> After spending several days playing with the negative voltages and testing 
> it, it doesn't work as I was hoping. It seems that a lot of people use AC 
> instead of DC and I can't find an example that uses DC for negative voltages. 
> This was the closest I could find, 
> http://threeneurons.wordpress.com/vfd-stuff/ , but it looks like its AC.

Yeah, that's AC.

> I was working with 2 tubes in series with a 30 ohm 2W resistor that has +5V 
> applied to the first filament and then the other end is grounded, and the 
> grid/anode was fed +30V with a HV5812 shift register.

I'm not quite sure what that lashup is, my best guess is one end of one 
filament is connected to +5V, the other end to a resistor, the other end of the 
resistor to the other filament, and the other end of the other filament to 
ground.  That's a kind of silly lashup.  I'd connect +5V - filament - filament 
- resistor - ground.  This would put the filaments at a slight positive bias, 
so grounding other elements would give solid cutoff.

You don't really need a resistor when running two IV-17 filaments in series, as 
they're rated for 2.4 volts, so two of them in series would need 4.8 volts - 
close enough to 5 volts.  If you wanted to use a series resistor to give them 
more precisely the right voltage, you'd only need 4 ohms.

> I put the -27V on the first end and GND on the other end of the filament, it 
> lit up pretty bright on the filament and I tried with different resistor 
> values but even the 2W resistors were heating up and the brightness of the 
> segments wasn't there.

You're dropping a lot of voltage there:



Note that you want the resistor connected to ground, and the filament connected 
to -27V, not the other way around.  The IV-17 wants 47mA of current*, which is 
the same as 0.047 amps.  You want to drop the difference between your negative 
supply (-27 volts) and the filament voltage (2.4 volts), so you'll want the 
resistor to drop 24.6 volts.  Dividing 24.6 volts by 0.047 amps yields 523 
ohms.  510 ohms is a common value, and would be close enough.  That resistor 
would be dissipating 0.046 amps times 24.6 volts, or nearly 1.2 watts - enough 
to make a 2 watt resistor fairly warm!  Since your resistor is dissipating more 
than ten times the power of the filament, and you need to make the filament 
hot, you're going to burn a lot of power this way (this is one of the reasons 
some of us like to use AC filament drive - then you can use capacitors to limit 
the filament power instead of resistors, and capacitors don't just waste power 
as heat).

Ideally, you'd have a supply that's 2.4 volts different from your -27V supply.  
So if you had a -27V supply and a -24.6V supply, that would work with no 
resistor.  Similarly, a -27V supply and a 29.4V supply.  Normally, you'd split 
the difference and have a -25.8V supply and a -28.2V supply, so the average 
filament voltage would be -27V, but it's not critical.

> I then tried -27V on the first filament and different combinations of -27V 
> and the +5V with resistors on the other end of the filament and the anodes 
> were grounded, but nothing was working or it was very dim segments.

Going from -27V to +5V only makes the problem worse - now you're trying to get 
2.4 volts out of a 32 volt difference.  -27V to ground is better.  You want to 
use the voltage that's closest to -27V available (as long as it's >= 2.4 volts 
different).  Since you only have the single negative supply, you're pretty much 
stuck with either the big watt-burning resistor or going to an AC filament 
drive.

> While I would love to get this working, It looks like I should use a second 
> MCU to handle the display/boost and another to handle all the other 
> operations.

I don't think another MCU is going to help with this problem.  You could use 
the MCU clock output to derive your AC filament drive, or some I/O pin that you 
have generating a pulsing signal.

* the current is key, that's what creates the heat that makes the filament 
thermionically emit electrons.

- John

<<inline: filament-dc.png>>

-- 
You received this message because you are subscribed to the Google Groups 
"neonixie-l" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to [email protected].
To post to this group, send an email to [email protected].
To view this discussion on the web, visit 
https://groups.google.com/d/msgid/neonixie-l/731DB77D-C08F-41BC-A09B-21D0E77FF88E%40mac.com.
For more options, visit https://groups.google.com/groups/opt_out.


Reply via email to