triode thermodynamic model

[anli]

Does such model exist? I mean a changing of triode parameters as function(s) of currently dissipating power. In SS world such dependence is known as "memory distortions".

[kevinkr]

Interesting question, I don't believe there is any exactly similar mechanism in vacuum tubes, however there are small parameter shifts that occur during warm up as internal electrode dimensions change.

This is well outside of my area of expertise. I would refer to Terman, RDH4 or one of the other old textbooks to determine whether there are any similar mechanisms present in standard or purpose designed tubes. (I'm not aware of anything similar to memory effect in transistors in the tube world. Thermal tails, etc.)

Note that like transistors and jfets tubes do exhibit some shift in dynamic transconductance as a function of the instantaneous current flowing through them.

[anli]

Originally posted by kevinkr Note that like transistors and jfets tubes do exhibit some shift in dynamic transconductance as a function of the instantaneous current flowing through them.

OK, anything depends on anything Here I'm interesting in thermo-modulation, and the main difference I can see (with respect to SS) are much longer time constants which describe power dissipation influence to a transconductance (you see, such thermal inertia model may be described with help of LPF with certain time constants, as at SS case). I think, at such situation "SS memory distortions" become "additional pseudo-reverberations", and last ones are one of the reasons tube audio-chains sound more "live" and "musical".

[EC8010]

I don't see there being a signal-dependent thermal effect in valves. Consider the size of the active bit of silicon inside a transistor, now consider the size of a valve electrode structure with similar ratings/usage. The thermal mass of the valve is much higher and any signal-induced heating is likely to be swamped by the nearby heater.

[anli]

Originally posted by EC8010
I don't see there being a signal-dependent thermal effect in valves. Consider the size of the active bit of silicon inside a transistor, now consider the size of a valve electrode structure with similar ratings/usage. The thermal mass of the valve is much higher and any signal-induced heating is likely to be swamped by the nearby heater.

The thermal modulation definitely exists (I know such measurements were accomplished). To estimate time constants we need qood square wave oscillator (with frequency, say, 0.1-0.2Hz) and ADC working from 0 Hz (i.e. with DC-open input) - probably good digital scope with recording will be suitable here. I have not such equipment.

I don't say about any influence at *sound* frequencies. I say about infra-low frequencies and about modulation of *sound enveloping*. And such modulation is, I think, similar to reverberation, if we say about our ears and something between them

[JohnAtwood]

There is probably more of an effect from anomalous behavior of the cathode rather than pure thermal effects. The oxide-cathode is a complex system that was only barely understood during the heyday of tubes. Look for references on the "cathode interface" problem for more on this. A good place to start is RCA's in-house publication "Electron Tubes". It can be found at: http://frank.pocnet.net/other/RCA/RC...be_Design.pdf. (Warning: 88Mbytes!)

- John Atwood

[anli]

Originally posted by JohnAtwood There is probably more of an effect from anomalous behavior of the cathode rather than pure thermal effects. The oxide-cathode is a complex system that was only barely understood during the heyday of tubes. Look for references on the "cathode interface" problem for more on this. A good place to start is RCA's in-house publication "Electron Tubes". It can be found at: http://frank.pocnet.net/other/RCA/RC...be_Design.pdf.

Thanks, I'll look for. At any case it is interesting to look at square-waves testing. Slow sound changings (say, with characteristic time interval about one second) are very sensible for our ears, and thermal modulation in "heavy" parts may be inside this frequencies range - say, 0.1-10Hz).

Originally posted by JohnAtwood Warning: 88Mbytes!

Ups.. It's rather costly for me

[d2134]

Here is a link. It may help you.
http://audioportal.spb.ru/plugins/p2...2_articleid=18

"Importantly also that that the basic parameters of lamp do not depend on the temperature of the anode or, otherwise, from the power being dissipated on it."

(This quote was traslated by babelfish)

[anli]

Originally posted by d2134
Here is a link. It may help you.
http://audioportal.spb.ru/plugins/p2...2_articleid=18

"Importantly also that that the basic parameters of lamp do not depend on the temperature of the anode or, otherwise, from the power being dissipated on it."

(This quote was traslated by babelfish)

At the same site there is a forum thread with opposite information. But nobody has published recorded square signal. To be more strictly, I don't know such publications. And this thread aim is to get such concrete info or to inspire tektronix owners to accomplish such measurements. Is there anybody brave?