on this?
Also, does anyone offer this as a service?

uTracer? Ronald's electronic project site

I built one to do just that, so I can model the actual tubes I will be using. It's a quick and easy way of generating models. If you use the Derk Reefman's "Extractmodel" you get Koren & Derk parameters for your tube. You can also chose the measurement range to get a best fit for your tube operating conditions so you can do some optimisation that way.

I have done some parameter extraction for folks who have sent me tubes to measure, it's not a "service" but I am happy to do it FOC because that way I get to learn more about tubes

sleiven · 2016-01-30 9:35 am

It would be interesting if you could utrace a TungSol reissue 6SN7. In my amp that 6SN7 outperforms any other 6SN7 I have got , NOS or not . The distorion is halved in my 6SL7-6SN7-300B amp.
Best regards
Sleiven

318742d1430686768-psvane-we-300b-kontra-tj-full-music-300b-n-capturex.jpg

jazbo8 · 2016-01-30 11:26 am

It could be just an one-off coincidence due to the tolerance of the tube. Perhaps you can send your tube to get traced if you are really interested in its characteristics.

sleiven · 2016-01-31 1:34 pm

Both tubes ( marked 13-01) I have show the same characteristic. Sold by Karltubes who stopped selling them now.

jazbo8 · 2016-02-03 4:50 pm

5654/6AK5/6J1 SPICE Models

Triode-Connected

Code:

*
* Generic triode model: 5654_T_AN
* Copyright 2003--2008 by Ayumi Nakabayashi, All rights reserved.
* Version 3.10, Generated on Wed Feb 03 16:25:08 2016
*                 Plate
*                 | Grid
*                 | | Cathode
*                 | | |
.SUBCKT 5654_T_AN A G K
BGG   GG   0 V=V(G,K)+0.2564934
BM1   M1   0 V=(0.020250832*(URAMP(V(A,K))+1e-10))**-1.0936778
BM2   M2   0 V=(0.57832935*(URAMP(V(GG)+URAMP(V(A,K))/20.822386)+1e-10))**2.5936778
BP    P    0 V=0.0042701484*(URAMP(V(GG)+URAMP(V(A,K))/36.004374)+1e-10)**1.5
BIK   IK   0 V=U(V(GG))*V(P)+(1-U(V(GG)))*0.0030226533*V(M1)*V(M2)
BIG   IG   0 V=0.0021350742*URAMP(V(G,K))**1.5*(URAMP(V(G,K))/(URAMP(V(A,K))+URAMP(V(G,K)))*1.2+0.4)
BIAK  A    K I=URAMP(V(IK,IG)-URAMP(V(IK,IG)-(0.002314205*URAMP(V(A,K))**1.5)))+1e-10*V(A,K)
BIGK  G    K I=V(IG)
* CAPS
CGA   G    A 0.02p
CGK   G    K 4.3p
CAK   A    K 2.4p
.ENDS

Pentode

Code:

*
* Generic pentode model: 5654_AN
* Copyright 2003--2008 by Ayumi Nakabayashi, All rights reserved.
* Version 3.10, Generated on Wed Feb 03 16:24:50 2016
*               Plate
*               | Screen Grid
*               | |  Control Grid
*               | |  |  Cathode
*               | |  |  |
.SUBCKT 5654_AN A G2 G1 K
BGG   GG   0 V=V(G1,K)+0.2564934
BM1   M1   0 V=(0.020250832*(URAMP(V(G2,K))+1e-10))**-1.0936778
BM2   M2   0 V=(0.57832935*(URAMP(V(GG)+URAMP(V(G2,K))/20.822386)))**2.5936778
BP    P    0 V=0.0042701484*(URAMP(V(GG)+URAMP(V(G2,K))/36.004374))**1.5
BIK   IK   0 V=U(V(GG))*V(P)+(1-U(V(GG)))*0.0030226533*V(M1)*V(M2)
BIG   IG   0 V=0.0021350742*URAMP(V(G1,K))**1.5*(URAMP(V(G1,K))/(URAMP(V(A,K))+URAMP(V(G1,K)))*1.2+0.4)
BIK2  IK2  0 V=V(IK,IG)*(1-0.4*(EXP(-URAMP(V(A,K))/URAMP(V(G2,K))*15)-EXP(-15)))
BIG2T IG2T 0 V=V(IK2)*(0.76634978*(1-URAMP(V(A,K))/(URAMP(V(A,K))+10))**1.5+0.23365022)
BIK3  IK3  0 V=V(IK2)*(URAMP(V(A,K))+3705)/(URAMP(V(G2,K))+3705)
BIK4  IK4  0 V=V(IK3)-URAMP(V(IK3)-(0.002314205*(URAMP(V(A,K))+URAMP(URAMP(V(G2,K))-URAMP(V(A,K))))**1.5))
BIP   IP   0 V=URAMP(V(IK4,IG2T)-URAMP(V(IK4,IG2T)-(0.002314205*URAMP(V(A,K))**1.5)))
BIAK  A    K I=V(IP)+1e-10*V(A,K)
BIG2  G2   K I=URAMP(V(IK4,IP))
BIGK  G1   K I=V(IG)
* CAPS
CGA   G1  A  0.02p
CGK   G1  K  2.6p
C12   G1  G2 1.7p
CAK   A   K  2.4p
.ENDS

sleiven · 2016-02-03 6:23 pm

jazbo8 said:
It could be just an one-off coincidence due to the tolerance of the tube. Perhaps you can send your tube to get traced if you are really interested in its characteristics.

Yes , who is willing to utrace it for me?
Best regards
Balle

Vacuum Tube SPICE Models

Member

Member

Member

Member

Member

Ex-Moderator

Member

Ex-Moderator

Member

Member

Attachments

Member

Member

Member

Member

Member

Member

Ex-Moderator

Member

Ex-Moderator

Member