Thomas, Rats! After that, it is all gravy for yours! I have to look at scraps of papers and find it. That was 3 years ago. I still need to find P75 PSU, video board I suspected either one is not working right making pixels noisy on tops of lit items. (usually first 2 lines of every lit pixels.) Silly I was disassembled the P75's PSU hence looking for another. Oh, I think it is not 200V, less than that. I'll have to find a working P70 (think my friend may still have it.) And you also need to have DC-DC converter that goes between panel and three DC power supplies. Your panel should have that small board behind the panel's. Both P70 and P75 panels interchange after moving the pins around on the power plug at panel's side. P70 is 386DX (either 16 or 20MHz), P75 is real 486DX 33 and uses different PSUs between them. I'm curious about your project using these plamsa panels? Cheers, Wizard

Thomas> Well, I'm an electronics student at the university of Twente Thomas> in Holland. I got the board from a fellow student, who Thomas> didn't know what to do with it. The screen runs on 25Mhz, Thomas> reading 4 bits of pixel data at a time. Almost half of the Thomas> connector pins are ground, and the others are default vga Thomas> signal pins. It shouldn't be too hard to get some of the Thomas> pixels to light, I'd really like to know if it still works. Thomas> I don't think i have the DC-DC converter (if it's not on the Thomas> picture), but I can build a new power supply if I know the Thomas> voltages as they are coming from the DC-DC converter. Btw, Thomas> if you have any pictures i'd love to see them. Too bad yours Thomas> isn't working anymore, it's such a cool device :-) I only know plasma panels from earlier days, when they were applied by their original inventor to the terminals of the PLATO system. Those were orange color, on/off only (no grayscale, or "orange scale" I suppose). The power for those panels was quite complex and fairly critical. It wasn't a DC voltage at all, but rather a high voltage AC square wave called the "sustainer voltage". The voltage was chosen so that a pixel, once lit, would stay lit (the sustainer would "sustain" the discharge) but pixels that were off would stay off. You could then turn pixels on or off by pulsing X and Y wires either in phase (turn on) or opposite phase (turn off) -- just like coincident current addressing in core memories. So those plasma panels had inherent memory, no refresh needed. I don't know how far the panel you're looking at differs from those PLATO panels. paul

Tony> [...] Tony> I have a plasma display on my HP Intyegral, and I am darn sure Tony> it gets continuosuly refreshed, The video controller, on logic Tony> board B, is an HP custom chip with some DRAM hung off it. The Tony> signals to the display itself are very similar to those used to Tony> drive a dot matrix LCD panel -- there seems to be a dot clock, Tony> horizontal sync (start of a new line), vertical sync, and a Tony> number of data lines. Yes, that certainly makes sense. I didn't mean that PC type plasma panels would use the inherent memory feature. But it still wouldn't surprise me to see the complicated high voltage AC main power supply just as on the original design. paul

It would surprise me though, :) These newer displays update pixels at 25Mhz, switching a high supply voltage on and off at this rate and getting a square wave is somewhat impossible. But even if you can, the radio magnetic interference emitted from the wires would mess up any signal far beyond the display board. You want to keep any signal leads short and next to a ground plane or some other grounded leads. My board even has shielding magnets glued to the display. There's also a rather large zener diode on my board, often used for voltage stabilization by keeping the voltage at a fixed level. I tried to get it to conduct by slowly incrementing my lab power supply. It reached it's max at 30V still nothing had happened. It does give another clue though about the display needing a large stable supply voltage. Btw, that LogicDart looks like a great tool, bit expensive for a student though :( Thomas. -----Oorspronkelijk bericht----- Van: cctalk-bounces(a)classiccmp.org [mailto:cctalk-bounces@classiccmp.org] Namens Tony Duell Verzonden: zaterdag 1 mei 2004 0:47 Aan: cctalk(a)classiccmp.org Onderwerp: Re: IBM ps/2 P70 Plasma screen voltages It wouldn't suprise me either... Alas I know little about the display panel in the Integral. The rest of the machine was mostly stnadard chips (68000, TTL, DRAMs, ROM, etc) and HP parts that I had data on (1LB3 HPIL chip, Thinkjet printer chipset, etc). The only thing I had to puzzle out was the display controller, and that didn't take too long However, the display seems to be all custom chips on the back (well, there were a couple I could identify, and many more I couldn't). And because of the high voltages, I didn't want to probe around too much with the LogicDart (which has a maximum input voltage of 40V -- it's fine for a board of logic, even if there are RS232 drivers on it, but not on a plasma panel). So I indentifed the signals on the input connector and left it at that. -tony