Simulation Software for hybrid (tubes and solid state) circuits?

[steve_qix]

Hi !

I have to design a circuit for a friend, but I don't want to actually build it.

In this case, it is a vacuum tube rf amplifier with a solid state source follower screen modulator.

What is good (and free!) that I can use to get this done ?

Thanks !!!!

Steve

[W1RKW]

micro cap 12.  it's free.
https://www.spectrum-soft.com/download/download.shtm

[Opcom]

I'd agree with that.
Depending on the tube you want to use, you may have to make the model yourself.
It comes with 12AX7, which I vave been playing with for 100mW input plate modulated part 15 AM broadcast transmitter. It's not  a simple program but is way more capable than I have been able to take advantage of!

[W1RKW]

There's also National Instruments Multisim Student Version for $40.  It is based off of Electronic Workbench.  It comes with some vacuum tube models, a fraction of Microcaps vacuum tube library, however but tubes can be modeled. It is a bit easier to use than Microcap. 

https://store.digilentinc.com/ni-multisim-student-edition-circuit-design-and-simulation-software

I have version 11 and it works well for what I do.

[W4AMV]

Hello Steve,

Take a look at LTspice. It is a free down load from Analog. It is easy to learn to use and has a built in help menu structure and a significant library of excellent examples.

There are tube models available from third parties and available for download. LTspice has built in symbols for triode, tetrode, pentode. There is an excellent users group via Google SIG.

Take a look at the following for some tube model and information:
http://www.normankoren.com/Audio/Tubemodspice_article.html

And here is an excellent resource:

http://bordodynov.ltwiki.org/

73' Alan

[DMOD]

Marshall Leach has a good explanation on constructing tube models in Spice: Spice Models for Vacuum Tube Amplifiers, Journal of the Audio Society, Vol. 43, No. 3, March 1995;

http://leachlegacy.ece.gatech.edu/papers/tubeamp/tubeamp.pdf and

https://digilander.libero.it/paeng/spice_models_for_vacuum_tubes.htm and

https://www.agsrhichome.bnl.gov/AP/ap_notes/ap_note_97.pdf

The Koren reference is a good one.

[M0VRF]

LTSpice of course, I know it's been mentioned already.

Massive user base, industry standard.

Surprised this question has been asked as it's been around a while, perhaps not as long as valves tho'!

JB.

[steve_qix]

Thanks !  This is good information.  I'm going to give LT Spice a try.  The project is at the point where I have to design the modulator.  The RF amplifier is all working with a pair of 6L6GC tubes.  Now it needs the screen modulator.

Interestingly, I can't actually build it because I don't have any tube equipment (at all) at the new location.  It's still in storage back in Townsend (which is not nearby to where I live now, else I would go and get a transmitter for testing).  So, up to this point, I've come up with the design, and someone else (who is far away) is building and testing.  

It's certainly an interesting process, particularly when things don't work at the other end, troubleshooting is remote, the design is completely untested, and my friend at the other end is not at all familiar with circuit design.  Schematics have been dictated by telephone, which results in some interesting anomalies  

I'll email the modulator schematic to my friend, as it is more complicated than the RF components (which included the oscillator, buffer and RF amplifier).

[steve_qix]

Ok, I have LT Spice.  I need a model for the FQA11N90C, which I have, however, it appears to be impossible to use this in LT Spice.  I cannot figure out how to do this  , which either means I am overly obtuse, or the way to do it is so obfuscated that I haven't been able to find the magic key 

I've searched around on the Internet for ways to import external PSpice models into LT Spice, and, although people say it can be done, I cannot seem to be able to do it.  I have found no clear instructions on exactly how this is done.  The help within LTSpice is of not help whatsoever.

The parameter names in the PSpice model are completely different from, and do not even begin to match the parameter names contained in the .sub models (for MOSFETs) that are part of LT Spice.  Otherwise, I would modify one of those.

Any ideas ? I'm sure this has been done, but it is alluding me very nicely !

Thanks!  Steve

[w9jsw]

Talk to WA1GFZ.

[W1RKW]

I think you can model it as a  .model type model vice .sub.  A .sub will be more complicated and tend to be geared toward more complicated devices such as op-amps and other IC type devices.
 
 the models for MOSFETs appear to reside in the CMP folder of the library.  The file of interest is the M standard file.  I tried modifying that file by using the FQA11N90 part number and using the parameters of  FQB11P06 part number that lives there but the FQA11N90 would not show up in the component menu. 

