music, guitars & other obsessions

Vacuum Tube Amplifier Design In the Cyber Age

Nakabayashi-son's book, unfortunately only available in Japanese

Nakabayashi-son’s book, unfortunately only available in Japanese

The title of this post came from Google Translate when I was reading Ayumi’s great website – it is full of information on vacuum tube models and circuit simulations. This might be old news to many of the readers, but I only came across his site late last year, and did not spend much time with it until a few days ago (that was definitely a mistake, it could have saved me a lot of time, of course, in hindsight…) It seems that the site has not been updated for awhile, and most of the articles were from the early 2000’s – so it definitely belongs in the  “oldies but goodies” category  Ayumi also provided many vacuum tube models at his site, so it is somewhat surprising to me that very few designers use his SPICE models, perhaps it is because the site is written in Japanese.

While looking at Bill Elliott’s paper, I noticed some similarities between his and Ayumi’s work – both methodologies relied on curve fitting and parameter optimization to generate the SPICE models. While Bill pretty much stuck to Koren’s models, and re-worked the equations so they can be solved in TI Derive. Ayumi, used an equivalent circuit approach and built the tube models accordingly,  the voltage-controlled current sources are based on the Langmuir-Childs equations and the whole process was meticulously documented and  explained on his website.

Here is the triode  equivalent circuit from Ayumi:

and for the pentode:

The open-source R statistical math package was used for the parameter optimization (and many other tasks), and it’s very fast even on my aging XP machine – so instant gratification is guaranteed 😉  Here is a screen shot of the program in action, not much to it really, as Ayumi has done all the hard work already, and all I needed to do was to enter a few lines of texts,  it could not be simpler!

pctube in action

And the model works like a charm in simulation as well, here is the one I did for the 805:

805 Plate

I will  try to build a few more models, especially for the pentodes to see how they fare. But it is looking very promising indeed. I think I am near the end of my quest to find the perfect tool for generating vacuum tube SPICE models.


5 responses

  1. Darren

    Where does one get “R”, and how about an overview of how to use it?

    February 5, 2014 at 10:01 am

  2. NoSpark

    So, three years later, how did Ayumi’s model design work out for you?
    I am just embarking on Ayumi’s modeling, any suggestions?
    R is still there and current


    March 15, 2017 at 6:33 am

    • jazzbo8

      The Ayumi models worked out great, with a few exceptions, they are the most accurate vacuum tube SPICE models currently available.

      March 15, 2017 at 12:51 pm

  3. NoSpark

    R is a trip… It sort of reminds me of APL, but in English! 🙂

    I am using pctube 1.21, apparently there has been an update.

    I am modelling a 1T4 pentode and have modelled a decent triode using digitized triode plate characteristics.

    In order to model the pentode connection do I need to digitize the pentode plate characteristics?

    May 16, 2017 at 11:52 am

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s