Vacuum fluorescent display to clock display

[neo]

i believe there was a misunderstanding i have no intention to use the transistor as a 2 input device what i mean was to combine outputs, like below above (reply#193)

That looks like two signals per transistor to me.  You've got an [An]signal and an [IOn] signal to each transistor then you are paralleling the emitters which cant ever work properly.

Try this:

I've attached the LTSPICE circuit in case you want to extend it or add pulsed sources ground etc. and simulate it

as i understand it theres the 5v logic signal that turns it on (saturation) then from there until it is turned off it acts as a short circuit
but as you said i dont grok transistors so im merely explaining why i thought that and not arguing for it

[Ian.M]

Do persevere with the SPI and Arduino.  Its a skill that's worth cracking because it opens up the use of so many clever chips that can save you much money and effort building complex circuits.  However to get your VFD clocks going can be done in pure logic + those Micrel High side drivers, and I think I agree you'll get more satisfaction that way.

I suggest you make it modular so you can 'borrow' the VFD as a complete subassembly for other projects - put the MIC2981 chips, the +HT boost circuit, the filament supply and the VFD all on one board with a pin header for all the segment and digit control lines to the MIC2981 chips.  IIRC your VFD had 21 grid and anode pins so that lot would fit nicely on a 34 pin or 40 pin 2 row box header you could cannibalise old floppy or IDE cables to make mating cables for, leaving enough pins for plenty of grounds and 5V pins.  (As a rule of thumb: Use twice as many ground pins as 5v pins.)  If you connect a 'bare' pizeo sounder disk to a spare MIC2981 output, with a lower value pulldown - say 33K, you could use it as a fairly loud beeper for an alarm.   Just gate a squarewave audio tone.

e.g. to do it in pure logic, 1.024KHz from your '4060 divider chain ANDed with the 2Hz final output of the '4060 to get a 'beep-beep-beep' effect and also with the output of your alarm latch which would be triggered by your alarm time match circuit.  The easiest way of entering the alarm time would be to decode each digit to individual output pins and use rotary switches or jumper wires to program it.  Hours >=24 = alarm off. The selected digit pins would all be ANDed together and applied to the set input of your alarm latch, with the reset input connected to a pushbutton.

[Ian.M]

as i understand it theres the 5v logic signal that turns it on (saturation) then from there until it is turned off it acts as a short circuit
but as you said i dont grok transistors so im merely explaining why i thought that and now arguing for it

OK, lets give you a sim to poke around in and hope it 'clicks' for you.

I've created a simulation test jig for the circuit that shifts some 5V logic signals around for all the segment inputs while alternating [IO0] and [IO1] to simulate multiplexing:



The top plot is all the logic signals.  They all use 5V levels so I've had to scale and offset them so they don't overlap vertically to make them clear enough to understand.
The bottom two plots are the VFD segment outputs straight from a P channel MOSFET.  The 10K resistors in the gate circuits act as a potential divider so the MOSFET gate drive is limited to 12V relative to the source to avoid blowing out the gate oxide. They also act as pullups for the NPN transistors. You can probe anywhere around the main part of the circuit, but be careful which plot pane you last selected as you really don't want raw voltages or currents in the digital pane.   If it all gets FUBARed, reload the plot settings file.

Ignore the gubbins in the box on the left  - its the simulation test jig and contains some fairly gnarly LTspice logic.

Edit: I've improved the test signals to make it easier to draw up a truth table from the plot, tweaked a few component values and added the grid drivers. See 'Discrete MUX2.asc'.  Copy the plot settings file from the zip and rename it to match.

[neo]

the now was a typo

[Ian.M]

Was the LTSPICE Discrete MUX simulation any help to you?
Do you use LTSPICE?

[neo]

i do use ltspice and its useful to have it mapped out, thank you

[Ian.M]

Bump: Full code posted for my SAMSUNG SSED HNM-05SS62T VFD clock https://www.eevblog.com/forum/projects/vacuum-fluorescent-display-multiplexing-problem/msg1041506/#msg1041506

[neo]

OK i have come up with an idea i don't know if it sucks or not. Take a 4017b counter and wire it in to where it pulses buffers and the grid pin, some vfds i have would be totally cool with only 15 or 18 volts others i could use the MIC2981 on. As for the control circuitry (buffers) a logic level signal goes in to a tri state buffer and then a MIC2981 circuit like described pretty much unchanged from your idea except for the means of pulsing grids. The appeal of a 4017b is that its expandable, capable up to 10 mhz at 10 volts vcc and (this is provided this idea doesnt suck) i actually understand everything thats going on here. Also as a post note is there any multiplexer chip than when you pulse clock it just changes and doesnt require anything more complex than that