After all, VFDs also have a heater, grids, anodes and are encased in glass. And they glow in tha dark.

VFDs are common on VCRs. I have a few of them I took from some broken VCRs. Last night I was working on how I could use them as vacuum triodes. I don’t have much experience with real vacuum tubes so I had to invent some, possibly wrong, arrangements, but I finally got something.

I’m not going deep into the structure of VFDs as I wouldn’t add anything to what is available on Wikipedia or on manufacturer’s websites. Just need to know that the heater (or filament) is made of thin straight metallic wires that emit electrons (thermionic effect) when heated by current. The electrons are accelerated by the electric field generated by a voltage applied between the heater and the metallic anodes, the metallic plates shaped as digits and pictograms and that are covered with fluorescent paint which glow when hit by the electrons.

A number of thin grids are placed between the cathode and groups of anodes with the purpose of screening or letting go the electric field generated by the voltage at the anodes.
A negative voltage between one of the grids and the cathode will generate an electric field opposite in sign with the anode-cathode one, reducing or voiding at all the latter. The electrons will be stopped and will not reach the anodes behind the grid and those digits will be dark. A positive voltage at the grids is actually necessary to pre-accelerate the electrons.

That said, I reached this final layout.

The grids are connected together as well as the anodes. I connected a headphone between the anodes and the positive of the anode voltage through a decoupling capacitor.
The grids are polarized by a 100k Ohm resistor to the positive and the audio fequency is fed into the grid through another decoupling capacitor.

The heater generally requires 2-3 Vdc. When powered from AC they give a more uniform brightness but in our case DC is better, to limit hum into the headphones.
The anode voltage may vary between 20 to 40 Vdc, depending on the model. The connections to the filament are the only ones to be really careful about as misplacing them and feeding with the anode voltage wil blow the heater. Looking closely through the glass of the display the filament and their connections to the connection pins can be easily found.

The power into the headphones is limited but I never expected more than this.
A novelty, nothing more, possibly.

Might behave better as low power preamplifying stage for a real vacuum tube power final stage.

That’s it, and I had fun.

Addendum (after comment from Hiro Protagonist) : The arrangements of the anodes in the drawing is simplified as the display I used has multiplexed digits, that is the segments and pictograms are paralleled inside the display to minimize connections to the driving IC is usual applications. This limits the use of the VFD as a multiple triode.

This is the radio controlled tank I’m designing around the Tamiya track & wheel set (p.n. 70100) and the relevant Tamiya twin motor gearbox (p.n. 70097).
I’m not detailing the model tank design, though I’m available if you ask, rather I’d like to show you the details of the motor controller I designed to convert the digital radio receiver PPM output pulses to a PWM drive suitable for the two Mabuchi FA-130RA the twin gearbox is equipped with.

This is a video of the tank in action.

And a kill cam video…

The video gives some hints on how it works together with output waveforms.

The microcontroller (Atmel ATtiny24) monitors the two servo outputs of the radio receiver and measure the pulse lenghts.
The pulse length is then converted to a 0% – 100% PWM pulse output. The dynamics expansion is achieved through a one-time calibration : the two radio control sticks controlling the left and right tracks are pushed and kept full forward then ‘max’ button is pressed : the maximum input pulse lenght is measured and a coefficient is inferred and stored to EEPROM.

Then the sticks are pulled and kept full backwards and the ‘min’ button is pressed and the minimum pulse lenght is measured and a second coefficient is evaluated and stored in non-volatile EEPROM memory.

These two EEPROM-stored coefficients are nonvolatile and they are retrived at power up and used to expand the dynamics. The trimming operation can be repeated if necessary.

If the motors behave weirdly after trimming, probably the two buttons have been switched in the operation, e.g. max button pressed while setting minimum.

In the picture above the two small buttons are top left.

The two calibration buttons can be replace by two small berg contact of the kind found on PC motherboards for configuration.

The microcontroller uses no crystal quartz as the one time calibration compensates for the internal oscillator tolerance.

The brdige driver I took from an HP printer. Allegro Microsystems have many equivalents in different and smaller packages : should you need directions, just ask.

The schematic and the source code is here. The HEX to be burnt into the ATtiny24.

The code was developed with Atmel’s AVR Studio 4.

This whole work is licensed under Creative Commons 3.0 license : Non-commercial, Share alike, Attribution

Alessandro Lambardi – 2008 , some rights reserved