Im afraid just the postage one way might will cost more than another cpu from ebay with shipping.
These two links will give you an idea on how that kind of repair looks like
https://hackaday.com/2017/10/15/get-down-to-t … al-chip-repair/
https://hackaday.com/2019/03/21/a-steady-hand … hip-work-again/
Replacement pin wont be as sturdy as original one so plugging into PLCC socket is off the table. You have to plug CPU first, then stick wire into missing spot, bend it and solder into prepared sanded off spot on the cpu.

But I suspect the socket being the source of your current troubles. I dont recall computer not posting before, and its almost unheard of to have a CPU that randomly sometimes works and sometimes glitches out without overclocking/undervolting.

Thank you for the guide. Funny enough the replacement cpu I ordered was from Poland 😀

That being said, I have some updates:
After some swearing, removing and putting back cards on the pc, I was able to make it constantly boot without hanging, so I plugged the isa eide card and I was able to get to msdos and use checkit to try troubleshoot the issues.

There are visual glitches plus an error apparently with video memory? The video card never had such issues (but another isa card is on the way to me I will get it in the afternoon).

So after some hassle I was able to take off the psu to give a full look to the motherboard (mainly to try find the manual), since it's one of many Suntac 286 motherboards that are around.
New cpu is an amd 286 12mhz, the old one was 10 mhz and I found it interesting that apparently there are no frequency sel jumpers.
I was still unable to take off the drive tray, I hate those oem custom things... any idea is appreciated since I thought the right manual was: HT-286r while apparently not.
I also noticed that after the boot despite clicking on and off the turbo button it stops responding, while during the memory count, if it is slowed down the memory count is very slow, while with other setting is very fast.
It is quite interesting since there never been have the infamous leaking battery on this particular model of motherboard, so we can exclude leaking inssues.

I'm going to give up for today.
I was able to boot with mfm hard drive on the pc, it boots fine constantly, I installed a new isa videocard, no video errors nor checkit shows any visual glitch, but then there is another issue, the floppy drive stopped working.
I tried several (which worked before on 386 machine), just once I was able to read an empty floppy, I also tried different cables just to be sure.
Whenever floppy I put in gives reading general error.
I also noticed that when I shut off the machine (with the power switch) for a second or so the screen become green or with garbled text.
Could it be the psu the issue maker? It was the first thing I tested and it gave fine results as voltages.

We have cycled in this one thread thru two? computers and multiple cards and drives going bad.
Its almost 50 year old electronics in need of servicing. Might be cracked joints, corroded connectors/connections, shorting capacitors, sockets and even chips going bad. Im afraid you cant go around buying more 50 year old hardware expecting to just swap few modules around. Like with vintage cars, you either collect them as is, pay thru the nose for restoration, or as a third option become expert in repairs yourself. That means mastering soldering and fault finding. Microscope, oscilloscope/logic analyzer, theoretical operational knowledge:
https://artofelectronics.net
https://archive.org/details/URP_4th_edition https://archive.org/details/upgradingrepairi0 … muel_2ndedition
http://vtda.org/books/Computing/Hardware/ISAS … DonAnderson.pdf
https://archive.org/details/DigitalDesignAndC … erArchitecture/
https://github.com/pinczakko/BIOS-Disassembly … jutsu-Uncovered

Another option are modern designs using vintage parts. Easier to debug
https://www.homebrew8088.com
http://www.malinov.com/Home/sergeys-projects http://www.malinov.com/Home/sergeys-projects/xi-8088 https://www.smbaker.com/building-an-xi-8088-pc https://www.youtube.com/watch?v=MW5lApN0gWc
https://github.com/skiselev/micro_8088 http://dosaidsoft.com/wp/2020/09/08/8088-pc-clone-part-1/

That definitely makes sense but those modern design using vintage parts are usually very expensive.
Probably the motherboard has some quirks and the only tool I have is a multimeter :p
That being said step forward have been made since now it boots fine every time, it boots on a freshly installed copy of dos 6.22 and the mfm hard drive seems to works fine.
After so much hassle I'm quite motivated to fully restore it but since my first computer (when I was 😎 it was a 386 i'm quite noob with a 286 which seems to have many different settings compared to the 386.
A thing I could try is to install the eide isa card boot to the cf card and then check if this one the floppy works, but the very same mfm card was working like a charme on the 386 build.

