Deunan wrote on 2023-07-21, 08:46:

Since you have a way to program chips, can you do a dump of the existing BIOS ROMs and link them here? Also, here's an AWARD BIO […]
Show full quote

Locutus wrote on 2023-07-20, 23:47:

I've already tried 286 BIOS PLUS from Tulip's MB (with similar chipset) with no success .
I only have four 27C128 EPROMs , therefore trying alternative ones (erasing/programing) is pretty time consuming 😉

Since you have a way to program chips, can you do a dump of the existing BIOS ROMs and link them here?
Also, here's an AWARD BIOS for this mobo, that I know will respond to CTRL+ALT+ESC because I have this exact mobo and it came with this BIOS. It does have some limitations though, there is no support for user-definable HDDs I'm afraid. So either you need to find something that matches the BIOS parameter table, or you need to use XTIDE. Or SCSI. My mobo works with MFM drive and controller but the SCSI card worked nice as well.

EDIT: Don't let the filename fool you, the date is from the last 16 bytes, where the first long jump is. The actual date of this BIOS is 10/20/87, this is AWARD 3.03.

Thanks - I'll try it.
I've edited my post with the picture of complete board and linked there BIOS 'images'.

The Phoenix setup s

Deunan wrote on 2023-07-20, 20:25:

It says seek failure, does it actually seek the floppy on boot? That is, the heads move forward and back? - If the heads don't m […]
Show full quote

Locutus wrote on 2023-07-20, 19:08:

The mainboard completes POST but there is another problem...
It has a very old PHOENIX BIOS with 'external' setup program.
But in order to run it I need to boot the system... I only have a 1.2MB 5.25" drive.

It says seek failure, does it actually seek the floppy on boot? That is, the heads move forward and back?
- If the heads don't move then you have some kind of FDC issue, try a different card maybe.
- If they do move then yes, it would indicate the BIOS is expecting to find a 40-track drive (so a 360k DD one).

BAD news is.... I borrowed NOS 360kB 5.25" drive.... it behaves exactly like 1.2MB one... "Diskette read error" 🙁
Tried on 2 different controllers, on other MB this 'combo' worked perfectly.

Predator99 wrote on 2023-07-21, 16:31:

Your BIOS does indeed not contain a setup. Its only 32kb in total. The Phoenix setup seems to be in another 32kb if you use 27C […]
Show full quote

Your BIOS does indeed not contain a setup. Its only 32kb in total.
The Phoenix setup seems to be in another 32kb if you use 27C256 ROMs.

There may be a jumper on your board to switch between 27C256 and 27C128?

I would go with a XT-IDE ROM. Put it on a network card and you can boot from CF or HD.

That is my plan...
I found in my 'supplies' PCB of XT-CF card:
http://www.malinov.com/Home/sergeys-projects/xt-cf-lite
I'll solder it and give it a try.

BTW. Has any one used this particular card ?

Locutus wrote on 2023-07-21, 22:26:

BAD news is.... I borrowed NOS 360kB 5.25" drive.... it behaves exactly like 1.2MB one... "Diskette read error" 🙁
Tried on 2 different controllers, on other MB this 'combo' worked perfectly.

Perhaps this BIOS expects a floppy not with OS but a "booter" (bootable floppy) with this setup progam. Maybe not even a booter, just written to specific sectors. So without this special floppy you won't be able to configure it.
Did you try the BIOS from my mobo? It's way more user-friendly and for IDE HDD support you could just try XTIDE. And it does support at least 1.2M drives (can't remember if it supports 1.44M).

Deunan wrote on 2023-07-22, 10:29:

Locutus wrote on 2023-07-21, 22:26:

BAD news is.... I borrowed NOS 360kB 5.25" drive.... it behaves exactly like 1.2MB one... "Diskette read error" 🙁
Tried on 2 different controllers, on other MB this 'combo' worked perfectly.

Did you try the BIOS from my mobo? It's way more user-friendly and for IDE HDD support you could just try XTIDE. And it does support at least 1.2M drives (can't remember if it supports 1.44M).

