Author Topic: Vacuum fluorescent display to clock display  (Read 53145 times)

0 Members and 1 Guest are viewing this topic.

Offline neoTopic starter

  • Super Contributor
  • ***
  • Posts: 1694
  • Country: us
  • The specialist.
Vacuum fluorescent display to clock display
« on: September 19, 2016, 03:41:53 am »
 I have a vfd that has 8, 14 segments and 11 indicators with only 25 pins. i know its multiplexed or whatever but how can i use individual segments on it without a micro controller. I already have the circuitry for the clock designed but the circuit outputs to individual segments not a multiplexed display
« Last Edit: September 19, 2016, 04:09:31 am by neo »
A hopeless addict (and slave) to TEA and a firm believer that high frequency is little more than modern hoodoo.
 

Offline jaycee

  • Regular Contributor
  • *
  • Posts: 206
  • Country: gb
Re: Vacuum fluorescent display to clock display
« Reply #1 on: September 19, 2016, 04:59:44 am »
You would need to make a circuit that briefly enables each grid, lighting the segments for the digits in that grid. I guess you could do it with a counter and shift register, and some multiplexers.... but a microcontroller would be much simpler.
 

Offline neoTopic starter

  • Super Contributor
  • ***
  • Posts: 1694
  • Country: us
  • The specialist.
Re: Vacuum fluorescent display to clock display
« Reply #2 on: September 19, 2016, 05:13:07 am »
but a microcontroller would be much simpler.

except their programmers are too expensive and even if i had one i would'nt know how to program
A hopeless addict (and slave) to TEA and a firm believer that high frequency is little more than modern hoodoo.
 

Offline Ian.M

  • Super Contributor
  • ***
  • Posts: 13090
Re: Vacuum fluorescent display to clock display
« Reply #3 on: September 19, 2016, 06:12:26 am »
Not being able to program in C is a valid reason, but cost of programmers is not.  e.g. an Arduino is just an Atmel MCU mounted on a board with a USB interface and a bootloader, so are you telling us you cant afford an Arduino Nano off Ebay?

You'll need a whole board full of logic to use it without a MCU.  Unless you can parallel some of its segment anodes to map to a 7 segment display pattern, you'll need to build a diode matrix ROM to store the segment pattern for each digit.  Assuming BCD input data, you'll also need six 4 bit latches with tristate outputs and an oscillator and decoded decade counter wired to reset at 6 to step through enabling one latch output and the corresponding digit grid to multiplex it.  Then you get to have fun with DIY high voltage drivers as most of the chips for that sort of interface are either obsolete and difficult to obtain or if current, highly integrated, with a MCU required to control them.

 

Offline neoTopic starter

  • Super Contributor
  • ***
  • Posts: 1694
  • Country: us
  • The specialist.
Re: Vacuum fluorescent display to clock display
« Reply #4 on: September 19, 2016, 07:10:44 am »
i have an arduino BUT it never did work it just says generic usb (paraphrase) in the menu and the arduino software looks a for a com1 port and can never find it because the arduino i got doesn't say "Here i am!". i kinda got made and gave up on it
A hopeless addict (and slave) to TEA and a firm believer that high frequency is little more than modern hoodoo.
 

Offline neoTopic starter

  • Super Contributor
  • ***
  • Posts: 1694
  • Country: us
  • The specialist.
Re: Vacuum fluorescent display to clock display
« Reply #5 on: September 19, 2016, 07:14:01 am »
i should note at this point to trying to learn how to program and do microcontrollers but only if i can actually do it and not stare at a screen for 3 hours trying to tell my computer that there is an arduino attached and failing. so as more eloquently put im open to it so long as i can do it without hitting a brick wall just trying to set it up.
A hopeless addict (and slave) to TEA and a firm believer that high frequency is little more than modern hoodoo.
 