So I had another idea, but apparently doesn't work.
The mfm card is this one: https://stason.org/TULARC/pc/hard-disk-floppy … ves-EV-332.html
Since the hard drive works, while floppy doesn't, what happens if I add an isa super i\o card?
I took a Gold Star Prime 2C Demo card, disabled anything except for the floppy, serial ports and game port.
The problem is that the EV-332 card won't allow the floppy controller being disabled but only change the address.
I put the jumper in the mfm card to give a later address compared to the goldstar card, doing so during the boot the floppy head seeks but then the bios give the floppy drive mismatch error.

Nemo1985 wrote on 2023-02-22, 09:19:

I was still unable to take off the drive tray, I hate those oem custom things... any idea is appreciated since I thought the right manual was: HT-286r while apparently not.

Removing the drive tray would require more visual cues than those you have currently on offer.

Nemo1985 wrote on 2023-02-22, 09:19:

I also noticed that after the boot despite clicking on and off the turbo button it stops responding, while during the memory count, if it is slowed down the memory count is very slow, while with other setting is very fast.

regarding the turbo button not working after boot, that could be due to a bios issue, or it might just be the way the chipset handles the turbo button.

Nemo1985 wrote on 2023-02-22, 09:19:

It is quite interesting since there never been have the infamous leaking battery on this particular model of motherboard, so we can exclude leaking inssues.

Are you kidding me? The fact that your mainboard already has a cr2032 battery holder soldered onto the location where the varta battery normally is located would indicate leaking issues do exist if the battery isn't removed before leaking.

weedeewee wrote on 2023-02-24, 10:40:

Are you kidding me? The fact that your mainboard already has a cr2032 battery holder soldered onto the location where the varta battery normally is located would indicate leaking issues do exist if the battery isn't removed before leaking.

I will have to disappoint you, the motherboard always had the coin battery holder, here is a page of the manual:

Also it's not the common 2032 battery holder since it's bigger.

About the other issues: I figured that the motherboard has an hard jumper to make the turbo button controlled by the keyboard (ctrl+alt+ for slow speed while minus for high).
I was also able to disassemble the case, it had some screws on the under side.

Nemo1985 wrote on 2023-02-24, 11:18:

I will have to disappoint you, the motherboard always had the coin battery holder, here is a page of the manual: photo_2023-02-2 […]
Show full quote

weedeewee wrote on 2023-02-24, 10:40:

Are you kidding me? The fact that your mainboard already has a cr2032 battery holder soldered onto the location where the varta battery normally is located would indicate leaking issues do exist if the battery isn't removed before leaking.

I will have to disappoint you, the motherboard always had the coin battery holder, here is a page of the manual:
photo_2023-02-24_12-17-25.jpg

Also it's not the common 2032 battery holder since it's bigger.

About the other issues: I figured that the motherboard has an hard jumper to make the turbo button controlled by the keyboard (ctrl+alt+ for slow speed while minus for high).
I was also able to disassemble the case, it had some screws on the under side.

well, looking at the photo of the board it does seem like the battery holder is an afterthought.
is it a CR2450 then ?

Do you have the full manual? does it contain which keycodes set the speed & io wait states, or are those more like bios settings.

weedeewee wrote on 2023-02-24, 13:37:

well, looking at the photo of the board it does seem like the battery holder is an afterthought. is it a CR2450 then ? […]
Show full quote

Nemo1985 wrote on 2023-02-24, 11:18:

I will have to disappoint you, the motherboard always had the coin battery holder, here is a page of the manual: photo_2023-02-2 […]
Show full quote

weedeewee wrote on 2023-02-24, 10:40:

Are you kidding me? The fact that your mainboard already has a cr2032 battery holder soldered onto the location where the varta battery normally is located would indicate leaking issues do exist if the battery isn't removed before leaking.

I will have to disappoint you, the motherboard always had the coin battery holder, here is a page of the manual:
photo_2023-02-24_12-17-25.jpg

Also it's not the common 2032 battery holder since it's bigger.

About the other issues: I figured that the motherboard has an hard jumper to make the turbo button controlled by the keyboard (ctrl+alt+ for slow speed while minus for high).
I was also able to disassemble the case, it had some screws on the under side.

well, looking at the photo of the board it does seem like the battery holder is an afterthought.
is it a CR2450 then ?