Hi there!
Not yeat.... 🙁
I have to erase EPROMs first.
This involves running my 'tank-like' DIY UV lamp :]

Locutus wrote on 2023-07-22, 13:47:

I have to erase EPROMs first.
This involves running my 'tank-like' DIY UV lamp :]

I have to erase some EPROMs too, usually I use an old tanning lamp but I'm going to try some experimenting with the sun. I will keep each EPROM in the direct sunlight for 5m and test it in the programmer, then add 5m more to be sure I have stable cells. I have used sunlight to erase EPROMs before but I want to have a better estimate on the time.

What you really need is *bootable* disk with barebones DOS with a bios configuration program copied to it.

By the way, you can also use 1.44MB 3.5" floppy drive on the 286 too, also boot with 720K floppy too. Back then when I was playing with 286 as toys, all I had was 720k floppy drive and does boot.

Cheers,

Deunan wrote on 2023-07-22, 14:27:

Locutus wrote on 2023-07-22, 13:47:

I have to erase EPROMs first.
This involves running my 'tank-like' DIY UV lamp :]

I have to erase some EPROMs too, usually I use an old tanning lamp but I'm going to try some experimenting with the sun. I will keep each EPROM in the direct sunlight for 5m and test it in the programmer, then add 5m more to be sure I have stable cells. I have used sunlight to erase EPROMs before but I want to have a better estimate on the time.

don't bother. it's been done already
https://hackaday.com/2016/09/14/staring-at-th … asing-an-eprom/

weedeewee wrote on 2023-07-22, 15:34:

don't bother. it's been done already
https://hackaday.com/2016/09/14/staring-at-th … asing-an-eprom/

See, this is exactly why I do bother. People come up with such ideas, do it once, and then it's taken as a gospel and nobody dares to question it.
I've done EPROM erasing with the sunlight before, in summer, and it took hours not weeks. And I didn't test it in-between so I don't even know exactly how many hours were needed for full erase.

The problem with that particular setup from the article is we don't know what kind of sun exposure this person has on their roof. At what time of the year (and which hemisphere). How was the EPROM burned, there are different methods and some leave more permanent charge trapped at the expense of erase count. That chip was powered all the time it seems which might have contributed to building a more robust charge trap, hence why it took so long. Not to mention the exact EPROM type since the older ones have bigger cells but also bigger structures behind the quarz window so can absorb UV more easily per given flux density. And this is just from the top of my head.

So, in short, I'd rather do it myself and come up with my own conclusions.

pentiumspeed wrote on 2023-07-22, 15:02:

What you really need is *bootable* disk with barebones DOS with a bios configuration program copied to it.
By the way, you can also use 1.44MB 3.5" floppy drive on the 286 too, also boot with 720K floppy too. Back then when I was playing with 286 as toys, all I had was 720k floppy drive and does boot.
Cheers,

Hi there.
I presume, you did not read the entire story 😉
My current struggle is to BOOT 😉
Parameters in CMOS are screwed.
I've tried 2 controllers, 360kB, 1.2MB, 1.44MB drive with no success 🙁
Same result .... "Diskette read failure".

Deunan wrote on 2023-07-22, 14:27:

Locutus wrote on 2023-07-22, 13:47:

I have to erase EPROMs first.
This involves running my 'tank-like' DIY UV lamp :]

I have to erase some EPROMs too, usually I use an old tanning lamp but I'm going to try some experimenting with the sun. I will keep each EPROM in the direct sunlight for 5m and test it in the programmer, then add 5m more to be sure I have stable cells. I have used sunlight to erase EPROMs before but I want to have a better estimate on the time.

Years ago I needed to produce ozone.... so I've used broken mercury-vapor lamp (quartz arc tube was intact) as a temporary measure.
Now I use it to erase EPROMs (2 minutes suffices).

Makeshifts last forever !

Deunan wrote on 2023-07-21, 08:46:

Since you have a way to program chips, can you do a dump of the existing BIOS ROMs and link them here? Also, here's an AWARD BIO […]
Show full quote

Locutus wrote on 2023-07-20, 23:47:

I've already tried 286 BIOS PLUS from Tulip's MB (with similar chipset) with no success .
I only have four 27C128 EPROMs , therefore trying alternative ones (erasing/programing) is pretty time consuming 😉

Since you have a way to program chips, can you do a dump of the existing BIOS ROMs and link them here?
Also, here's an AWARD BIOS for this mobo, that I know will respond to CTRL+ALT+ESC because I have this exact mobo and it came with this BIOS. It does have some limitations though, there is no support for user-definable HDDs I'm afraid. So either you need to find something that matches the BIOS parameter table, or you need to use XTIDE. Or SCSI. My mobo works with MFM drive and controller but the SCSI card worked nice as well.
EDIT: Don't let the filename fool you, the date is from the last 16 bytes, where the first long jump is. The actual date of this BIOS is 10/20/87, this is AWARD 3.03.