Offline FreddyVictor

  • Regular Contributor
  • *
  • Posts: 164
  • Country: gb
Re: Vacuum fluorescent display to clock display
« Reply #6 on: September 19, 2016, 07:32:54 am »
You need to get over to https://www.arduino.cc/en/Guide/HomePage
and https://www.arduino.cc/en/Guide/Troubleshooting

as for COM ports, they are relatively basic communication ports - not like full-blown USB - so you need to tell the Arduino IDE which com port your Arduino is located on
https://www.arduino.cc/en/Guide/Troubleshooting#toc16 is relevant

Also, why 2 threads on same subject ?
 

Offline neoTopic starter

  • Super Contributor
  • ***
  • Posts: 1694
  • Country: us
  • The specialist.
Re: Vacuum fluorescent display to clock display
« Reply #7 on: September 19, 2016, 07:40:34 am »
Also, why 2 threads on same subject ?

quite an astute observation the answer is i have a very bad internet speed and it glitched so i made them not so the same seeing as how it wouldnt let me delete one, this one is specifically about ways to make it work as a clock display while the other is just in general. Back to the arduino i have no clue how to set a com port and none of those links tell me directly how to turn standard usb to com
A hopeless addict (and slave) to TEA and a firm believer that high frequency is little more than modern hoodoo.
 

Offline FreddyVictor

  • Regular Contributor
  • *
  • Posts: 164
  • Country: gb
Re: Vacuum fluorescent display to clock display
« Reply #8 on: September 19, 2016, 10:45:03 am »
Also, why 2 threads on same subject ?

quite an astute observation the answer is i have a very bad internet speed and it glitched so i made them not so the same seeing as how it wouldnt let me delete one, this one is specifically about ways to make it work as a clock display while the other is just in general. Back to the arduino i have no clue how to set a com port and none of those links tell me directly how to turn standard usb to com

hey, no problem !

In Arduino IDE, on Menu, go: Tools/port
and choose correct one (must have arduino plugged in!)

Good luck with the project :)
 

Offline neoTopic starter

  • Super Contributor
  • ***
  • Posts: 1694
  • Country: us
  • The specialist.
Re: Vacuum fluorescent display to clock display
« Reply #9 on: September 19, 2016, 05:37:15 pm »
Freddyvictor, you are a genius it actually worked now we can get past the arduino not working and get on the fact i dont know how to program for a multiplexed vfd
A hopeless addict (and slave) to TEA and a firm believer that high frequency is little more than modern hoodoo.
 

Offline Ian.M

  • Super Contributor
  • ***
  • Posts: 13090
Re: Vacuum fluorescent display to clock display
« Reply #10 on: September 19, 2016, 07:02:48 pm »
There's no real difference between programming for multiplexed LEDs and for multiplexed VFDs.   All the difference is in the display interface hardware, + possibly whether or not to invert the segmet and digit drive outputs, so grab 4x 7 segment LEDs, a proto-shield and 8 1K resistors and get coding.  The segments wont be very bright with 1K resistors and 25% duty cycle, but for experimenting wit the code, its desirable to simplify it as much as possible by omitting external digit driver transistors, and it isn't a good idea to go much over 20mA per pin on the Arduino, which will happen when displaying an 8 with the d.p. lit as well.
 

Offline neoTopic starter

  • Super Contributor
  • ***
  • Posts: 1694
  • Country: us
  • The specialist.
Re: Vacuum fluorescent display to clock display
« Reply #11 on: September 20, 2016, 03:27:44 am »
is the proto shield necessary? can the 7 segment displays be common anode? also in pursuit of experiment i got two regular leds to switch between one another so fast i cant see it, 5 ms, so that has to be progress especially given that i begun my work of learning programming this morning
« Last Edit: September 20, 2016, 03:29:25 am by neo »
A hopeless addict (and slave) to TEA and a firm believer that high frequency is little more than modern hoodoo.
 