Do you have the full manual? does it contain which keycodes set the speed & io wait states, or are those more like bios settings.

I've found the full manual here:
Suntac 80286 Mainboards it's the HT-286r.
There is no frequency selection with the keyboard combination it can be selected the speed between 6 and 12 mhz, that's it.
There is the 0-1 wait state but the jumper is already set to 1, also the memory mounted on the board is rated for a lower speed than adviced.
I'm quite done with this pc , I also tried another psu (just in case) but the floppy drive won't work, while the hard drive works like a charme :\

Nemo1985 wrote on 2023-02-24, 13:59:

I've found the full manual here: Suntac 80286 Mainboards it's the HT-286r. There is no frequency selection with the keyboard com […]
Show full quote

I've found the full manual here:
Suntac 80286 Mainboards it's the HT-286r.
There is no frequency selection with the keyboard combination it can be selected the speed between 6 and 12 mhz, that's it.
There is the 0-1 wait state but the jumper is already set to 1, also the memory mounted on the board is rated for a lower speed than adviced.
I'm quite done with this pc , I also tried another psu (just in case) but the floppy drive won't work, while the hard drive works like a charme :\

Thanks,
I wasn't talking about the wait state, but the IO wait state, JP3 on the manual page you posted.

I haven't read through the whole thread, so if you have a working floppy/controller combination, you could try if that is working on this mainboard.
if that isn't working then... try to get it to boot from a hard drive/CF card/other storage device and add a soundblaster into the pc for which you have diagnostics software installed on the storage device and boot it and run the diagnostics. the dma tests in particular.
also there's the IMD software http://dunfield.classiccmp.org/img/index.htm imagedisk you can try for some further insights.

Though , since you're done with this pc. meh. Do whatever pleases ya.

The jp3 was set for keyboard control, since the jp1 is hard wired on the same way I thought to have a full speed or very low speed.
The problem with this computer is that with the provided mfm card now the hard drive works like a charme (a mfm without any bad sector!), I was able to achieve such behaviour installing the mfm card on the 386 with the hard drive, here also the floppy drive worked. On that pc the very same card, cable and floppy drive doesn't work.
I also tried with a cf card it works only with a specific super i\o controller. Otherwise it hangs after starting ms-dos.

There is probably something wrong I do not know if it's with the settings or with the motherboard but since I have no oscilloscope those are things I can't check.
I've almost 10kg of stuff to try, time to move on...

Back talking about the floppy drive, I somewhat fear it might be a 360kb drive (despite it shouldn't be due to age).
Anyway I actually have 2 floppy disks 5.25, both seems to be low density.
Also the first time I tried to put in the following disk, it was read fine, but the second time... boom, impossible to read and trying to mess with it gave this result:

After the damage was done I kept messing with it and the drive started to make a noise like grinding sand while "trying to read it".
Then I followed this guide (https://dfarq.homeip.net/5-25-floppy-drive-maintenance/) and cleaned the drive.

Then I tried to read the second disk, which worked fine, but apparently it is suffering from the same fate (I noticed just from the picture, to eyes it looks ok):

I'm sure the floppy were not stored in a safe way (a basement for years). So they can have the infamous mold, but how can I be sure of that and not that is the drive which kills the floppy?
I could buy 10 nos floppy or 34 used but "untested".

Any good guess what is killing what and how to proceed?

Some floppies can shed the magnetic material, it just peels off the mylar substrate. This is because of poor storage conditions, although I've seen some reports of poor quality media that will do that even with proper storage - so age is a factor as well. Nothing can be done about that, AFAIK. The floppy is gone and trying to read it will only foul the heads - which might even cause a build-up of dirt so big, and so quickly, that the force of the rotating floppy will rip the head off (too much friction and the head will be grabbed and pulled sideways).

Sadly there is no way to tell in advance how bad the situation is, you just have to be vigilant and listen to any unusual noises like squealing coming from the drive when the floppy is being accessed. Inspect the disk surface and drive heads after putting in unknown floppy, residue from one floppy can easily kill other ones that were good. This is what gets a lot of ZIP floppies (and drive heads, sadly).

If you really need to make a copy, and it looks like the floppy might still survive for a bit, you can use IMD to make an image to HDD. Your floppy and drive, you judge the risk. But do remember to clean the heads again afterwards to avoid damaging other floopies.