Hi.
I've tried with BIOS you provided.
Same behaviour as with Tulip's - "dead silent" 😒

Time to try XTIDE...

working ROM...

Predator99 wrote on 2023-07-24, 12:29:

Did you try to insert the ROMs in different order? Very unlikely nothing is displayed in the Post card if this is a working ROM...

I've swaped the chips - no difference - "dead silent"....

This is very odd, the BIOS should work. I can't tell what my chips were, exactly, as my mobo has either ROMs or OTP EPROMs with custom markings, but I can see on the photo yours had 27C128 with 200ns speed rating. These are pretty slow and I assume your replacement parts are also 200ns or faster? Perhaps the chips degraded and are not in spec anymore, that will most likely not show on the programmer during verification (the access time is slower than 250ns) but might not work in the mobo anymore.

Depending on how much time and money you want to sink into this project I would suggest getting some FLASH-based EPROM emulator chips. Like Winbond W27E512 (there are others too, not sure now if SST27SF512 belongs here or not). These chips emulate EPROM pinout (and programming method) perfectly but are actually electrically erasable - like EEPROMS. The only downside is these chips are usually somewhat limited in the number of erase cycles (like older EPROMs it's about 100 or so). Other than that they require no UV, less power, and are faster. With a couple of these you can experiment much more easily and also be reasonably sure it's not access timing issue.

I do notice some differences in jumper settings between your mobo and mine, but I'm not sure what J10 to J14 do, I don't have a manual and I didn't really do any proper investigation as the mobo works as-is. FYI J2 is turbo switch, J3 is external battery, J5 is PC SPEAKER, J6 is KEYLOCK / POWER LED, J15 is RESET SWITCH and J16 is TURBO LED (1: A+, 2: K-)

Deunan wrote on 2023-07-24, 13:08:

This is very odd, the BIOS should work. I can't tell what my chips were, exactly, as my mobo has either ROMs or OTP EPROMs with custom markings, but I can see on the photo yours had 27C128 with 200ns speed rating. These are pretty slow and I assume your replacement parts are also 200ns or faster? Perhaps the chips degraded and are not in spec anymore, that will most likely not show on the programmer during verification (the access time is slower than 250ns) but might not work in the mobo anymore.
Depending on how much time and money you want to sink into this project I would suggest getting some FLASH-based EPROM emulator chips. Like Winbond W27E512 (there are others too, not sure now if SST27SF512 belongs here or not). These chips emulate EPROM pinout (and programming method) perfectly but are actually electrically erasable - like EEPROMS. The only downside is these chips are usually somewhat limited in the number of erase cycles (like older EPROMs it's about 100 or so). Other than that they require no UV, less power, and are faster. With a couple of these you can experiment much more easily and also be reasonably sure it's not access timing issue.
I do notice some differences in jumper settings between your mobo and mine, but I'm not sure what J10 to J14 do, I don't have a manual and I didn't really do any proper investigation as the mobo works as-is. FYI J2 is turbo switch, J3 is external battery, J5 is PC SPEAKER, J6 is KEYLOCK / POWER LED, J15 is RESET SWITCH and J16 is TURBO LED (1: A+, 2: K-)

I'll 'check' those EPROMs by erasing them and programing with 'original BIOS'.
I'm playing with this borad for fun so there is no limit for time / expences 😉

I've soldered 'MALINOV' board but there seems to be a problem with it / BIOS image.... 🙁

BAD Checksum error has been solved !
One step further...

Time to play with xtidecfg...

Locutus wrote on 2023-07-24, 15:00:

Time to play with xtidecfg...

I'm not familiar with this design but it uses less chips than newer XTIDE revisions, and was done in 2013 so quite possibly it needs the config to be set to "XTIDE rev1".

64k ROMs.