I recently acquired some VFD 7-segment display tubes (6x 10mm IV-6, 6x 18mm IV-11 and 10x 18mm IV-22 russian VFD tubes). VFD displays need a filament voltage (about 1V@50-100mA) and a positive 25-50VDC grid & anode voltage to light the displays. The filament voltage is not a problem (serial resistor from +5VDC logic voltage), but i'd liked to find a simple schematic for a 30-50V switched PSU (about 40-50mA) from a 9-12VDC input voltage like this 100-200VDC Nixie PSU. I made this Nixie PSU and it works great, but I can't find any switched PSU NE555 applications notes... Anybody? CD
by changing R5 in your linked circuit to a lower value, around 47k, would give you ~30-40v output. a resistor devider sets the voltage cutoff when the output votage devided by the network gets to .6v at the transistors base. to keep that same tartget voltage, we must keep the same resistor ratio. 1470 = 1k pot at max + 470. at so: Code: 170 * ( 1470 / 220,000 ) = 1.135 < our goal ratio (170 / 5) * ( 1470 / ( 220,000 / 5 ) ) = 1.135 <by deff, same ratio 34 * ( 1470 / [b]44,000[/b] ) = 1.135 <algebra ;) now becasue we cant get a 44k resistor, we use 47k. at 10% tollerence, it wont make much diffrence. NOTE: every VFD display that i have seen in a production unit has used 120v mains for power with a stepdown of some sort... i guess this is easier then steping stuff up.
Yes, I figured this out this morning... Thanks for your calculations, will make the converter and do some tests with the above resistor values. Most VFD displays need a constant anode-grid voltage of 25-30VDC in direct drive mode, and up to 70-80VDC in pulsed multiplex drive mode. A stepdown converter would work too, except mains is 230VAC overhere, and dual secondary transformers with 7.5-9VAC and 24-48VAC coils are not that commonly available... Older VFD devices that I saw had even three secondary coils on the transformer (1-2VAC coil for filament voltage, one coil for logic voltage, and one coil for anode-grid voltage). That's why I thought about a stepup converter from the same 9VDC input voltage (from a wall wart for example) as for the +5VDC logic voltage. CD
Watch out with using a resistor to drop the voltage to the heater (filament). THe resistance will change as the it warms up. A better way would just to put five of them in series.
Yeah, good idea! IV-6 tubes need a filament current of 50mA@1V, IV-11 & IV-22 tubes need 100mA@1.15V, 6 digits would make 300 or 600mA in parallel... To minimize the resistance change due to resistor heating, I always use a bigger wattage resistor than needed (ex. dissipated power= 0.4W, a 0.5W resistor can handle it but will heat a lot, therefore I would use 1W resistor) CD
No no, i'm saying the resistance of the tube would change. You essentially HAVE to use come sort of regulated power supply. Also, i said series, not parallel. WIth 5 tubes with 1V heaters, you would not need a resistor, just connect directly to a 5V source. Putting heaters in series is a old tube design method, and works fine.
Okay, misunderstood you about the changing resistance... But once the filament is heated up and at working temperature its value doesn't change much (IV-22: filament resistance cold=5.2 Ohm, at working temp= about 11.5 Ohm), it's just like an incandescent light bulb. Anyway filament resistance is increasing with temp, so no worry for excessive filament current. Since I have 6 tubes to drive I have to use two times 3 filaments in series onto the regulated 5VDC logic voltage with a resistor: IV-6: (Vf=1.05V): 39 Ohm/0.25W for If=50mA filament current IV-11 & IV-22: (Vf=1.15V): 15 Ohm/0.25W for If=100mA filament current I will use the same PSU schematic based on the above NE555 Nixie PSU but redesigned for 50-70V/100mA anode-grid voltage in the final clock schematic as the digits are multiplexed (1/6 or 16% on-time). For those who do not know what VFD displays look like: IV-6 (10mm digit): IV-11 (18mm digit): CD
Thanks for the link to that PSU. I'm getting some IN-13 that have a 140vdc anode voltage and was wondering how I was going to power them....now I don't need to worry about it Don't know if you've seen this site or not, but it seems to have a lot of good stuff and links to some good information/projects. http://www.sphere.bc.ca/test/nixies2.html
Whoops, didn't remember which way it went, If the resistance increased or decreased. Anyways, looking good! The VFDs you hav are much larger than the ones i'm playing with. Mine are tiny and unmarked.