Offline Ian.M

  • Super Contributor
  • ***
  • Posts: 13090
Re: Vacuum fluorescent display to clock display
« Reply #12 on: September 20, 2016, 03:37:16 am »
You can do it without the protoshield - you can even patch across to a solderless breadboard.   It doesn't matter whether the dispays are common anode or common cathode as long as all the digits are the same type and you know which you have and their pinout.   The only difference between the two is the connections, and whether or not the segment patterns and the digit drive signal are inverted.
 
The following users thanked this post: neo

Offline neoTopic starter

  • Super Contributor
  • ***
  • Posts: 1694
  • Country: us
  • The specialist.
Re: Vacuum fluorescent display to clock display
« Reply #13 on: September 20, 2016, 04:36:57 am »
i think i got it figured out but by my conclusion it is an impossibility to use this with a simple logic clock because it would take too many inputs from the clock circuitry
A hopeless addict (and slave) to TEA and a firm believer that high frequency is little more than modern hoodoo.
 

Offline neoTopic starter

  • Super Contributor
  • ***
  • Posts: 1694
  • Country: us
  • The specialist.
Re: Vacuum fluorescent display to clock display
« Reply #14 on: September 20, 2016, 06:02:54 am »
in further news i have figured out next to nothing and i cant figure which pins are which on my vfd display
A hopeless addict (and slave) to TEA and a firm believer that high frequency is little more than modern hoodoo.
 

Offline Ian.M

  • Super Contributor
  • ***
  • Posts: 13090
Re: Vacuum fluorescent display to clock display
« Reply #15 on: September 20, 2016, 06:34:22 am »
Pictures of youur VFD would be good, well lit, closeup and in focus.  Please take one from the front and another from the back.  If we cant come up with a datasheet, its fairly easy to identify VFD electrodes experimentally.
 
The following users thanked this post: neo

Offline neoTopic starter

  • Super Contributor
  • ***
  • Posts: 1694
  • Country: us
  • The specialist.
Re: Vacuum fluorescent display to clock display
« Reply #16 on: September 20, 2016, 07:14:35 am »
i just had a breakthrough and mapped out all the pins as it turns out i was making it wayyy too complex and with what i assumed it was doing i didnt have a chance of figuring it out might as well have been  :horse: haha   
to those who can't read my writing (just about everyone except me) 8 and 9 are the special pins 1 has no know connection, perhaps a ground?
    the 14 segments i code named them and no i didnt miss the i, its used to to mark the special in my head but i didnt write it, but their the same across all grids
 17=d 16=0 15=k 14=n 13=e 12=c 11=g 10= 8=i 9=i 7=f 6=b 5=m 4=j 3=l 2=a 1=nc

all for the samsung hna-08ms16 of which i have two because of some guy taking apart a stereo that broke and giving to me for the price of shipping.
« Last Edit: September 20, 2016, 07:25:39 am by neo »
A hopeless addict (and slave) to TEA and a firm believer that high frequency is little more than modern hoodoo.
 

Offline neoTopic starter

  • Super Contributor
  • ***
  • Posts: 1694
  • Country: us
  • The specialist.
Re: Vacuum fluorescent display to clock display
« Reply #17 on: September 20, 2016, 07:23:58 am »
Also as i understand it i have to cycle the grid but how do i use my clock (schematic included) output to display on this, changing as little as possible preferably. i know ill have to move the transistors to after the microcontroller, well maybe no have to but theyd do me more good there im sure. I have a vague thought of multiplexing ics but thats just a thought i would kinda like to avoid the programming nightmare of a chip big enough to do a one to one. that would be somewhere near 100 io pins. and i know im asking alot of questions but this IS my first microcontroller project, i may as well be in the desert on a horse with no name.
A hopeless addict (and slave) to TEA and a firm believer that high frequency is little more than modern hoodoo.
 

