--- In NEONIXIE-L@..., "Terry" <terry+yg@...> wrote:
--- In NEONIXIE-L@..., "figureloop" <crobc@> wrote:
What are you planning to do with them? I am working on a 12 IV-17 PCB and somewhat universal VFD driver that will be done soon (hopefully, but I keep wasting time on the internet).
I plan on (eventually) producing clock kits and assembled clocks. Since these tubes are a lot less expensive than nixies (and are alpha-numeric) I think they'll be popular.
Do you have a www? I don't right now, but will probably get one going again within the next year.
(general stuff)
(my race car, which keeps me away from electronics)
Interesting.
(work - not much there that's visible to the public)
That's encouraging for a numitron. I have been thinking about how I might make a constant current driver, but for each segment that's a lot of parts and work! Apparently filament life is inversely proportional to volts to the fourth power, so a large change in lifetime can be afforded by just a slight moderation in voltage.
Take a look at the Allegro A6278/9 parts. Quickly - the A6278's are pre-eol and one of the A6279's is last-time buy. I believe in using dedicated drivers (like the Supertex on Nixies) - it frees me up to play with the cool stuff in software and not have to worry about a program error burning out a tube or more.
I just bought 240 Supertex HV5812 since the A6812 was in last time buy status. I doubt I'll plan on any quantity production of Numitron clocks (ugh, the last thing I need is more ideas!), so if anything I may look at a higher voltage driver for Nixies. My first clock design just used MPSA42 transistors and 'HC595s.
About the processor screwing the tubes, this is an interesting design issue that I will be grappling with in my VFD driver. I wanted to accomplish a biased AC filament drive, with no magnetics. So I have a drive scheme that will start out being guided by a microcontroller PWM output. If the PWM duty gets set too high, then it is pretty bad--it could take out all the VFD filaments! But if the PWM output simply locks up, then it is no problem since the power drive is AC coupled.
For a final design I'll probably make a dedicated PWM circuit not dependent on the uC, but then again, having the uC do it makes it possible to gently ramp the filament drive.
A tradeoff might be to just ensure that the supply voltage for the filament power circuit is such that close to full duty % is needed to power the filaments in the first place. Then a little bit of overdrive caused by a development error won't be catastrophic. I think that once a piece of software is ready for prime time, it's possible to place high confidence that it will control power electronics without faults. This is the direction things are going these days.
