Path: utzoo!utgpu!jarvis.csri.toronto.edu!rutgers!sun-barr!texsun!pollux!killer!chasm From: chasm@killer.DALLAS.TX.US (Charles Marslett) Newsgroups: comp.sys.ibm.pc Subject: Re: Orchid ProDesigner VGA+ (Problems.. & vugif 1.00) Summary: Check with Orchid -- new drivers are out! Keywords: VGA problems & vugif 1.00 problem Message-ID: <8504@killer.DALLAS.TX.US> Date: 1 Jul 89 04:57:28 GMT References: <24760@srcsip.UUCP> Organization: The Unix(R) Connection, Dallas, Texas Lines: 139 In article <24760@srcsip.UUCP>, rogers@falcon.SRC.Honeywell.COM (Brynn Rogers) writes: :: In article t36273c@saha.hut.fi (Ari Huttunen) writes: :: > :: >Hey! I've been an (un)happy owner of the Orchid ProDesigner VGA+ for some :: >time and I'd like to know if any of you have had similar experiences.. :: :: I also bought a Orchird Prodesigner, and made it a plus for $52 :: I am happy with it though. :: :: >The main problem is that the VGA-card can't seem to handle 10MHz bus speed.. :: >(I have a Mitac MPC 2000s motherboard) Instead I have to lower the speed to :: >6MHz to get the card working. This seems a little rough. Surely a new VGA :: >card should be able to handle 10MHz!? If I use 10MHz speed, in a graphics :: >program a part (10%) of the pixels go anywhere in the screen. This happens :: >also if the program writes through the BIOS.. If I put the card in an 8-bit :: >slot, it doesn't have this problem. (even in 10MHz speed..) Using the jumper :: >to set 8-bit operation doesn't have the same effect. This appears to be a problem with the Tseng Labs ET3000 chip in the board. STB (the company I work for) makes a similar board and it also has problems with several faster 386 boxes. Some work great, some don't, and speed is only part of the issue (a 25 MHz Everex works, a 16 MHz Zenith works, but a 20 MHz Micronics doesn't?). In our case, at least, the problem only goes away completely with our 8-bit card (even in an 8-bit slot, the 16-bit card has occasional misplaced pixels). :: My DTK clone runs the bus at 10 Mhz also. Standerd bus speed (and the speed :: most cards are designed to handle) is 8 Mhz. As I mentioned above, some do and some don't, I'll post a message if I ever figure out what it takes to fix the problem! :: In graphics mode maybe 1% of my pixels miss there mark, and in text mode :: I have seen 2 characters (out of zillions) with different color/attribute. :: :: I called Orchird and they said that it might help if their card was in the :: slot closest to the power supply. I moved it and it seemed to help marginally, :: but still didn't fix everything. Here I can only agree, nothing I have tried has really fixed the problem. :: What DID fix the problem completely is to use the VGASPEED.SYS :: driver that comes with the board to load the BIOS into RAM. eats about :: 30k of your memory, but works great. This is equivalent to a fix we have installed on a couple of boards, but it does not seem to fully solve the problem (the difficulty is that in some graphics modes, especially, parts of the BIOS ROM are not readable). :: >Problem 2: in 800x600x16 and 800x600x256 graphics modes there is a plainly :: >visible dark vertical line at the right side of the picture. About .5 cm :: >from the side. (2.5 cm = 1 inch, if you didn't know ;) I can't get it to :: >dissapper. And I'm quite certain it DOESN'T come from the monitor! (Which is, :: >by the way, TVM MD-11A) :: :: I haven't seen anything like this. I am useing a princeton Ultra-sync untill :: I have enough money for a NANAO 9070s. Note that Most VGA moniters are :: not rated for 800x600, just 800x560 (if that even). moniters that are :: rated for 800x560 may give wierd results like this. :: I haven't seen any really decent moniters that can handle 800x600 for :: less then $500 [for mail order or wholesale] If you are seeing this when running Ventura Publisher or other software that turns on the border, it probably is a monitor problem (in a sense). What is happening here is that the border actually extends past the start of the hori- zontal retrace, so a shadow of the overscan color wraps back over the first (or last) few pixels on the screen. A similar problem can occur on the left side of the screen if the video is unblanked before the retrace completes. In both cases, the problem arises if the configuration tables for 800x600 are "tuned" to a monitor that has a faster retrace than the one you are using. ::