An AC filament supply is usually preferable, especially in long tubes
meant to be multiplexed. If you use a DC filament supply in tubes like
these, you'll find that one side is brighter than the other. The same
thing happens with strings of smaller tubes that are run in series (one
side of the tube string is brighter than the other side). However, if
you are using single tubes and supply their filaments in parallel, you
won't see much of a brightness gradient from DC.
Theoretically, running the filaments in DC will shorten the lifespan of
the tube.
So, as with any design, take it apart and decide what works best for
you. If you're familiar with linear AC power supplies, it might make the
most sense to run the filaments in series off of a transformer. This
might make particular sense if you are planning to run the VFD's
"direct", because you can use a small transformer and a voltage doubler
for your HT as well.
You can also use various power amplifier circuits to generate an AC
filament supply from DC.
The technique that I used on my multiplexed VFD clock (which uses
individual tubes) was one that I learned on this list. Simply build a
beefy 5vdc supply (you'll want it to be a bucking switching supply).
Power each tube's filament in parallel from the 5v supply, in series
with a voltage dropping resistor. Since the filament will draw a fixed
amount of current, one can easily calculate the value of the resistor
needed to lower the 5v supply to the supply voltage appropriate for the
filament. This has the disadvantage of wasting a lot of power in the
resistor, however. Make sure that the power dissipation rating for the
resistors used is sufficient!
The advantage to this technique is that it is simple. The filaments are
biased ~3vdc in relation to the board's "ground", so simply grounding
the grid on a VFD will work very well to turn it off. The grid will then
be -3v in relation to the tube.
For the HT supply, I used the same MC34063 supply that I use for nixie
clocks. Instead of a voltage divider on the feedback line that is
calculated for 180vdc, I used one calculated for ~50vdc.
You'll need to figure out a technique for sourcing your HT voltage to
drive the VFD's. I liked the MAX6921 because it was free (sampled) and
available in a PLCC package. I do all of my building on protoboard, so
the cheap availability of DIP PLCC sockets was a primary decision maker
in my choice of driver IC.
-Adam
On 12/19/2011 4:03 AM, MrNixie (UK) wrote:
Dylan, First off, treat the power circuits for the "HT" and heaters
separately. otherwise any circuit that you might come up with is going
to be a compromise, and loading on one voltage rail is going to effect
the other. One problem at a time is best!
Each IV 11 tube has a 1.5V(100ma) heater. So there is your first
challenge - how to drive those? Put all the filaments in parallel and
then design a 1.5V 600mA power supply? Don't think so! Put them all
in series, so now you have to feed your tubes with a 9V 100mA supply?
That starts to sound more likely. But now each tube is "sitting on" a
different heater voltage, so tubes at one end of your array are going
to see a different voltage step of from their heaters to the HT. This
then causes those tubes at the "high" end of your heater chain to glow
less brightly. This is noticeable on multidigit arrays like the IV18
(8 digit mux tube) for example. You might get away with it where
you've got shorter strings of heaters in series - how about two
strings of three heaters each, driven via a diode dropper (0.6V) from
a handy 5V supply? That might work well.
This brightness gradient across an array is also exasperated when the
tubes are directly driven, since the HT voltage is lower, when
compared to the total voltage across your heater chain.
Another way around any brightness gradient is to drive the heaters via
AC. if you don't already have it from your power supply, you have to
make it! And that's what the MAX 628 is all about - you use it a bit
like a four diode bridge rectifier in reverse - DC in, AC out (with a
square wave control signal, of course). That was, over the AC cycle,
each tube "sees" the same average HT voltage.
There are lots of small SMPS designs out there, either using 555's or
something a little more esoteric. Personally, I favour the UC3483
switcher. I've used it for years, and it's bullet-proof. But pick one
that you've worked with before, and try that. And Yes, 1000uH is going
to be WAY too big - come down to something nearer 100uH, that can
handle about 1A peak. your choice of FET looks fine.
Have a look at other designs out there, and experiment!
WARNING! Want to know the QUICKEST way to fry a VFD? Put 5V DC across
the 1.5V heater... marvel at the LIGHT BULB you have just discovered
(briefly), before the filament POPS. Conversely, it is hard to blow up
a VFD with too much HT voltage (within reason!) at least for short
periods of time. DOUBLE CHECK your heater circuits to make sure they
cannot over-volt, or I will be back to say "I told you so..."
Laurence
www.MrNixie.com
On Dec 19, 9:43 am, Grahame Marsh<[email protected]>
wrote:
I am interested in building a IV-11 clock from scratch, and have Maxim 1771s,
inductors around 1000 uH, and IRFD220s on hand but am open to ordering whatever
parts I need. I'm most interested in solutions for generating the grid
voltage, but ideally am looking for a nice circuit for both grid and filament.
But it is true that you and drive the filament with DC but you get a
brightness gradient across tubes in series or long multisegment VFDs?
I've never seen how severe the gradient is. The anode/grid voltage can
be generated with a simple power supply by selecting the transformer
voltage correctly.
Grahame
--
You received this message because you are subscribed to the Google Groups
"neonixie-l" group.
To post to this group, send an email to [email protected].
To unsubscribe from this group, send email to
[email protected].
For more options, visit this group at
http://groups.google.com/group/neonixie-l?hl=en-GB.
