VFD Tube-like effects?

[Frankenbike]

I've been researching possibly doing some projects using Korg's Nutube VFD triode tubealikes.

Then I found out people have been building amps using displays from clocks and old video recorders, because those are also triodes. Each segment of a 7 segment display, could be its own tube. And a clock display has at least 25 of those.

So, what could you do with 1" x 4" board full of 25 tubes? BTW, no one is thinking like that, that I've found so far.

Here's the sort of thing I've found that inspires this line of thought.
https://www.instructables.com/id/VFD-Amplifier-a-Tube-Amp-From-VCR-Screens/

[teemuk]

Unfortunately, each segment is likely not an individual tube but VFD with 8 segments could include e.g. 8 plates, or 8 grids, which are switched on and off at rate that looks continuous to human eyes.

[Rob Strand]

The main issue is the heater forms a common cathode for all segments.   In order to stop hum injection you need to run the heater with DC (or very high frequency AC).   With a DC heater each segment operates slightly differently because each sees a different DC bias relative to ground.  I think you can see the segments not being equally bright on the site you linked.  The heaters are thin wires that span across all the segments.  From what I remember, the two stage amp on that site uses two VFD's ie a separate VFD for each tube.

[teemuk]

This is quite useful referation of operation of generic VFDs for those who attempt to reverse engineer.

https://www.noritake-elec.com/technology/general-technical-information/vfd-operation

Ofcourse Noritake Nutube is just simplified to bare dual triode because its aimed for plain amplification.

[Frankenbike]

Let's examine the pros here. If at worst a single VFD has a single grid and cathode, IVL2-7/5s can be bought for $1.70 each in a 10 pack from eBay, or around $25 from Amazon. Other VFDs are similarly priced. Appearance becomes a not terribly expensive consideration, as dot matrix VFDs could be fun to watch.

If you're looking at pre-amp tubes, $1.70 is pretty cheap. They also are relatively flat, run relatively cool, use much less power than a 12AX7. You can fit a lot of IVL2-7/5s in a medium sized stompbox.

Some VFDs have discrete grids or cathodes, most don't according to my research. I wish it was easier to find which had discreet cathode/grid/plate per numeral. The ones that do could provide a lot of bang for the buck. Running different EQ, on different pre-amps, that might have weird cross talk characteristics that aren't exactly hi-fidelity, might be a unique and newly sought after sound

At worst, I can get IVL2-7/5 VFDs in a 10 pack for $17-$25. That's a hell of a lot cheaper than a $50 6P1 NuTube. So if the ones with discrete grids/cathodes are too expensive, I can use a bunch of those. They're 2.25 x 1.1 inches, and each can be run on two AA alkalines. There are other types that are not dissimilarly priced.

If I'm the only one to think of this, the turnaround to a design and execution might take awhile. If the fidelity is crap, we'll see what kind of distortion pedal it makes

[Frankenbike]

The main issue is the heater forms a common cathode for all segments.   In order to stop hum injection you need to run the heater with DC (or very high frequency AC).   With a DC heater each segment operates slightly differently because each sees a different DC bias relative to ground.  I think you can see the segments not being equally bright on the site you linked.  The heaters are thin wires that span across all the segments.  From what I remember, the two stage amp on that site uses two VFD's ie a separate VFD for each tube.

Well now you're just increasing possibilities. If each sees a slightly different bias, how would that sound if both channels had the same signal? Can that be exploited some way. Also, appreciate the solutions to the problems you point out. High frequency circuits for heaters are not out of the question. Use another VFD for the oscillator. At $2.50 each for an IVL2-7/5 from eBay, you can use them for ridiculous things. And make sure they're all visible, because that would be quite the light show at your feet.

Remember, we are not talking exclusively about high fidelity applications. Pedals are one area where new discoveries in distortion are valuable.

