There are (Texas and Samsung) RAM (and parity) chips onboard already right of the empty DIL sockets. The DIL sockets can be populated to upgrade RAM.

You should replace the missing BIOS EPROMs first and get everything working, then care about RAM upgrade.

You will also need a ISA riser card, otherwise you are restricted to a VIDEO card and the onboard floppy controller and no HD.

majestyk wrote on 2024-06-14, 04:11:

There are (Texas and Samsung) RAM (and parity) chips onboard already right of the empty DIL sockets. The DIL sockets can be populated to upgrade RAM.

Hi, thank you. Despite it came to my mind "Where are the BIOS eeproms" I thought, the 28dip-wide sockets are maybe for RAM, because there are four instead of the expected two for 32k odd and even. Are the two sockets for further ROM, like Basic or for xt-ide suitable?

I also thought that maybe the smaller dip20 or dip16 are as well for dram what actually is the case. (DIP or DIL?)

majestyk wrote on 2024-06-14, 04:11:

You should replace the missing BIOS EPROMs first and get everything working, then care about RAM upgrade.

Yes, I then found an image with EPROMs inserted, so that I'm omitting the idea, that those sockets are for RAM.

Also I thought, that the issue isn't due to lacking RAM, but probably due to lacking a BIOS. But I 'm not sich experienced, that I could easily identify those RAM chips.

majestyk wrote on 2024-06-14, 04:11:

You will also need a ISA riser card, otherwise you are restricted to a VIDEO card and the onboard floppy controller and no HD.

I got one with three ISA slots. I did removed the Mainboard from that PC, because I wanted that Foto and for investigating the issue.

So can you please tell me, what I have to take account of with reprogramming the EPROMs with off and even BIOS?

I bought a programmer some time ago but never used it or tried to dive into that operation.

I also don't know, if there are suitable and non suitable EPROMs. I purchased yesterday six "EPROM 27C256-100 32Kx8 100ns M27C256B-10F1 CMOS von STM DIC28".

I'm interested in learning how to repair defective legacy vintage Hardware, though I'm a complete beginner.

If your programmer supports the STM chips, you´re one step further.
Then you have to find image files of the BIOS. It´s either two files 32KB each or one file 64KB. (you can split it into 2 halves before programming the chips.)
This could be a challenge, since this BIOS must support both the C&T chipset and the I/O stuff like floppy controller, COM and LPT ports...

For a start you could browse the Vogons database for similar mainboards:
80286 BIOS image collection

Majestyk, thank you. I'll try then #8 286-MS0010-2.zip, because it comes with phoenix v3.06. Isaw a photo of the startup screen of that Unisys computer with phoenix v3.07.

I found also two on minuszerodegree.

https://www.minuszerodegrees.net/at_clone_bio … _clone_bios.htm

BIOS name Phoenix 80286 ROM BIOS Version 3.05
Screen photo Photo
Supplier offensive_Jerk at the Vintage Computer Federation Forums (VCFF), on OCT23
BIOS chip type 27128 (quantity: 2)
BIOS contains the strings "80286 ROM BIOS Version 3.05"
"SRC 286 10MHZ"
"12/05/86"
BIOS download Phoenix 80286 ROM BIOS Version 3.05.zip
Comment For AT-class motherboards that are hardware compatible with the IBM AT (IBM 5170).
This includes AT motherboards containing chipsets that do not require the BIOS/POST to do chipset-specific initialisation.
Some later motherboard chipsets require chipset-specific initialisation, and this BIOS will not do that initialisation.
Comment Some other BIOS' in this category (no support for chipsets requiring chipset-specific initialisation) are at here.
Comment A later version of this BIOS appears to be the '80286 ROM BIOS Version 3.07' above.

--

BIOS source Motherboard: From a Mitac branded 286 clone
Supplier bobba84 at the Vintage Computer Federation Forums (VCFF), on MAY17
BIOS chip type 27256 (quantity: 2)
BIOS contains the strings "MITAC Inc." and "(R04)"
"80286 ROM BIOS Version 3.07" and "(c) 1985,1986 Phoenix Technologies Ltd"
BIOS download Phoenix 80286 ROM BIOS Version 3.07 (R04).zip
BIOS chip photo Photo
Comment The chip type suggests that this is a 64 KB sized BIOS, but it is actually 32 KB sized (because first half of both ODD and EVEN is empty.)
Comment An earlier version of this BIOS appears to be the '80286 ROM BIOS Version 3.05' below.

Or Mr BIOS 286er?

Hi, I recently got my EPROMs and did the programming with the BIOS files from minuszerodegree with the Phoenix 3.07 BIOS.

I first put them into the upper two sockets from the 4 sockets and it didn't boot up.

So I switched off and even and then got it running.

It didn't boot up from 1.44 floppy drive, so I tried with a 5.¼" 360k drive and it did boot up from that.

I then prepared an image with qemu that boot up the Unisys startup program from a 4.01 msdos system for a pw² 500/16 with a 386 I also have. Unfortunately not every file fits on that 360k floppy. I may attach a 1.2M drive instead to it to gain the full Unisys support floppy with the Phoenix setup.com file to get the BIOS configured.

I could set the BIOS, but it eventually didn't keep the info in the CMOS and indeed the battery seems to be discharged.

