I'm leaving for DC tomorrow so it'll be some time before I can post new photos, circuit boards, or source code. But here's where things stand:
New CRT
Chris had a lead on some inexpensive Tektronics XY displays. It took awhile for the seller to ship two of them, but they arrived and are great! Bigger than the scope, without a reticule, and much sharper than the scope. The downside is that the XY bandwidth is said to be 2MHz. They're re-badged but I think they're Tek 620s. Anybody got a PDF of the service manual they'd send me?
New board
The other problem is that these displays don't have the same degree of flexibility that the scope has—The only gain adjustment is inside the box, and its range is not that large. So I was forced to make an amplifier board. It uses 3/4 of a ST TS274ACN quad op-amp, in a non-inverting configuration. Even though the amplifier's output is supposed to go within 50mV of the low supply voltage, I went ahead and used a split-voltage supply. The easiest way I had of doing that was to use my good old PC power supply, with its "GND" hooked to the op-amp's VCC-, its +12V hooked to the op-amp's VCC+, and its +5V hooked to the GND of the VGA signals. This gives around 6.5V of swing, which is enough to cover the screen. A second design would probably use a single input voltage (up to 16V) and a 1.2V regulator to split the supply, giving around 14V of swing. This would be enough to use the monitor from a real classic vector game like Asteroids, which would be great if those monitors were affordable. One recently went on ebay for $250 shipped, which is about twice what I'd consider spending.
Same software
With a few tweaks for the inputs—increasing voltage on X or Y moves down or to the left, increasing voltage on Z increases brightness—I'm running more or less the same software as I was for the 'scope. The problem I'm dealing with now is that when there beam moves from 0,0 there's a noticible overshoot, especially when the destination value is over 50% of the full range. Chris suggests that this error is like that of improperly tuned scope probes, while I wonder if it's either inherent in the VGA DAC (the effect was visible before the amplifier was added, and doesn't seem to have been exacerbated by its addition) or related to the transient seen when Z changes. For now my plan is to try to fix this in software.
Future directions
Once I've done that, I want to fine-tune the patches to MAME—always use the full screen width, instead of a width dictated by the game ratio, for instance. Supporting MESS or testing on Windows might be next steps after that. While the vector terminal is neat, the "real" use of this project in my mind is for emulation. Why, someone might even pay money for a vector MAME box. You can legitimately buy the ROMs for Asteroids and a few more Atari vector games from StarROMs, though sadly Star Wars is not on their list. Oh well, color games aren't ever going to work under my scheme anyway. While they're running their "15 free credits" program, for around ten bucks you can get Asteroids, Battlezone and Lunar Lander. I'm not affiliated with StarROMs.
Running really odd video modes is another thing that might improve the quality
of the display (since the CRT is forced to (0,0) less frequently). Apparently
the ATI graphics chipset in my laptop is capable of a display of 8192x128 pixels
(or more lines if necessary) though when I configured my X server to use this
mode and display a regular X application, the scope showed nothing on the color
outputs.
Entry first conceived on 26 May 2005, 0:36 UTC, last modified on 15 January 2012, 3:46 UTC
Website Copyright © 2004-2024 Jeff Epler