3
\$\begingroup\$

I managed to pick up a few cheap VFDs (Vacuum Fluoro Displays), but the information about them is scarce. What makes matters worse is that I am a post vacuum tube era child, so my knowledge in the area of using and driving vacuum tubes is limited.

The display is labelled Futaba 11-ST-26ZA and a (joke) of a datasheet can be found here: Datasheet

The specifications from the datasheet are:

Ei: 4V

If: 18mA-100mA

Ek: 4.5 VDC

Duty: 1/15

Idigit = 1.3 mA

Vgrid: 12 VDC

Vseg: 12 VDC

So my understanding is that I need to do the following to drive the VFD with a micrcontroller of some sort:

  1. Apply 4V signal between filament +'ve and -'ve terminals, limit the current to 18-100mA with a inline resistor appropriately sized, and use PWM at frequency of 15Hz and duty cycle 1/15. I was planning on switching 4V at required frequency and duty cycle through a MOSFET to the filament.

  2. Apply 12 VDC to each segment I want to turn on (again, for instance via MOSFET switch).

  3. Apply 12 VDC to the grid (digit) I want to turn on, as in 2.

Once I get the VFD working, I can implement multiplexing and driving the segments in the MCU properly.

Questions: Is my above plan reasonable? Do I need to somehow limit the current to the grid and to the segments (inline resistor - as with filament) or is this not necessary?

Thanks for feedback in advance! -Igor

\$\endgroup\$
8
  • 1
    \$\begingroup\$ Here's a pretty good tutorial. \$\endgroup\$
    – EM Fields
    Commented Apr 23, 2016 at 9:56
  • \$\begingroup\$ You need to shoot the electrons from the hot wire toward the segments. You need a potential between those two that you are not describing. So I do not think your approach will work. BTW, why did you start with just the VFD tube? Futaba, Noritaki & others make plenty of VFD tube + PCB displays with easy-to-use/commonly-available Hitachi like interfaces. \$\endgroup\$
    – st2000
    Commented Apr 23, 2016 at 13:01
  • \$\begingroup\$ Isn't applying 12v to the segments and to the grid going to allow the electrons from the filament (hot wire) to shoot through to the segments? I managed to source the VFD tubes very cheaply, had no cheap local source for the VFDs with integrated controller. \$\endgroup\$
    – IgorEE
    Commented Apr 23, 2016 at 13:11
  • \$\begingroup\$ I stumbled across the same Noritaki pdf as @EM Fields. He doesn't appear short on point. But I am, so... \$\endgroup\$
    – st2000
    Commented Apr 23, 2016 at 13:29
  • 1
    \$\begingroup\$ @IgorEE, This is why I buy/scavenge tubes + PCBs! Guessing, Ei appears to be the voltage across the filament while Ek appears to be the filament voltage w.r.t. the anode. Guessing some more, the duty cycle can effect the brightness of the display. So can the cathode to anode potential. I'm guessing Futaba is worried you will short the duty cycle and consequently turn up the voltage to get a decent brightness. And that this may shorten the display's life. Keep in mind we're talking about a Futaba spec sheet and trying to draw conclusions using a Noritake white paper. \$\endgroup\$
    – st2000
    Commented Apr 23, 2016 at 14:01

4 Answers 4

Reset to default
4
\$\begingroup\$

Firstly, thanks for all the feedback.

I went and procured some parts at the local shop today to give this a go, and I got the display to light up using only a microcontroller, a few N-FETs and a open collector HEX buffer.

I have attached a schematic to show what I did and a few pics of the breadboard forest. My goal was to learn how to get this display to light up, and now I can focus on making something useful with it.

Notes on the Schematic:

  1. The half bridges (or full bridge if you will), is driven by complementary PWM signals. Here it is important to ensure your PWMs are complementary and maybe insert some dead time to prevent shoot through. The bridge drives the 4.5 VAC through the filament.

  2. The HEX buffers pulled up to 12 V and connected to the VFD Grids and Segments drive the grids and segments. The other side of the buffers is connected straight to a MCU of your choice. Make sure the MCU multiplexes the grids and segments (I guess thats the duty 1/15 spec in the datasheet). If you let a single grid+segments ON for a long time it might damage the display. I haven't tried it yet.

  3. I'm using the 12V from a wall-wart, but a charge pump or boost converter would be more elegant perhaps.

As a proof of concept it worked, and the displays lights up. Hope someone might find this useful for future reference!

enter image description here

enter image description hereenter image description here

\$\endgroup\$
1
  • 1
    \$\begingroup\$ Don't forget to accept an answer if your problem is solved. You can accept your own if it's the best! \$\endgroup\$
    – Transistor
    Commented Sep 19, 2018 at 6:05
1
\$\begingroup\$

You need to make a electrical connection between the filament, grid and phosphorus coated segment. This allows electrons to "fly" off the filament (cathode), "hit" the phosphorus coated segment (anode) and make the segment light up. enter image description here

Your particular display uses the same lines for similar segments. It also has separate grids for each digit. So you need to multiplex your display in a more complex manner. enter image description here

These images are from the same Norikate pdf as listed in the above comments.

\$\endgroup\$
0
\$\begingroup\$

I believe the filament runs 100% of the time, is biased by Ek, and "Duty" refers to the digit duty cycle. That is, each digit should be gated on for only one-fifteenth of a frame, a frame being the time it takes to display all 11 digits sequentially.

The way it works is that the segment lines connect all of the like segments of all of the digits together, and then when the grid for a particular digit is strobed, electrons are attracted from the hot filament by the grid and and then accelerated toward the segments that have 12 volts on them, and which glow when the electrons hit the phosphor coating on the segments.

Since there are current limits specified, I'd assume that they'd need to be enforced one way or the other, and resistors seem to be a sensible way to do it, as with LEDs.

\$\endgroup\$
0
\$\begingroup\$

I am running 8 digits of one of these exact displays using arduino. I am using 2.8 volts dc to run the fillaments at 100% duty, this voltage is acheived by placing 3 signal diodes in series with the 5 volt supply, I am using an anode voltage of 20 volts on both the grids and the segments, acheived via a dc-dc boost regulator. The 20 volts powers 2 UDN611A VFD driver IC's, which convert the 5 volt logic level signal to the 20 volts needed for the display, and 2 595 shift registers which feed the UDN6118A inputs, one for the grids and one for the 7 segments and decimal. You MUST ensure that the fillament supply voltage, the high Anode voltage and the Arduino all share common ground. I personally found using 12 volts on the anodes resulted in a very dull display once you multiplex 8 digits. PS Digit duty cycle is 1/8, I am simply scanning across all 8 digits repeatedly. Works great, segments light evenly.

\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.