Offline Ian.M

  • Super Contributor
  • ***
  • Posts: 13090
Re: Vacuum fluorescent display to clock display
« Reply #18 on: September 20, 2016, 09:00:03 am »
14 segment 'starburst' displays,  + 8 digits = 22 pins.  There are at least 9 symbols + d.p. and colon, so they cant be done as an extra segment in each digit.  They must be connected as 'segments' in an extra 'digit' so one more pin.  + two for the filament = 25 pins. 

Segment lines are phosphor cored anodes, all in parallel for the same segment position and digit pins go to the grids for each character. 

As many hobbyist friendly MCUs don't have enough pins, assuming you aren't going to use a dedicated VFD driver chip, the easy option would be to use a chain of 74HC595 octal SIPO shift registers as an I/O expander  to provide the 23 segment and digit signals,  It has a SPI compatible interface with a latch enable pin so all outputs can be updated simultaneously.   You then need a high voltage driver, but you don't know how high until you either characterise your display or find a datasheet.  It needs to be able to sink and source as floating grids can cause problems.  There are various options here depending on the voltages involved.

As you're used 4026 combined decade counters and 7 segment decoders in your clock, there is NO easy way to interface it to this display.  Just about the only option short of a full ground-up redesign of your clock, would be a 42 input PISO shift register (built from six individual octal ones) to read all the segment patterns under MCU control, but that's frankly crazy and a far better choice would be to feed the MCU a precision clock signal and have it keep time in code, replacing all the logic in your clock.
 

Offline helius

  • Super Contributor
  • ***
  • Posts: 3675
  • Country: us
Re: Vacuum fluorescent display to clock display
« Reply #19 on: September 20, 2016, 09:05:28 am »
It's a lot easier (and more likely to work) to use VFD driver chips like MAX6932 instead of rolling your own drivers.
Controlling them with a microcontroller (or using programmable logic) is very easy since they use a pretty normal 4-wire interface.
Identifying the pins on the VFD goes like:
1. identify the filament pins. usually there are a group on each side of the display that have a pretty low resistance between them. Use continuity mode on your DMM
2. from those remaining, identify the grids and anodes. You can tell them apart because when multiple grids are powered together with one anode pin, the same segment will be lit in more than one character. And vice-versa when you power multiple anodes and one grid pin.
 

Offline neoTopic starter

  • Super Contributor
  • ***
  • Posts: 1694
  • Country: us
  • The specialist.
Re: Vacuum fluorescent display to clock display
« Reply #20 on: September 20, 2016, 09:10:08 am »
It's a lot easier (and more likely to work) to use VFD driver chips like MAX6932 instead of rolling your own drivers.
Controlling them with a microcontroller (or using programmable logic) is very easy since they use a pretty normal 4-wire interface.
Identifying the pins on the VFD goes like:
1. identify the filament pins. usually there are a group on each side of the display that have a pretty low resistance between them. Use continuity mode on your DMM
2. from those remaining, identify the grids and anodes. You can tell them apart because when multiple grids are powered together with one anode pin, the same segment will be lit in more than one character. And vice-versa when you power multiple anodes and one grid pin.

ill consider a driver chip because that sounds very sane only problem is it doesn't turn up on ebay so if you could be kind enough to point me in the direction of an obtainable chip it could work
for everything else ive already plotted the pins
« Last Edit: September 20, 2016, 09:22:55 am by neo »
A hopeless addict (and slave) to TEA and a firm believer that high frequency is little more than modern hoodoo.
 

Offline FreddyVictor

  • Regular Contributor
  • *
  • Posts: 164
  • Country: gb
Re: Vacuum fluorescent display to clock display
« Reply #21 on: September 21, 2016, 10:24:34 am »
MAX6932 isn't the easiest part to get hold of AFAICT
however, you can daisy-chain other more-readily available chips together to achieve the number of outputs
MAX6921 has 20 outputs, so putting 2 of these together would give 40 outputs

If this insufficient pins, then you could decide to either not use some of the pins or, join some up so 2 or more elements are on at the same time
eg the ':' elements may be 2 separate pins but you could connect them together

note these are MAXIM chips so not overly cheap !
 

