• Please review our updated Terms and Rules here

XT-IDE Help

Holmes said:
I can't even get to my MFM HD or boot to a floppy. If I unplug my XT-IDE, I'm sure I could boot OK, but then how would I restore the universal BIOS?
Click to expand...
Remove the jumper that enables the BIOS chip. Boot the machine. Whilst the machine is running, replace the jumper. Now flash the BIOS.
 
That did the trick! Thanks a million. When I was building it, I was wondering what the purpose of those jumpers were. "When am I ever going to need these?". Now I'm thanking my lucky stars they're there.

Still can't figure out why I'm having so much trouble trying to flash the hargle BIOS, and why the associated utilites (like FINDCARD.COM and FLASH.COM) are not working on my XTIDE/EEPROM.

If I can't figure it out, I guess I'll have to wait until the universal BIOS has a version for the speed modification.

Anyone have any ideas on what else I can try to flash my EEPROM with hargle's BIOS or why I'm having these issues?
 
Chuck,

No, I'm afraid I don't. I'm probably going to have to get one of those, especially if I'm going to build an N8VEM.

I'm wondering, however, why I am running into this problem at all and it seems no-one else is. So do you think it's just a slight difference between my EEPROM and others which is revealed with the different BIOS versions/flash programs?

I'm new to building electronics, so is this just "one of those things" and I should ignore it, find a way around it, and move on?
 
As you get the correct BIOS with no ROM erros, then it's not the flashing that's the problem. If the flash failed, the BIOS wouldn't have loaded at all (in most cases), which it does in this case.

The modded BIOS you use, is that the one with special 286/V20 instructions?
 
When I tried using FLASH.COM with the /A switch, it DID fail. I ended up flashing the hargle BIOS 0.11 with the IDECFG utility from aitotat's universal BIOS package.

The version of the hargle BIOS I tried was NOT the modded version. It was the latest version 0.11, which is linked to on the XTIDE wiki (here's the link from the wiki: http://www.waste.org/~winkles/xtide_011.zip). Looking at the version history, I didn't see anything about 286/V20.

What I was doing was I wanted to test my setup with the original hargle BIOS because the speed modded BIOS is based off the hargle BIOS and I figured it was a closer match. I have NOT made ChuckG's speed mod yet. I was just testing with the original hargle BIOS to check if it would work BEFORE I made the hardware change and flashed the modded BIOS.
 
Holmes said:
When I tried using FLASH.COM with the /A switch, it DID fail. I ended up flashing the hargle BIOS 0.11 with the IDECFG utility from aitotat's universal BIOS package.

The version of the hargle BIOS I tried was NOT the modded version. It was the latest version 0.11, which is linked to on the XTIDE wiki (here's the link from the wiki: http://www.waste.org/~winkles/xtide_011.zip). Looking at the version history, I didn't see anything about 286/V20.

What I was doing was I wanted to test my setup with the original hargle BIOS because the speed modded BIOS is based off the hargle BIOS and I figured it was a closer match. I have NOT made ChuckG's speed mod yet. I was just testing with the original hargle BIOS to check if it would work BEFORE I made the hardware change and flashed the modded BIOS.
Click to expand...

Ok... The /A parameter hasn't really been tested so it should only theoretically work.

I don't know if IDECFG alters some data in the ROM image (or problably code in this case), but I'm pretty sure it does. Because of this, it's generally a bad idea to use it with Hargle's BIOS.

What you can try to do is to disable the data-protection using IDECFG. If you do this, you should be able to use FLASH.COM just fine without the /A parameter at all. If you do this, you may want to open the "Write enable" jumper if the flash is successful.
 
Per,

Yeah, I wasn't sure if flashing Hargle's BIOS with IDECFG was dangerous. Glad I didn't destroy anything.

I never noticed you could disable the write protection on the EEPROM with IDECFG. OK, I'll give this a try.

You also suggested disabling the write-protection jumper on the XTIDE. Is this is because I'm disabling the EEPROM chip's data-protection?

Thanks!
 
Holmes said:
You also suggested disabling the write-protection jumper on the XTIDE. Is this is because I'm disabling the EEPROM chip's data-protection?
Click to expand...

Yea. On the prototype where this jumper is not present, the smallest eletrical disturbances in the system can cause various parts of the EEPROM to be erased.

The setting for data protection is the "SDP command" option under the "Flash EEPROM" menu.
 
Last edited:
I really don't like the idea of modifying the BIOS as the configuration process. When I've done a card that requires storage of configuration data, I've found that a little 8-pin serial EEPROM works better. My concern is working with with the XTIDE, tossing it in a box and then wondering why it doesn't work with the new system you just got a few years later.
 
