Trying to find VDD on 4-Line-by-96 point VFT

[Cliff Matthews]

I hope to run this VFT from my Arduino without the TTL character board (since it's missing the ROM). I have powered this up with 5-volts only and verified many clocking signals are all over the pcb. My next step is to guess what VDD (+ or -) I should send to get the VFT to turn on.
On the VFT pcb pic, I've noted pins on the 20 pin header in case anyone's seen this configuration before. Any ideas?

[Paul Moir]

I think I see a filament supply on the VFD board:  could you verify this with just the 5v supply on?  Should be about 3v AC - you could check with an oscilloscope.  It'll be between two of the pins that are just about shorted on the VFD display itself. 

Does VDD run to P1 on that power connector on the old driver board?   That could be the VFD's anode supply (typically 20-30Vish).  Is it connected to the capacitor you called out?  That will verify it's polarity.

Betcha your 9th line there clocks in the 8 bits.  Is each character column 7+1 pixel high?

[bktemp]

Interesting type of display. I have never seen anything similar.
From the connections of the VFD it looks like a 32x96 matrix. Therefore it is most likely 1/32 multiplex.
1/32 multiplex needs around 35-60V, so 30-40V ist a good starting point unless you can measure it somewhere on the PCB the display was connected to. If the display is dim you can later increase the voltage but I would start with a lower voltage to avoid damaging pixels if the multiplexing fails.
Those 2 large ICs seem to be the grid driver for scanning the 32 rows. There are 16 pins + 2 NC on each side going to the pixel rows on the display.
The SN74390 2x 4bit counter is probably counting the line numbers.
I can not see the pins from the top and bottom of display on the PCB. There must be a second PCB hidden between the display and the back PCB.
There is this 10 pin flatflex cable. I don't think there is a stepup hidden on the second PCB.
VDD is thererfore most likely around 50V.
For the timing signals I expect at least the following signals:
- Bit clock for shifting in 96bit per row
- Bit Data in (maybe more than 1 input)
- Latch clock for storing the 96bits and maybe triggering the row counter but maybe it is seperate
- Row counter Reset
- Enable/Blanking signal for dimming and disabling the outputs to avoid ghosting when changing rows

The filament could be derived from the 5V logic supply or there are 2 additional signals for an external AC supply.

[Cliff Matthews]

VDD does run to pin 1 on the power connector. Pins 2 and 3 are gnd and 1 is 5+
Those 2 large IC's are actually 4 x 17pin ceramic sips of unknown data (each 2 are stacked and folded over for space).

There is another board sandwiched in, just behind the VFT - I will take a side pic later..

[dom0]

Interesting type of display. I have never seen anything similar.

Looks like a front luminous (reverse) VFD to me. The anodes are directly at the front glass, the grids behind those and the filament is at the rear. This has the huge advantage that you won't ever see the grid structure or the filament glowing (if it's very dark). Contrast is also better, since every part of the glass that isn't covered by an anode is painted black.

[helius]

I'll go out on a limb and say that the points KA-1, KA-2, KK1, and KK2 have something to do with the filament.

[amyk]

Those 2 large IC's are actually 4 x 17pin ceramic sips of unknown data (each 2 are stacked and folded over for space).

What are the markings on the other side of them?

[Cliff Matthews]

What are the markings on the other side of them?

They are glued together, I just tried to separate with a razor blade - no go won't budge..

[Cliff Matthews]

Here's the bottom view..

[helius]

You've probably already surmised that the 32 pins on the sides are for the grid electrodes, one for each line (each group of 8 are for a line of characters, 7 high + 1 underline), and the 90 (96?) pins on the top and bottom are for anodes, each anode is a vertical strip. The flex cable must be for a shift register that controls the anode drivers on the board epoxied into the middle of the sandwich. The four ceramic SIPs are grid drivers.
I'd suggest testing the KA-* KK* pins pairwise for resistance, those will be the filaments
One other idea: if there is a net that is common to a particular pin on each of the SIPs, and also goes over the flex cable, it must be the Vbb.

[Cliff Matthews]

You've probably already surmised that the 32 pins on the sides are for the grid electrodes, one for each line (each group of 8 are for a line of characters, 7 high + 1 underline), and the 90 (96?) pins on the top and bottom are for anodes, each anode is a vertical strip. The flex cable must be for a shift register that controls the anode drivers on the board epoxied into the middle of the sandwich. The four ceramic SIPs are grid drivers.
I'd suggest testing the KA-* KK* pins pairwise for resistance, those will be the filaments

I get 2 megs from kk1 to kk2.
VDD is indeed positive and is sent to KA-1 and KA2 (each after passing through a 2 meg resistor).
VDD also goes directly to the last pin of each of those ceramic SIPs.   

[Cliff Matthews]

I think I see a filament supply on the VFD board:  could you verify this with just the 5v supply on?  Should be about 3v AC - you could check with an oscilloscope.  It'll be between two of the pins that are just about shorted on the VFD display itself. 

Does VDD run to P1 on that power connector on the old driver board?   That could be the VFD's anode supply (typically 20-30Vish).  Is it connected to the capacitor you called out?  That will verify it's polarity.

Betcha your 9th line there clocks in the 8 bits.  Is each character column 7+1 pixel high?

I didn't get 3v AC it was 330mv AC and 4.5vdc.

Hooked up 5v and 32v VDD but no pretty blue, not even a flicker.  After 10 mins the 17-pin sips were barely even warm.. Looks like further testing will involve bread boarding TTL or re-soldering header pins on the logic board and finding an old 2716 Eprom.
To the shelf..