Offline Buriedcode

  • Super Contributor
  • ***
  • Posts: 1684
  • Country: gb
Re: Vacuum fluorescent display to clock display
« Reply #22 on: September 21, 2016, 12:57:09 pm »
I see Ian M is the only one so far to mention voltage.  VFD's are not LED's, they require grid voltages 50-70V taht require HV buffers to cope.  They are a bitch to drive without a dedicated chip (and hard to find transformers) to do the job for you.  I managed to get one running from a car stereo display, it took far too long and even then I re-used the onboard power supply.

Do yourself a favour and start off with an LED display, it'll be a good excuse to start using microcontrollers.  One you have that, then you can work on the power supply for the VFD
 
The following users thanked this post: Ian.M

Offline neoTopic starter

  • Super Contributor
  • ***
  • Posts: 1694
  • Country: us
  • The specialist.
Re: Vacuum fluorescent display to clock display
« Reply #23 on: September 22, 2016, 03:13:59 am »
 i think i have the power supply figured out, nothing stops me from using multiple transformers of which i have plenty, i also made my own chip out of transistors (schematic/explanation below)
 the arduino code that makes my plan work for those into that kinda thing (tested, not on the vfd yet but ive yet to find any faults in my design or plan)
 
int IO1 = 4;
int IO2 = 5;
int IO3 = 6;
int IO4 = 7;
int IO5 = 8;
int IO6 = 9;
int IO7 = 10;
int IO8 = 11;


void setup() {
  pinMode (IO1, OUTPUT);
  pinMode (IO2, OUTPUT);
  pinMode (IO3, OUTPUT);
  pinMode (IO4, OUTPUT);
  pinMode (IO5, OUTPUT);
  pinMode (IO6, OUTPUT);
  pinMode (IO7, OUTPUT);
  pinMode (IO8, OUTPUT);
}

void loop()
 {
 digitalWrite (IO1, HIGH);
 delay (1);
 digitalWrite (IO1, LOW);
 digitalWrite (IO2,HIGH);
 delay (1);
 digitalWrite (IO2, LOW);
 digitalWrite (IO3,HIGH);
 delay (1);
 digitalWrite (IO3, LOW);
 digitalWrite (IO4, HIGH);
 delay (1);
 digitalWrite (IO4, LOW);
 digitalWrite (IO5, HIGH);
 delay (1);
 digitalWrite (IO5, LOW);
 digitalWrite (IO6, HIGH);
 delay (1);
 digitalWrite (IO6, LOW);
 digitalWrite (IO7,HIGH);
 delay (1);
 digitalWrite (IO7, LOW);
 digitalWrite (IO8,HIGH);
 delay (1);
 digitalWrite (IO8, LOW);
 delay (1);
 
}

thank you freddy for that part number, but overall what do you think of my solution?
« Last Edit: September 22, 2016, 03:16:52 am by neo »
A hopeless addict (and slave) to TEA and a firm believer that high frequency is little more than modern hoodoo.
 

Offline FreddyVictor

  • Regular Contributor
  • *
  • Posts: 164
  • Country: gb
Re: Vacuum fluorescent display to clock display
« Reply #24 on: September 22, 2016, 08:34:07 am »
i think i have the power supply figured out, nothing stops me from using multiple transformers of which i have plenty, i also made my own chip out of transistors (schematic/explanation below)
 the arduino code that makes my plan work for those into that kinda thing (tested, not on the vfd yet but ive yet to find any faults in my design or plan)
 <snip>
thank you freddy for that part number, but overall what do you think of my solution?
If you're going for discrete transistor solution, then you'll need help from others on this forum as I'm not an EE, more programmer !

my experience of VFD's is from this project which I reproduced

I can't see a schematic so it's not really possible to put the code you posted into context
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf