First post, by WJG6260
- Rank
- Member
Hello everyone,
I've been reading this forum for years and finally managed to start something of a collection of my own. Amidst my growing interest in general old hardware things-mostly spurned by wanting to just learn more generally about computer hardware and electronics-I recently purchased an FX-3000 motherboard. I know Phil and some other members have discussed this board in pretty great depth, but I've had some weird quirks with mine since the start and wanted to ask for some general advice. For reference, it's this board: https://stason.org/TULARC/pc/motherboards/E/E … 86-US-3486.html
This is and was my first 386 board of any kind; nearly all of my experience is with 486s and their associated boards.
When I purchased the board, the seller described it as being new. I'm under the impression he was either correct, or that the board was used only very slightly. What's strange is that the original bios that the board came with was an AMIBIOS dated sometime in 1990. It, however, seemed incorrect for the board and a better match for the ECS US 3486, as it had settings for an OPTi chipset. Further, it only recognized the first 640k of RAM, no matter how many varying configuration I tried and how many different 30-pin SIMMs I used. I grabbed an EEPROM and flashed it with the bios from Phil's board, and then later tried the MR-Bios. I found the MR-Bios extremely preferable.
I've been tinkering with older parts for a while, but managed to make a mistake while testing out the board. I popped in an 80mhz oscillator, and decided to try the board with my 386DX-40. I have relatively poor eyesight, and managed to assume that the white triangle on the end of my 386DX-40 was the corner for pins A and 1. I did not notice, until the board just would not boot, that the markings were on the opposite end. I powered the board down, turned the CPU around and tried again. The 80mhz oscillator was quite hot to the touch, so much so that I let it cool off, removed it, and popped in another that I had. I had a POST card in the machine, and it just read "---," so I presumed the board might be dead.
But I then tried my TI 486DLC at 33mhz, with a 66mhz oscillator, and it POSTed! The memory count was accurate, and everything worked. I did some stress testing, and surely enough there were no issues there either. I powered the board down again, popped in a Cyrix Fasmath Cx83D87-40P, and that worked too. I then switched back to the 386DX-40, and, sure enough, it posted at 33mhz.
After testing various other configurations, it seems that the board POSTs just fine and operates with any combination of 386 and 387 or with any 486 with an oscillator less than 80mhz.
I'm suspecting-but not certain-that I may have damaged an internal trace in the board, especially given the heat from the oscillator. In that case, I am okay with the outcome, seeing as the board works at 33mhz or less, and seems perfectly suited for use there.
That being said, I'd love to get it working at 40mhz. Any and all thoughts on this would be greatly appreciated!
I've attached a few photos for reference. There should be one of each CPU I tried, and the board as it sits now. I put it away temporarily, so there's nothing attached to it at the moment.
Thank you all again for all of your help!