Thermionic Converter For RMS Power Measurement

[ion]

The idea came to me that a ordinary electronic vacuum tube (valve) as used in old equipment, radios, TV's etc. can be used as a power monitor by putting the power input into the filament and reading the emission directly from cathode to plate in millivolts or microamps with a DMM. No plate supply required!

This gives us a nice thermal RMS converter without having to mess with thermocouples, thermistors, infra red diodes etc. The filaments on some have very low inductance due to folded construction.

From about half a Watt to 3 Watts it is usable, and as always you should prepare a chart of your own particular tube as your "standard".

In this particular study I took a type 6AL5 dual diode, and put increments of voltage into the filament, up to 8 volts which is a bit higher than rating but that's okay, I've got boxes of them.

Next I will try some interesting tubes as used in portable battery radios of the 40's and 50's, pre transistor. These were the "1, 2, 3" series tubes. These tubes have very low filament current ratings.

Maybe I should have called this "Fun with vacuum tubes" because it is fun to find new uses for old.

Before the naysayers step in and point out all the potential problems, I'm just having fun, I know there are problems around every corner, and I can if I wished point many of them out e.g non-linear filament resistance, not ruler straight, etc but lets have some fun on this journey. 
Perhaps Graham is old enough to appreciate this? 

Maybe I should start making youtube videos of this stuff and begin watching those pennies from heaven rolling in.  A little dishonest title like " Free Energy From the Vacuum (tube) Giant Overunity Heat your house Free" ............well, not my style.

https://www.youtube.com/watch?v=EXIK9YkwIqA

[Vasik041]

Very nice idea

Valves are undeservedly forgotten. They have many interesting and unique properties.

[ion]

Very nice idea

Valves are undeservedly forgotten. They have many interesting and unique properties.

Thanks Vasik, not as linear as the TC method but I've got thousands of all types in my attic to play with, so it's fun.

Long obsolete but still prized in certain applications, I just highly appreciate the technology of valves.

One interesting aspect of valves is their ability to accelerate electrons, but there are many other properties worth studying.

Regards

[Vasik041]

One interesting aspect of valves is their ability to accelerate electrons, but there are many other properties worth studying.
Regards

It is offtopic here, but may be it will be interesting for you (you can delete if it is not).
Some time ago I was collecting everything about negative resistance and among others things I found this (please see attached document).
It was mentioned somewhere in relation to Tesla's black box which powered his electric car.

Regards

[CITFTA]

And don't forget they are EMP resistant!  So the old tube stuff will still work after all the solid state stuff is fried.

[TinselKoala]

I like it!   

[ion]

Looks like my choice of the 6AL5 was not the best, it was what I had lying on the bench. In the attic I found a JRC9004 UHF Acorn diode which outperformed the 6AL5 by a mile. Only 150 mA at 6.3 volts and well over a volt output. This breaks the mW / mV ratio now greater than 1 mV per mW. Data posted soon. Have to find those portable radio tubes for comparison and I also have some new subminiature tubes with flying leads.

Besides from a DC power supply, I ran the tube directly from my SG, at fixed voltage and the output was nearly constant all the way to 5 MHz Sine, the limit of the HP generator. This tube has less than a microhenry filament inductance. Data to come.

It occurred to me that I could put this tube under Arduino control and data acquisition using a Max6675 or other 16 bit converter for A/D input. Then it might be possible to actively bias the tube into the linear portion and apply some linearization to the curve. Resolving power input would then be about 0 to 3000 milliWatts with 16 bit resolution over that range, or uW resolution.

[gyula]

Hi ION,

Interesting and good idea to use the filament of a vacuum tube for dissipating an unknown power and learn about it.   

You surely recall car radios in which the valves had 12.6 V heater voltage, this feature may extend the useful measuring range either down or upwards.  Here is a link to such valve types, you may have some of them in your junk box:
http://www.junkbox.com/electronics/lowvoltagetubes.shtml   
Your further idea of putting the tube under Arduino (or perhaps other) control is also good. 

Thanks for sharing your ideas.

Gyula

[ion]

Dear Gyula
Thanks for suggesting the low plate voltage tubes. I'm sure I have some in my stash.

On another note, I think a tube that has been properly characterized over it's preferred range of operation can be used at much higher input power providing the power attenuation ratio of the external switched "soak" resistors is factored into the final reading with the appropriate range switching indicators or scale selection.

Under Arduino control, I see a calibration cycle easily built  into the test set so that aging effects and other factors can be taken into consideration.

Using bias, I have been able to take out the dead zone from 0 to 2.5 volts therefore position the start of the curve closer to the start of emission or position the zero anywhere I wish on the curve. This allows for the registering of very low power devices.

Regards

[muDped]

Gyula,

Many thanks for the link you've provided to
the Junkbox!


The article about the "space charge tubes"
contains links to other resources which are
extremely interesting.  I have a special affection
for one-transistor-regenerative-receivers but
had not yet discovered the Macrohenrydyne.

You've done it again!

[TinselKoala]

Looks like my choice of the 6AL5 was not the best, it was what I had lying on the bench. In the attic I found a JRC9004 UHF Acorn diode which outperformed the 6AL5 by a mile. Only 150 mA at 6.3 volts and well over a volt output. This breaks the mW / mV ratio now greater than 1 mV per mW. Data posted soon. Have to find those portable radio tubes for comparison and I also have some new subminiature tubes with flying leads.

Besides from a DC power supply, I ran the tube directly from my SG, at fixed voltage and the output was nearly constant all the way to 5 MHz Sine, the limit of the HP generator. This tube has less than a microhenry filament inductance. Data to come.

It occurred to me that I could put this tube under Arduino control and data acquisition using a Max6675 or other 16 bit converter for A/D input. Then it might be possible to actively bias the tube into the linear portion and apply some linearization to the curve. Resolving power input would then be about 0 to 3000 milliWatts with 16 bit resolution over that range, or uW resolution.

I have a couple of 2D21 tubes as spares for my StroboTac. What do you think? (It might be difficult for me to find a socket though...)

[ion]

I have a couple of 2D21 tubes as spares for my StroboTac. What do you think? (It might be difficult for me to find a socket though...)

The 2D21 is a gas filled thyratron (SCR like) tube so I have no idea how those might work. If you like to experiment this is fun stuff but I think there are much better ways to go from heat to mV.

e.g. the Fluke 540B transfer standard uses a vacuum tube with a built in thermocouple. Also there are tubes made for vacuum gauges that also have a filament with TC attached but they are open on one end to detect the number of molecules present. You could pull a vacuum and seal off the end for high output or use as is, but I suspect the output would vary with humidity if left open, so not so good.

http://www.ebay.com/sch/i.html?_odkw=vacuum+gauge+tube&_sop=15&_osacat=0&_from=R40&_trksid=p2045573.m570.l1313.TR1.TRC0.A0.H0.Xvacuum+gauge+thermocouple.TRS0&_nkw=vacuum+gauge+thermocouple&_sacat=0

I'm still working out the bugs on the thermionic emission method, and it looks promising if done with bias on the filament. I tested that and it works fine and allows zero suppression to any point on the curve.

There is a world of tests needed to be done with this method, and I really don't have the time to spend.
e.g. a small bias on the grid will hurl electrons at the plate with greater speed and number....haven't tried that yet.

The thermocouple-on-resistor method works so well and is so linear, I think that is the way to go if on a budget.

Regards