[merlinb]

The heater is a non issue. Bias is a free choice, you can always bias the grid to whatever you want, so each stage has the same bias with respect to heater, or not, whatever suits you. Since this is for guitar you'll almost certainly want to tweak the biasing anyway, to get the best sound not some arbitrary voltage figure. (Typically with these tubes the grid needs to be positive, unlike most tube circuit's you've seen!) Don't worry about biasing and filaments at this stage. First, pick a tube; then you'll know how many elements you have to work with, and we can help you from there.

Each segment of a 7 segment display, could be its own tube.

Typically no. Most VFDs have all the similar anodes connected together, meaning you really only get one independent triode per bottle. In other words, if you tried to use a 7-seg as seven triodes you would find the grid of each triode would also control the anodes of all it's neighbours... in all possible combinations. You'd get noise and oscillation, if anything at all. So it's better to think of a VFD as one triode per bottle when it comes to audio.
And let's face it, the appeal of using a VFD is mainly the form factor, not the cost. You can buy an old TV tube for the same price as an old VFD and it will make a better amplifier, even at low voltages. But it will be inconveniently tall and round rather than flat, and it won't light up.

[Rob Strand]

Well now you're just increasing possibilities. If each sees a slightly different bias, is how would that sound if both channels had the same signal? Can that be exploited some way. Also, appreciate the solutions to the problems you point out.

These things are going to sound different to real tubes anyway so you will need to experiment.

The different bias currents can be compensated for by paralleling grids or using different value anode resistors.

As far as biasing options go, you can shift the DC point of the cathode.     There were old budget configurations which used common cathodes with a bypass cap.  For example look at the first stage of this amp,
https://drtube.com/schematics/marshall/1986u.gif

Since the cathodes carry the heater current you can only use small value cathode resistors.  A diode in the cathode might work for fixed DC bias.   It could turn out the easiest solution of grounding one end of the heater and picking the anode of the best digit is fine!

Getting a good sound from a VFD is largely unexplored area so you will need experiment using general principles as a guide.  The anode voltage is going to come into play for sure.    All this is just opinion.  The site you posted is a good starting point.

[Frankenbike]

And let's face it, the appeal of using a VFD is mainly the form factor, not the cost. You can buy an old TV tube for the same price as an old VFD and it will make a better amplifier, even at low voltages. But it will be inconveniently tall and round rather than flat, and it won't light up.

Uh, power usage also. Heat dissipation. Durability. Possible long life. Possible new and desirable sound. Form factor is pretty important sometimes. There are plentiful answers as to "why" it's a worthwhile experiment by multiple people with multiple goals. At worst, the device only produces garbage sound. Even then, sometimes people want even that.

Some VFDs are cheap. $2.50 each for an IVL2-7/5. Whether they can be useful is another issue. It's 1" x 2" and pretty flat. So if more get used, so be it. They use a LOT less power than tubes. The main purpose is analog sound shaping, so I know a class A,B or D amp might be needed. It's already a frankentube device, so it's not going to pass anyone's purity test.

[merlinb]

Uh, power usage also. Heat dissipation. Durability. Possible long life.  They use a LOT less power than tubes.

OK I admit that's true for the heater, which is more hungry in a tube and accounts for nearly all the heat. Even so, a VFD is likely to use 50mA for the filament so you're already looking at wall power rather than battery power, in which case we probably don't care whether it's 50mA or 300mA. Durability and lifetime won't be much different for a tube running at the same voltages. (A pedant might even point out that the filament in a VFD is more fragile than in a tube!)

[tubegeek]

More discussion on this topic recently here:

https://www.diystompboxes.com/smfforum/index.php?topic=123699.0

For my 2 cents:

Russian subminiature dual triodes aren't crazy expensive, and make a great alternative for a "real" tube application. For a "crazy" tube application, the VFDs seem to me like interesting semi-uncharted waters to splash around in.

Got a variable power supply, signal generator and a scope? It seems to me you'll probably need them to get further than "I tried this and it sounded like that."