• Please review our updated Terms and Rules here

8-Bit IDE Controller

The default address for the XTIDE is 300h. If you haven't changed the settings (which requires a new BIOS upgrade to tell it where the IO space is) then you should not be conflicting with the SB card. Is it maybe the MIDI stuff that is conflicting?

Anyway, there are 16 possible IO settings for the XTIDE, you should be able to find one somewhere that doesn't conflict with the SB card at all. Again, just make sure you update the BIOS at the same time as you move the IO space setting. I'd suggest keeping the SB at its factory defaults for maximum compatibility.
 
Thanks for the answers. My XTIDE is still at 300h. SB suggests 200h or 220h during installation. But both dont work. So i guess i have to try different settings in XTIDE. Which are these 16 possible settings?
 
cauldron_ said:
Thanks for the answers. My XTIDE is still at 300h. SB suggests 200h or 220h during installation. But both don't work. So i guess i have to try different settings in XTIDE. Which are these 16 possible settings?
Click to expand...
FWIW: Small tip - Start with a bar-bones system and pull all unnecessary peripherals out of your pc; i.e., sound card, etc. The sound card setup is fairly flexible. Use Microsoft's MSD.EXE to see where your assets are.
 
are there other cards in the machine? NIC card perhaps?
Take out everything you don't need to boot, including the XTIDE and boot to a floppy and run your tests. If it still doesn't work, then yea, it could be a bad sound card, but you really can't rule out some conflict with something else on the motherboard or even the floppy controller, whatever. You may have to try it on other machines just to verify it.
 
cauldron_ said:
I changed the base address to 320h. But it stills doesnt work. It seems that it cannot detect the sound card. I think there is no problem with XTIDE. The sound card setings are the factory defaults Maybe is a bad card or incompatible? (http://stason.org/TULARC/pc/sound-cards-multimedia/CREATIVE-LABS-INC-Sound-Card-SOUNDBLASTER-PRO-2-CT.html)
http://imageshack.us/photo/my-images/705/0d824.jpg/
http://imageshack.us/photo/my-images/98/0d825.jpg/
Click to expand...

i can say that i use the same CT1600 SB Pro 2 with the XT-IDE card, and it definitely works for me, in both a 5150 and a 5160.
 
Just for interest, there has been some exciting developments over on the 'dangerous prototypes' version of the board - discussion thread here. Basically there is a working prototype out there which used a programmable 'CPLD' chip to replace many of the 74-series ICs on the original version.
 
If I had the luxury of a CPLD on board, I'd rework the read/write logic so that word I/O was transparently done to the drive without the loopy "Chuck mod", so that both reads and writes would be sped up.
 
Hi Chuck, is this something you could be persuaded to offer some input on over there? Sounds like a great idea!
 
We have talked about the CPLD version on and off :) I posted some early work in a thread here. I would like to consider a second version with no discrete logic at all, only a large CPLD. That would essentially be a blank slate for you to design on. I'd really appreciate any input to the design process. There are a ton of great suggestions already, but I'm afraid I don't have enough background to address all of them.

I can make some of the easier changes like replacing logic with CPLD, header changes, interface/power pins/connections for CF cards, that kind of stuff. I'm lost when it comes to what extra pins you'd need if you wanted to toy with DMA.
 
I've already built a CPLD XT-IDE with no discrete logic for the PCjr - with Chuck's read optimization also applied in the write direction. Only problem is it's been sitting on my desk for a month without any time for me to test it. I've also layed out a board and written the code for an ISA version with literally only 2 ICs (CPLD & ROM). And everything is 5V through-hole - no SMT. I can provide the layout and the code to anyone who wants it.

I've finally moved into my new apartment closer to work this week. So more time is opening up without 2+ hours of commuting each day.
 
It's really dependent on the I/O pin count needed. I used either an Atmel ATF1504AS or ATF1508AS in a 84 pin PLCC. The PCjr design needs the 128 cells for some additional logic to decode RAM, paged flash, and a secondary PLD for POST display and RTC. The ISA design fits into the 64 cell part. ATF150x's come in a 5V version with on-board reg to make core voltage. Though their future availability may become volatile. But PLCCs choices in general are becoming more and more end of life. And I was shooting for a SMT free / hobby assembly friendly approach. You can buy ByteBlaster II knock off cables on eBay for $16 shipped that work with the Atmel ISP software.

Using a 3.3V part is still an option with a through-hole LDO reg. The only requirement is the part must have independent I/O voltage supplies (@5V). A 3.3V part with 3.3V I/O won't work even if it's qualified as 5V tolerant. It's likely it can't supply enough drive current for IDE. And the clamping diodes used to make 3.3V inputs 5V tolerant typically require external current limiting resistance on each pin.
 
Interesting, thanks. I'm not sure on the finer details but the dangerous prototypes board does seem to be working with a 3.3v CPLD.
 
I haven't looked at their schematic lately or part choice. Both problems may initially work. The low current drive problem may not work with all hard drives. Running 5V clamping 3V3 parts at 5V w/o limiting the sink current will also work, but it will burn the part out over a short time.
 
Just from the standpoint of protecting a part that needs programming, I'd use some buffers between a 3 volt part and the drive. I wouldn't worry about modern IDE drives, but some crufty old IDE drive may not be as kind.

Of course, this raises the parts count, but it's also just good practice.
 
You must log in or register to reply here.
Back
Top