• 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.

Tube spice models

Jump to Latest
#21
Cauhtemoc said:


It is written in Tcl, so the actual program is the .tcl file. You need an interpreter to run these files. One example is ActiveTcl, which is freely availible from ActiveState.
Click to expand...

Yes that is correct and I use the activestate version of tcl without problems.

Curve Captor does not care what the application of the tube is, it just captures curves. Note that you must manually enter the inter-electrode capacitances into the model in order to have a fully valid ac model.

The overall accuracy of the model depends on well you are actually able to match the curves which is controlled by amongst other things the complexity of the polynomial equations of the models used and your patience.

Lots of patience is required and the program is a little difficult to use, if you misplace a point on a curve you can ignore it or start over depending on how many errors you've made. (Sometimes points wildly off of the target curve will be ignored in the final curve fitting part of the process - if they're not you must redo your curve fitting entry.)

My overall experience with curve fitting pentodes has been frustrating, I think the underlying math used in the models often does not allow a very good fit in the cutoff region and it is very difficult to get a good match in this region.
 
#22
janneman said:

What AX year/issue was that? I have all of them. Do you have a copy of the Ham radio article?

Jan Didden
Click to expand...

It was August 2004 -- I've since gone on to build one for USB with a Measurement Computing controller -- the HV amplifier was a real kluge but it worked like a charm.

The HR article was by Joe Carr sometime in 1978 and I have it on CD-ROM. The only ham magazines I keep now are from when I was in high school.
 
#23
Concerning Curve Captor, kevinkr said:

Lots of patience is required and the program is a little difficult to use, if you misplace a point on a curve you can ignore it or start over depending on how many errors you've made. (Sometimes points wildly off of the target curve will be ignored in the final curve fitting part of the process - if they're not you must redo your curve fitting entry.)
Click to expand...

You can tweak the most recently placed point with the arrow keys on your keyboard.

My overall experience with curve fitting pentodes has been frustrating, I think the underlying math used in the models often does not allow a very good fit in the cutoff region and it is very difficult to get a good match in this region.
Click to expand...

I didn't think that Curve Captor would create pentode models.
 
#25
#26
janneman said:



Hi Jack,

Thanks. I actually have all the material including the tubes and data sheets. I was just trying to find spice models so I could sim the amp before building; at least to weed out my most stupid mistakes 😉

What AX year/issue was that? I have all of them. Do you have a copy of the Ham radio article?

Jan Didden
Click to expand...


Got it. Good article, clever thing. I'd appreciate a copy of that Ham article if possible, jan atta linearaudio dotta nl.

Jan Didden
 
#28
Hi Jack,
You're tweaking us right?😀

For those new to the program you just select and run the curve captor application, IIRC if windoz is configured correctly it will automatically start TCL for you. TCL is just a programming language which can run widgets and lots of other things..


Curve Captor here: http://sourceforge.net/projects/curvecaptor/

Install TCL first and then Curve Captor..

It's been quite a while since I last used CC and I will admit there is a very steep learning curve.. There's quite a bit more on it here:

Pretty much everything you need to know should be on this thread:

http://www.diyaudio.com/forums/showthread.php?postid=632704
 
#29
Hi Jan,
I suggest You take a look at Acoustat dd amp. It will go 10kVpp but no more than 10mA. If You consider very low amplitude at 20kHz in music
material this should be sufficient. I have made some changes in original
schematics (with a help of a friend) and now is very stable. You can find my posts regarding this.
I think 1500Vrms is not nearly enough and You will end up building something else with more swing or going back to step-up transformers. Original 6HB5 pentode has been
replaced with 6HV5 triode .
Regards.
 
#30
sasha said:
Hi Jan,
I suggest You take a look at Acoustat dd amp. It will go 10kVpp but no more than 10mA. If You consider very low amplitude at 20kHz in music
material this should be sufficient. I have made some changes in original
schematics (with a help of a friend) and now is very stable. You can find my posts regarding this.
I think 1500Vrms is not nearly enough and You will end up building something else with more swing or going back to step-up transformers. Original 6HB5 pentode has been
replaced with 6HV5 triode .
Regards.
Click to expand...

Hello Sasha,

Thanks, I'll look up your posts. Is there a circuit diagram there also?
I was going to built this for my Martin Logan Sequell II's, which as far as I know would have enough with 1500VRMS. But I'll take another look at it.
Thanks for the heads-up.

Jan Didden
 
#34
janneman said:
Hi,

I am looking for tube spice modesl for a couple of radial beam power tetrodes, no luck so far. Anybody can help me:

4X150A / 7034 (Eimac probably)
8072 (RCA probably)

TIA,

Jan Didden
Click to expand...

Hi Jan,

Did you ever find or create a model for the 4x150/4cx250? I also have been looking for a spice model for this tube and there seems to be nothing out there. I think I may need to create one.

Michael
 
#40
jackinnj said:
My son, the quant, found Eurequa on the web -- a neat program for taking data and fitting equations -- here's a youtube tutorial:
YouTube - Introduction to Eureqa (1/2)

and the organization's website:
Eureqa | Cornell Computational Synthesis Laboratory
Click to expand...

Hello,
Quant? Is that some kind of savant? (joke!)
The Eureqa artificial intelligence thing is interesting.

Tell us more. Have you put tube data into Eureka and put the extracted formula into SPICE and modeled tube performance?

Several years ago I studied tube modeling in SPICE. I ran across a fellow tube geek with a Los Alamos National Lab e-mail address. The guy wrote a Visual Basic program that would import a bitmap of triode plate curves. With the bitmap imported you would tweak the typical tube SPICE parameters until the model curves would match the imported bitmap. Then you would put the parameters into the spice model.
I modeled several tubes using this method (12B4A and 5670 are two). Then I built breadboard circuits and compared test values with spice results. It was fun to see theory match practice. I even plotted some triode strapped 5763 data in Excel, printed the curves and scanned the curves into a bitmap. With the bitmap in hand I extracted the SPICE parameters from the Visual Basic program and put that into SPICE. Now I have a 5763 Single End Triode amplifier on the shelf collecting dust. It was fun!
DT
All just for fun!
 
You must log in or register to reply here.
Top Bottom