VFD tubes/displays and multiplexing

[09d08]

Hi, I am currently thinking of building a VFD clock with 4 IV-4 tubes but I'm having a hard time deciding whether I should multiplex the segments or drive them statically. My main concern is the how the drive method affects the lifetime of the tubes. I couldn't find any comparison detailing this aspect online. The translated datasheet I have states that the maximum grid/segment voltages are 2 times higher (50V vs 25V) and that the maximum grid current is also much higher(40mA instead of 6mA) when multiplexing. Static drive has a specified brightness of 500Kd/m^2 whereas it's 300Kd/m^2 multiplexing but honestly I have no idea whether any of these are related to the expected lifetime of the tubes. You probably need to use higher voltages to get the same brightness as multiplexing reduces the average voltage so that explains it.

So long story short, does anyone know whether multiplexing VFD segments or driving them statically is better in terms of lifetime assuming the brightness is the same in both cases?

Thanks.

[PKTKS]

The really critical factor to VFD life cycle is the cathode.

Cathode of  all tubes wear out... VFDs are not different.

It is highly recommended to run the heather at the lowest possible
voltage cf specifications.. and almost all VFDs use AC drive heathers.

The multiplex or not multiplex question affects more the logic
used than VFD life .. with the introduction of VLSI controllers
and multi character grids.. multiplex is the only solution to
minimize pin count.. almost all large VFDs are multiplexed
and due to using modern ICs.. they should be as well

So the VFD project is far more critical in voltage drives
than logic.. due to modern MCUs you can even drive them
with simple serials.. I2C SPI ...

Paul

[ledtester]

Here's a data point:

"VFD CLOCK YULIA IV-12"
https://kama-labs.com/vfd_iv-12-6

In the features section it states:

* 6x IV-12 Russian VFD tubes (made in 1990-93) in sockets
...
* 20 years lifetime of tubes
...
* Fade at night (increase lifetime of tubes in twice!)
...
* IV-12 tubes works in static mode

[09d08]

Cathode of  all tubes wear out... VFDs are not different.

It is highly recommended to run the heather at the lowest possible
voltage cf specifications.. and almost all VFDs use AC drive heathers.

I see, I'll just drive them statically then. 4 HV5812s should be enough. The datasheet specifies a heater voltage between 2.45V and 2.75V, I'll use 2.45 to be safe. Do you know if it's bad to use a DC  for the heater filament? Some pages mention uneven illumination but I didn't notice that while testing one of the tubes on a breadboard. Not having to generate a sine wave would make things a lot easier.

[PKTKS]

Do you know if it's bad to use a DC  for the heater filament? Some pages mention uneven illumination but I didn't notice that while testing one of the tubes on a breadboard. Not having to generate a sine wave would make things a lot easier.

YES very bad... possible for some short periods..

But for continuous operation it will damage the filaments..
they may even glow red..

Recommend minimum AC voltage..
usually a small transformer with middle ground is frequently used (AC/GND/AC)
this is the best method...

I am aware that some odd exotic models may be different ..
usually 99% follows the rule  AC filament drive.

Paul

[09d08]

YES very bad... possible for some short periods..

But for continuous operation it will damage the filaments..
they may even glow red..

Oh well, then those voltages for the filament in the datasheet must have been RMS voltages. A miniature center-tapped transformer and an inverter should do the trick then.

[PKTKS]

YES very bad... possible for some short periods..

But for continuous operation it will damage the filaments..
they may even glow red..

Oh well, then those voltages for the filament in the datasheet must have been RMS voltages. A miniature center-tapped transformer and an inverter should do the trick then.

Google for an ancient NORITAKE appnote about driving filaments..

They have exactly these wave forms discussed...
How important they are and how to use them

Paul

[09d08]

YES very bad... possible for some short periods..

But for continuous operation it will damage the filaments..
they may even glow red..

Oh well, then those voltages for the filament in the datasheet must have been RMS voltages. A miniature center-tapped transformer and an inverter should do the trick then.

NORITAKE appnote about driving filaments..
Paul

I was actually reading that document yesterday, they recommend AC drive to avoid brightness gradients which can supposedly happen with longer tubes. I didn't observe this while testing and the filaments were glowing so faintly that I had to cover the tube with my hands to barely see them glow. 2.45V RMS AC and 2.45V DC shouldn't be any different in terms of power dissipated in the filament, right? Does driving the filament with DC really wear the cathode out faster and reduce the lifetime of the tube? Sorry if I'm getting repetitive, it's just a little confusing because every single VFD clock that I've seen uses a low voltage DC supply to drive the filaments.

[CaptDon]

I have seen more D.C. filament supplies used in the older clocks, maybe 50/50 A.C. vs D.C., but concerning multiplexing let's look at a numerical example, not real actual values but perhaps close. Let's say with all segments lit a static driven tube drew 5 milliamps total of anode current. If you now multiplex 4 tubes, each with a 25% duty cycle you could safely raise that total peak current to 20 milliamps. In reality to avoid ghosting there should be a tiny amount of dead time from the time you blank an already lit tube until you provide the on state for the next tube with its segment drive. If you don't do this and the 'number' being displayed is a different number than what is displayed on the tube earlier and later in the multiplex line-up you may see a bit of 'superimposed' character ghosting. Since there is no way to control Ion Bombardment of the phosphor at very high segment currents the phosphor itself will slowly degrade if driven hard. It is very easy to use pulse width modulation to control brightness. The peak current would be the same but for a shorter duty cycle which would decrease brightness and increase tube life. In a 4 tube line up you could do something like 23% max duty and 6% min duty to get a 4:1 brightness range and at 23% max duty you can avoid ghosting.

[PKTKS]

Oh well, then those voltages for the filament in the datasheet must have been RMS voltages. A miniature I was actually reading that document yesterday, they recommend AC drive to avoid brightness gradients which can supposedly happen with longer tubes. I didn't observe this while testing and the filaments were glowing so faintly that I had to cover the tube with my hands to barely see them glow. 2.45V RMS AC and 2.45V DC shouldn't be any different in terms of power dissipated in the filament, right? Does driving the filament with DC really wear the cathode out faster and reduce the lifetime of the tube? Sorry if I'm getting repetitive, it's just a little confusing because every single VFD clock that I've seen uses a low voltage DC supply to drive the filaments.

Well in your specific case using 4 independent tubes that should be a easy decision...

But unlike this particular individual case.. VFDs share a common cathode (unlike regular tubes)
and  MULTI DIGIT VFDs  can not be driven with UNEVEN POTENTIAL

Quoting NORITAKE
Luminance varies with the filament voltage (Ef) as shown in Fig.10. Since the lifetime of a VFD is dictated by the extent of evaporation of oxide materials coated onto the tungsten filament wires, it is critical that the filament voltage is supplied within the specified ratings.

Current drain from the anodes and grids to the filaments can cause ghost illumination so a bias voltage is applied to the filaments to raise them above ground.


Old school VFDs like this one I shot below.. if driven with DC POTENTIAL
will have UNEVEN brightness and the higher potential will erode the cathode sooner...
Along with the AC EVEN drive.. you also need a bias offset to prevent artifacts...

Your INDIVIDUAL tubes should not be that hard...

Paul