You can also use PWM to reduce the voltage to the filaments. I have seen a VFD PSU design that used a transformer instead of an inductor in a buck regulator to generate both filament voltage and anode voltage at the same time.
I've just had this argument with myself, undervolting a 12V bulb to last longer. Resistor has the advantage AFAICS over say diodes in that with a cold (low resistance) filament the resistor drops a larger fraction of the supply voltage, gradually giving a bigger share of the volts to the filament as it warms up, so the switch-on shock is less. With case fans it's the opposite, you want the switch-on shock so they start running.
Turbokeu, Don't know if you have come across this site, but there seems to be some real nice setups for nixie-type displays, and schematics: http://www.electricstuff.co.uk/nixiegallery.html I came across this 9-12vdc nixie power supply on the above site...I might try this one!! http://www.electricstuff.co.uk/nixclock.html
i think the word you are looking for is varistor? holy cow, that is the ghettoest VFD ive ever seen. boy have they come a long way, they have graphical VFD's now. next time i see one of these in a vcr or somehting im gonna save it, i didnt relise how cool they were
They aren't ghetto, they are Old-Skool. You won't see one of these in a VCR. They haven't been made for many years.
You're welcome. That nixie PSU is very simple and works very well. Several components in the schematic are not critical and can be replaced by other. The BUZ73 NMOSFET can be replaced by an IRF730, IRF740, I used an IRF840 (use a MOSFET with the lowest possible Rdson) The Fast Recovery diode UF4004 can be replaced by a BYV95C or BA159. The inductor L1 is the most critical component: use a 100µH, near 1A or more DC current inductor (not a choke!) with preferably a low DC resistance (Epcos, Toko, Panasonic). With the IRF840 and a Panasonic ELC09D101F inductor (100µH/0.82A/0.2 Ohm) or Toko 622LY-101K (100µH/0.91A/0.28 Ohm) I get easily 15mA output current from 100 to 250VDC output voltage. Remember the 140V anode voltage of your nixie is the typical anode voltage, you need to apply about 170-180VDC with a suitable anode resistor (see the datasheet for the recommended cathode current). Yes, I know Sphere, lots of nixies, but too expensive... CD
Indeed... I will use (software) PWM to the grid voltage to vary the display luminosity (in direct mode the displays work well with grid voltages between 10 and 30VDC for min-max luminosity. CD
Maybe ghetto for you, I like them very much... And yes, I have other VFD's. My latest from eBay, a huge Futaba 2x40 chars, 30 Euro (about 36us$ - 20GBP): CD
Yeah, I've another issue concerning the filament power supply of my 7-segment VFD displays. Most of the filaments of these displays need 1 to 1.5V at a current of 20 to 150mA (longer display tubes with more digits do need even up to 5V glow voltage). Following the info that I found on the internet it is advised to feed the filaments with AC voltage to minimize the luminosity gradient across the different digits in the tube. With one-digit tubes (like my IV-6, IV-11 and IV-22 tubes) this isn't much a concern as the filament(s) is (are) very short, but with multidigit VFD tubes like my IW-18 (8+1 digits), ILW1-5/7 (4 digits + semicolon), ILC1-6/7 (6 digits + 2 semicolons) this gradient is really noticable with DC-filament feeding. Before filaments were mostly feeded in parallel by a separate 1 to 1.5V winding from a transformer, but actually such transformers are quite difficult to source. I first thought to feed the filaments (when using multiple tubes) in series with a higher AC-voltage (ex. 6VAC with a little series resistor) but this situation isn't ideal because of the potential gradient in the filament as opposed to the 20-30V of the anodes. Anybody with experience in VFD's (or any other genius idea)? CD
if you only need a volt or two, just wind a few turns of wire around a pre-existing transformer. it shouldn't take more than 10 turns.