Chuck(G) said:
I really don't like the idea of modifying the BIOS as the configuration process. When I've done a card that requires storage of configuration data, I've found that a little 8-pin serial EEPROM works better. My concern is working with with the XTIDE, tossing it in a box and then wondering why it doesn't work with the new system you just got a few years later.
Click to expand...
We thought a little about this when the card were in the design phase. The main reason why it's stored in the BIOS is because it turned out to be impossible with the curent design to successfully autodetect the I/O port base (due to unused I/O ports returns different values for different systems).

When Hargle's BIOS was made, a given area in the EEPROM was being reserved for these settings. The big problem is that Aitotat's BIOS was developed completely independent from this, and thus it uses different settings located in different locations of the EEPROM. Because of this, the different flashing programs should only be used with their respective BIOS files.
 
Last edited:
OK, here's my update. It didn't work.

I disabled the EEPROM data protection via IDECFG.COM as per suggested. Then I flashed my BIOS with the original Hargle BIOS v0.11 with FLASH.COM.

It gave me a different error this time:

"Verify failed, unexpected byte found at the following address:
D000:0000 55<>7F

Press any key to reboot your system."

The last time I tried this, it gave me about 10 unexpected byte errors like this. When I rebooted, however, it have me an XTIDE error, but allowed me to boot from my other disks, so it was easy to fix this time. This is slightly better, since the last time it locked my system up on reboot.

Anyone have any other thoughts?
 
My only thought is that you still might have an intermittent connection somewhere. Are you in the USA? I (or any number of other folks) could burn you a 2764 EPROM with your configuration and mail it to you.
 
ChuckG,

Thanks for the generous offer! I'll definitely keep it in mind.

I'd love to figure out what's wrong before resorting to that. It's probably not worth sending it out to anyone, since my setup is currently working with the Universal BIOS. I was just interested in implementing your speed modification. What can I say, I love speed! I can just wait until the Universal BIOS has a version that supports the speed mod.

Just a question, though. If you think it's an intermittent connection, does that mean that if I repeat the flash with the same settings then the errors I get should be different each time? Would that be a good way to determine if that's the problem?

Thanks.
 
For all we know, FLASH.COM might be incompatible with your computer.

If you try the following sequence:
1. FLASH.COM - Hargle BIOS
2. IDECFG.COM - Aitotat's Universal BIOS
3. FLASH.COM - Hargle BIOS
4. IDECFG.COM - Aitotat's Universal BIOS
5. FLASH.COM - Hargle BIOS
6. IDECFG.COM - Aitotat's Universal BIOS

If during that sequence, you always encounted errors during FLASH.COM, but none with IDECFG.COM (assumption: it does a verify after write), then that suggests to me that the issue is not an intermittent connection.
 
Holmes said:
Just a question, though. If you think it's an intermittent connection, does that mean that if I repeat the flash with the same settings then the errors I get should be different each time? Would that be a good way to determine if that's the problem?.
Click to expand...

Wiggle and flex the card and see if you get exactly the same error message. If it makes a difference, you have your answer.
 
I'm much more comfortable with repeating the cycle like Modem7 is suggesting.

ChuckG, the thought of doing that to my XTIDE makes me a bit queasy! I don't think I've got the guts to do that intentionally. I can't help but wonder if that would create problems. That being said, I've taken the card out and put it back in at least a dozen times now which involves a bit of pushing and shoving, which is about as much as I'm comfortable doing right now.

I'll report back if I discover anything else.

Thanks everyone!
 
"Verify failed, unexpected byte found at the following address:
D000:0000 55<>7F
Click to expand...
This is the first byte of the ROM, but no matter which flash eeprom writer you're using, the value *must* be 55 or the ROM won't load. We don't know if the bits got stuck to 1 during the write or the read though.

I'd do this:
Rip the EEPROM out of the working card that you have and stick it in this one. If you don't see the boot menu, boot to a floppy disk, load up debug.exe and type the following at the - prompt:

-d d000:0

this will dump out the contents at memory location D000:0 where the eeprom is located. You *should* see data where the first two bytes are 55 AA. If you see 7F, FF, ... especially where the last 4 bits always make an F in the data, you have a problem on a few of the data lines. Hitting all the pins on the sockets and reflowing the existing solder that is there will usually do the trick. Just heat up each pin just enough so that you can move the pin inside the hole.
 
hargle,

OK. I can try that in a few days when I return home.

One clarification, though. Both of my XTIDE cards ARE working. The only problem I am having is that they only work with the universal BIOS 1.1.5 and IDECFG.COM, and they repeatedly fail with all my attempts to flash using BIOS 0.11 with FLASH.COM, even after removing the EEPROM write protection.

If what you said about the 1st byte is true, then I shouldn't be able to flash with the universal BIOS either, right?
 
You must log in or register to reply here.
Back
Top