I think enough has been said to convince us that an EDID solution is possible. We know enough to say that no features need to be added to the board or the chip to interface with a DDC injection box. That's important to keeping the OGD1 release moving, and continuing with Verilog code. Besides being one method of supporting fixed-frequency monitors for users, one thread indicated that the DDC injection box is useful to the project as a development test tool. Many details need to be investigated and decided, but they're not on the critical path to OGC1. That's important. One question is whether there's a PCB-mount mating connector for the DVI connector on the video board, with low reflections. If so, a PCB could be put in-line with the DVI cable and pass the analog signals through. If not, a better answer would be a breakout cable, with a DVI connector plugging into the board, and the video signals going one way to the VGA or BNC connectors, and the DDC lines going out another branch to a connector that plugs into the tool. The tool then would deal only with low-speed data, and could use a common type of shielded connector such as a DE9. Software Integrators would certainly have the expertise to design and manufacture a breakout cable. Maybe JRT or Howard has the answers about availability of both sexes of DVI connector for PCB mount.
-------------- Original message ---------------------- From: Dieter <[EMAIL PROTECTED]> > > A DVI-to-analog adapter is normally a cable. Good quality ones are > > available off the shelf. I haven't seen any that included a board or a box. > > However, if you're satisfied that an adapter can be made that splits > > off the DDC signals without affecting the video signals, then we can say > > that a solution exists in principle. > > Are DVI connectors available with solder cups and a screw-together cover? > Similar to those available for DB-25, DB-9, HD-15. Or do you need some > specialized $500 crimp tool? On the cable side, probably the latter. Assembling that connector to the coax lines without fouling up the impedance match has to involve some esoteric black magic. > We can test a prototype using the easy to find and doubtlessly less > expensive HD-15 connector and some existing brand-x graphics port. > > > I imagine the hardware handshake is complex enough to be most > > easily implemented with a microcontroller and on-board firmware. > > The first question is what do monitors have? Some small chip that just > does the DDC? Problem is it might be one-time programmable, and/or > require a specialized burner machine. Probably doesn't help us. We don't want a specialized programmed chip, we want direct user access to the data bits. > And is likely to be surface mount. > Has anyone come up with a good prototyping board for surface mount chips, > like the very handy ones for DIP packages? There are some in the Digi-Key catalog. We use them here in the lab. However, this will be a simple board to design, once we have all the information together. I see no reason to bother with a hand-built prototype. > > More than > > one user interface is possible. Probably the most universal solution would > > be to put in both the RS-232 port and the DIP switches; that way only one > > board would have to be laid out to satisfy different sorts of users. > > It occurs to me that if the tool is built with through-hole DIP ICs and > > switches, it can be assembled with ordinary home-shop tools, and sold as a > > kit. > > If you're thinking of making a PCB, devices could be surface mount, *IF* > the pins are spaced widely enough. The real problem isn't surface mount > vs through hole, it is the size. For de-soldering, surface mount is actually > easier than through-hole. The biggest problem soldering is finding a small > enough tip for your soldering iron. You also want skinny solder and a good > magnifying glass, but those are easy to find. It could be done either way, but I just find it easier to work on DIPs than SOICs. Also, DIP shift registers match the footprint of DIP switches better, and make the board layout easier. And, DIPs allow IC sockets, which may be worthwhile if the board gets used with the case off. Although, I think an external tool should have a case to help protect the insides against ESD. > But for a prototype, those solderless breadboards for DIPs are very nice. > > > One question: can it get its power from the OGC1, if the display > > doesn't send any power up the cable? > > According to others on the list, it is supposed to work with the display > turned off. We'd need to find a spec for the Voltage and for how much > current draw is allowed. Current drain shouldn't be a problem, as long as we don't try to run LEDs and things like that off it. 150 microamps is fairly typical for a small microcontroller. The shift registers used to interface the DIP switches to the microcontroller would take much less than that. I suppose the RS-232 interface might take as much as 10 mA to run the charge pumps. _______________________________________________ Open-graphics mailing list [email protected] http://lists.duskglow.com/mailman/listinfo/open-graphics List service provided by Duskglow Consulting, LLC (www.duskglow.com)