If your drive reacts to control signals you can rather easily tell if it's DD or HD (you're very unlikely to have a rare QD). Put a floppy in the drive, any floppy is fine provided it does not shed (even unformatted one) and run head alignment in IMD. Ignore the readout, just use the keys to seek/move the head. If you get a full swing of the arm with just 40 steps (and double-step is off in IMD settings) then it's a DD/360k drive. If the heads move only halfway then obviously the drive can do 80 steps so it's HD one.

Thank you for your complete answer.
My main concern was if the floppy drive damaged the disks, or the opposite. Since those floppies were looking good before (and it read good the first time I put it on the floppy drive).
I cleaned the head after the mess and it was a bit dirty (cotton fioc and IPA).
I only have those 2 floppy (I have a third one that is 1.2mb but was stored in the same conditions as the other ones), so I'm not sure if I should try it or not.
I will probably wait to receive the new floppies to give a proper test.

So I've tried another 4 floppies sent from a friend.
2 didn't work and they got scratched.
2 worked but when I took them out were scratched too. In a less but still visibile way.
I do wonder, is it possible that is the floppy drive that scratches the disks?
Because the guy told me that their floppy were kept in optimal conditions.

Never worked

Never worked

It worked for a brief time, just to write dir

works fine but has scratches already.

Obviously before put them into floppy drive they were not ringed like that.
I only show one face but also the lower one has the same thing.

After every try I clean the heads with a cotton swab wet in ipa 99%.

This is the status of the lower head (I was unable to take a picture of the other side):

The picture is pixellated but it's visible a black line, also the foam on the side seem quite old and most useless?

Nemo1985 wrote on 2023-03-02, 17:48:

After every try I clean the heads with a cotton swab wet in ipa 99%.

And that should remove any kind of dirt the heads might have. It is possible to lift the upper head via that arm it uses to rest, but in most cases it only lifts a bit a you should not try to force it further. Also, do not touch the head screws if you can't re-align the drive afterwards. And do be patient and careful while cleaning, the upper head is held in place by a spring (a thin, flat metal pad with cuts to allow it to tilt vs the arm a bit) - it can get bent or ripped off. It's not butterfly wing kind of fragile but don't scrub with any force.

So, scratches on the center ring, or near it (visible in the center hole) you can ignore. Some drives have more "aggressive" clamping mechanism than others and might leave bite marks. It's normal. Rings visible in the head cutout are not. Since this is on the upper side it would seem the upper head has something stuck to it.

Heads are like a letter U, but very shallow. One side (the dark strip) is the magnetic coil, the other is a flat rest for the opposite head. Both these parts must be clean and flat, on both heads. If you have a tool like dental mirror it would help you inspect the uppper head without the need to disassemble anything on the drive.

Thank you very much for your answer.
I wasn't worried about the scratches near the center hole (or the inner parts).
The main problem is instead the outer parts that get scratched on every disk, it's where the head moves when it starts to read the disk (I usually use it on dos so I just use a: and if I get a prompt I write dir to view the content).
I ordered a dental mirror and I should get it tomorrow, I'm not so confident to disassemble the drive, for sure I won't touch the screws for the alignment since I won't be able to align it back.
In your opinion how many chances there are that it's the drive damaging the disks?
The only definitive test I'd have in mind would be buy new disks but they may have the same problems if they were not stored properly (despite of being sealed?)

Nemo1985 wrote on 2023-03-02, 20:32:

In your opinion how many chances there are that it's the drive damaging the disks?

As I mentioned before, it's not uncommon for one bad floppy to foul the heads and that in turn ruins other ones. But in general a fresh deposit like this should clean right up with IPA-soaked swab. So it's possible the heads were fouled before, perhaps by accumulation of dust during use, and that left some residue that hardened and is difficult to remove. But IPA and some rubbing should get it off eventually.

Wikipedia has a nice photo of the upper head, hopefuly this link will work: https://upload.wikimedia.org/wikipedia/common … -write_head.jpg

The two flat surfaces are the beige colored ones, the rest is not on the same level (though obviously a lump in there could also scratch the surface). But those two must be clean - use some magnifying glass and an extra source of light if you can't tell. Inspect/clean the side edges of the head too, all the corners here should be rounded a bit and not sharp. The glue you see on the photo should not be sensitive to IPA (or at least I've never had any problems with that whatsoever).