Could this be the issue with 1.44 floppy drives, also after a soft restart? It didn't keep the floppy parameters and therefore only the 360k floppy drive works?

Also I got a memory error on the bootup screen.

Good job so far ! Yes if it does not keep the parameters then replace the battery and start over. Many boards will loose settings if battery is dead if a soft reset with psu on occurs....

Hm, I now attached 3,7V to the CMOS, and it seems, that some capacitors now , after attaching it, also keeps the values for some time, as I noticed.

I got one MFM drive type 1 and it seems to be defective. Also I tried different MFM controller cards. One seems to be either not supported by the board or BIOS or is defective, because the bios tells "hard disk controller failure" and with the other controller "hard disk failure" instead.

I also had, I don't know why a gate A20 keyboard failure with only detected 64k instead of 512k RAM. I reseated the IC and it did boot normally again.

I don't have any manual found for that board and therefore I don't know what exactly is up with the floppy controller.

It only plays well with my teac fd 55 configured for 360k. I attached also a Panasonic ju-475-4, but that doesn't work. Also an alps or another 1.44M floppy drive doesn't work. I first thought they might be 2.88M drives and the BIOS doesn't support it. It seems also that it doesn't support a drive b:, but that may be due to my confusion. Also I didn't tried two 360k drives yet.

Also it seems, that it makes completely no difference if a: is configured as 1.2m, 360k or 1.44m. it always only boots from the 360k.

So I don't know what that is related to, but I don't know if there is a jumper to disable the controller. So I'm not able to test a second floppy controller. I did attach one and configured it for drive b: but that didn't work either, although the drive seems to be seeked on post or bootup.

So I can imagine, that it might be a Controller issue, and it is maybe only capable to handle 360k, but I doubt that. I saw pictures of this computer with a 1.44 3.5" floppy drive and I think the computer is to new for such limitations. So maybe the BIOS doesn't fit that well as I expected.

Maybe there is a way to get the real BIOS and Manuals for that board alongside with the fitting setup program?

Startup from the pw² 500/16 with DOS 4.01 seems to work to some degree like also the setup.com that came with the Phoenix BIOS 3.07 zip file from minuszerodegree.

I have a couple of LPX mainboards (Unisys / AT&T) where you have to connect the floppy drive to the first - non-twisted - connector of the 34-pin cable.

It behaves really strange. The in board floppy controller works differently, if the mfm Controller wd-1006V-MM1 is connected to the riser card.

Also I got a lot of gate A20 errors and halts. I thought it was maybe related to overheating and bad seating of the keyboard controller.

But then it worked somewhat "flawlessy" and I also got the 3.5" floppy working as b:.

Strange also, that the floppies are working on boot, but if I take them out to read another floppy, it can no longer read it if returned, and also the other one isn't accessible. But the other one then is bootable and dir works as expected. The message is always "sector not found" even if I booed up from that 1.2m floppy with msdos 4.01 and can die and access that floppy after bootup, if I reject it and put it right back and close the slot, this error message appears and it is no longer accessible. I have to reboot. This particular floppy is a HD formatted in that drive as 1.2M.

I'm not sure if this is related to the 1.2M drive or the 360k formatted HD floppies, or if that has to do with the main board floppy controller.

I'll test my equipment. Maybe another floppy drive will help, or changing cables. Sometimes I can access another floppy in the Sanyo floppy drive, but then it couldn't access the bootup floppy again, because it cannot find the sector.

Your board is a newer rev than the 300 286 board manual I have (also in vogons in library as scans) but it shows three jumpers and one is for floppy controller.
The manual does say 1.2mb OR 1.44mb floppy drive and appears to use a straight short cable (from the poor images). Here is the section on the switches and jumpers.
You need to find which jumper is JP3 to enable/disable the onboard controller.

Hello, thank you majestic and horun for your helpful hints.

I now think, that the issue comes merely from the 5.¼" floppy drive itself, then from the controller on board.

If I change the floppies in the 3.5" the system recognizes the media change. In the 5.¼" it says in those cases "sector not found" and I have to reboot, even if I take the same floppy for a second out and put it back in the slot again.

I think maybe a false jumper setting.

I'll try with another floppy drive.

I also managed to revive a Seagate st -225. It didn't react on type 2, 6, 25, 39 or other BIOS drive types with 615 cylinder and 4 heads and 17 sectors (those types only varies in precomp, Landzone for parking).

I did read about the thing, that those MFM drives often are bond to their controllers on that they're low Level formatted. But I didn't think that they behave that dead, if not connected to that former controller.

I attached it to the MFM controller from the 500/16 (that computer shipped with a dead(?) NEC drive, so I had it spare). A WD 1006V-MM1.

I think, yesterday my tests were like sgtfmt3 and checkit didn't recognize the drive, but today suddenly llfmt from the Unisys startup disk did. Okay I used some not that hard shaking, maybe that loosened the head actuator? It produced some clicks noises, unfortunately the spindle only runs at 90° fast, not with the PCB at bottom. Maybe some oil can help?

I achieved to low level format it by llfmt and then I could fdisk and format it. Now I use calibrat from Norton utilities because I set the interleave at 1:1 but it needs 2:1 to be 88% faster.

I'm currently processing it with 80 patterns in calibrate.

Thanks for the manual hint!