CMP path is: C:\Program Files\LTC\LTspiceXVII\lib\cmp

As John said, Frank/GFZ may  know.

[W4AMV]

Please see this:

https://electronics.stackexchange.com/questions/397354/how-can-i-convert-pspice-model-into-ltspice-model/397365

There is a video too.

Look at the Google users group as well for LTspice.

This is a .lib file and is straightforward to place into LT.

[M0VRF]

Yes you can import any manufacturers spice model into the simulation. There used to be an amazing Yahoo user group and guess it's moved (Google/groups.io?) but not sure where.

I modelled class E eight or so years ago and the simulations were quite accurate.

JB.

[steve_qix]

I want to thank everyone who posted in this topic.  It is VERY helpful.  While I still am not able to import the Fairchild .lib file into LTSpice, I did use an existing MOSFET model, and changed the parameters to match the FQA, and added a line to one of the LTSpice files.

This is the circuit I have come up with for the screen modulator.  It simulates perfectly.  Whether it will work in real life - well,  we'll see :-)  If anyone sees any anomalies or things that may not work (or should be changed) please let me know.

The circuit takes about .8V p-p to achieve full output, which should be more than enough to modulate most (if not all) screen grids that we would be using.  The circuit features a DC coupled negative peak limiter, screen voltage adjustment and overload and arc-back protection.

[KB1VWC]

Is thé voltage source labeled v6 the correct polarity? Maybe + - swapped?

Steve

[steve_qix]

Is thé voltage source labeled v6 the correct polarity? Maybe + - swapped?

Steve

I agree, that is CONFUSING !  The negative voltage is actually coming from the + terminal of the supply, which IS confusing.  I originally had the supply turned around, with the + terminal at ground, and then it said +12V, which was ALSO confusing.  Ideally, there should be a negative power supply symbol (and maybe there is).

Thanks !!  Steve

[steve_qix]

I fixed it.  Created a new asy file for a negative power supply.  The schematic is updated.  You may have to refresh the page to see the changes.

Thanks for the feedback.

[WD5JKO]

Steve,

    Consider softening that negative clip circuit. Here is a concept I have used:

http://amfone.net/Amforum/index.php?topic=30945.0

    The RF tube screen voltage required to get 100% negative screen modulation varies from tube type, and is likely something negative. Still with the 3 stage soft clipper, you could still start while positive, like +10v (soft), 0v (moderate), and -10v (hard). The pot could be used to vary all three thresholds.

Jim
Wd5JKO

[steve_qix]

Hi Jim,

Wow, that's an interesting approach.  I can't say I've ever heard anything on the air using this.  I've always used "hard" limiting (often with a low pass audio filter following the neg peak limiter).

Any idea of whether the IM distortion is higher, lower, the same with soft limiting ?

Oh, I made more changes to the schematic.  Had to add another voltage amplifier to get a bit more gain. The p-p voltage input voltage required ended up exceeding 1 V  p-p.  Now, it is considerably less.

[WD5JKO]

Hi Jim,
Wow, that's an interesting approach.  I can't say I've ever heard anything on the air using this.  I've always used "hard" limiting (often with a low pass audio filter following the neg peak limiter).
Any idea of whether the IM distortion is higher, lower, the same with soft limiting ?

   Steve, I have worked a lot with soft negative cycle "attenuation" versus a hard clip, and to me, the hard clip is always more harsh (subjective). AM IMD tests with sine waves are no good in that either method gives you bad numbers, one way more abrupt, and the other more gradual. Thinking about it, doing a voice plot on a spectrum analyzer (FFT) could be useful. A pre-recorded repeating syllable, "hay", or "Fred" over and over while varying the level, and clip threshold(s) could be fun. Close in high resolution needs a slow sweep and small resolution bandwidth. Since the syllable "hay" or "Fred" is brief, compromises must be made to see what is happening. The Flex 5000 might be better than a spectrum analyzer.

   I attach a plot of a single level soft clip on a plate modulated rig. Audio reports are excellent, but IMD and THD at that level with a sine wave not so good, but excellent pre-clip. Looking at my signal on a web SDR and it is clean. Also added pre-clip THD and IMD for same Tx.

Jim
Wd5JKO

[steve_qix]

Update:  My friend built the circuit (which I never built or tested), and it worked perfectly the first time, and as expected.