VOGONS


SUNTAC 80286 Mainboard help

Topic actions

First post, by Predator99

User metadata
Rank l33t
Rank
l33t

This week I got this nice SUNTAC 80286 Mainboard. I think have tried everything I can do but it doesnt work 🙁

- keyboard LEDs are flashing on power-up
- nothing on screen
- no beep-code

I already have
- tried different VGAs
- re-seated all chips in sockets
- tried different jumper settings
- populated the SIPPs
- checked the BIOS EEPROM. Put them in the BOOT-ROM socket of a NE2000 card and read the content - see attached. Seems to be OK.

Filename
SUNTAC 80286.zip
File size
43.44 KiB
Downloads
192 downloads
File license
Fair use/fair dealing exception

I was not able to find the documentation of this board anywhere. Especially the SIPP option does not seem to be very common.
This one here is very similar, but with different memory configuration:
http://toastytech.com/manuals/Magitronic%20B233%20Manual.pdf

Memory banks 0 and 1 seem to be completely populated, so this should be OK. The 8x"KM44C256AP" are 1024 kb. This option is not even mentioned in the manual above. I do do not have other chips available to test. I have received the board without the SIPPs shown on the pictures, these were added by myself.

The Power/Turbo LEDs are also not working. The LED is only flashing when connected to the two left Speaker-PINs (+5V / GND).

Does anybody have a documentation for such board or any idea what else I can try? I dont know the manufacurer, the BIOS Chips are labeled with "TCL" and on the back there is noted "TMF M21 94V-0 8939".

Any idea is welcome 😀

Many thanks!

DSC_0209.JPG
Filename
DSC_0209.JPG
File size
4.17 MiB
Views
3437 views
File license
Fair use/fair dealing exception
DSC_0210.JPG
Filename
DSC_0210.JPG
File size
4.04 MiB
Views
3437 views
File license
Fair use/fair dealing exception
DSC_0213 (1).JPG
Filename
DSC_0213 (1).JPG
File size
4.5 MiB
Views
3437 views
File license
Fair use/fair dealing exception

Reply 1 of 40, by kenrouholo

User metadata
Rank Member
Rank
Member

Next thing I'd try is replacing that Dallas RTC which will unfortunately require soldering. While it's not a 100% chance that this is the problem, still, your best case scenario is that RTC having a little bit of charge left. It's either dead or close to dead, and many older machines with these things won't boot if the battery inside the RTC is dead. They are fairly cheap, you just have to be able to solder them or have a friend do so.

Maybe someone else can offer additional possibilities.

Note: Don't confuse the Dallas RTC with the 14T5101 delay line that is in the corner. They look similar but are totally different parts.

Edit: If you do replace the RTC, if you want, you should be able to find an IC socket that the RTC fits, solder that in, and then you could replace the RTCs easily - if you might still have this machine in 10-20 years when the RTC battery dies again, that'd be handy, or if you resold it, it might be worth a little extra to some people.

Yes, I always ramble this much.

Reply 2 of 40, by Predator99

User metadata
Rank l33t
Rank
l33t

Thanks for your reply! This Dallas Chip is already in a socket and I already tried to replace by another one I had in stock (dont know if this is still working, however).

But I am nearly sure that an empty Dallas will not prevent the system from powering up. It will prevent passing the POST, but an error message should appear on the screen? Or do you have other experiences?

Reply 3 of 40, by kenrouholo

User metadata
Rank Member
Rank
Member

I haven't personally encountered such a situation, but most of my experience with old computers is from when they weren't old computers. I'm sort of just getting back into retro computers for a nostalgia factor. But I have read accounts of others finding some boards that won't boot at all until the RTC is fixed, while others will generally work fine and just might forget BIOS settings and the time. It depends on the board.

Here's what you can do... Look at this image (which I did not create; I found this with a Google search): http://www.mcamafia.de/mcapage0/jpg/ds1287rw.gif (image comes from this page http://www.mcamafia.de/mcapage0/dsrework.htm )

You should be able to measure the battery with pins 20 (positive) and 16 (negative)... (Edit: Unless those pins aren't actually exposed, which, on second look at that pic, could be the case... Edit 2: EEVBlog forum people (love EEVBlog) talk about being able to measure it so there is likely a way to do so reasonably easily). Do you have a multimeter so you can do that?

The page actually talks about modding an external battery into one, which you could attempt if you wanted (if yours measures bad), or you can just use the info of which pins to test. That info could likely be found in the datasheet for the part, too, but this was just one of the first Google results I got which answered my question.

Yes, I always ramble this much.

Reply 4 of 40, by Predator99

User metadata
Rank l33t
Rank
l33t

Just gave it a quick try. Battery is indeed totally empty. I connected a 3V with a wire held by my hands... 😀 But nothing. Maybe I will try to get a soldering iron tomorrow.

Yes, maybe there are boards that will not boot without the RTC..but I dont think they will not power up without it?

Reply 5 of 40, by kenrouholo

User metadata
Rank Member
Rank
Member

If you want to find out which chips those BIOS chips are, you're going to need to remove that sticker with a razor blade, but you'll have to be super careful if it's an EPROM. EPROMs are erased with UV light and have a window on the top so manufacturers put stickers on them to stop them from being erased. But some manufacturers also put stickers on PROMs and EEPROMs (a 286 won't likely have EEPROMs though) just to be a pain in the ass (okay that might not be their reason but it IS a pain in the ass). If you do remove the stickers, feel for an indentation in the top and avoid peeling the sticker up that far. Also do it in as dark an environment as possible, and what lighting there is should not be directly overhead, and also don't use fluorescent lighting because it has more UV content than incandescent or LED bulbs.

The board is getting some degree of power so I do think there's a chance it initializes the keyboard and then looks for a signal from the RTC, but I am not saying this with 100% certainty, just that I do think it is a possibility.

I have no idea about jumpers for this board but I think the empty socket is for if the manufacturer wanted to use one larger [E]PROM chip instead of the split low+high (that's what the L and H stand for, LowROM and HighROM), and if I'm right then there will be at least one jumper setting that. Maybe follow some traces from jumpers and see if any jumpers have traces that could go either to the large socket or the two populated sockets; if you find them then that may control that. It would likely be one of the jumpers near those chips if so. It's not necessarily set incorrectly but if you do find a jumper that does this, it should obviously be set to the lines that go to the 2 chips.

You could also try swapping the two ROMs, but I doubt that's it. But maybe someone removed them in the past and put them back in incorrectly... Probably not. Edit: The PDF you linked mentions which chip goes into which socket (they'll be labeled on the board but I can't see the labels in the pic). Double check but it's probably not the issue.

You say this is a SUNTAC board but I suppose from the marking on the back it was likely manufactured by TMF, perhaps outsourced by Suntac.

Sorry if none of this is helpful... 286s aren't my forte, but I'll let you know if I think of anything else. Heck most of this was just me rambling 🙁 Sorry. I'm looking at the board pics and trying to get some info on its components so that's why I mentioned some random stuff I thought of while doing that.

Yes, I always ramble this much.

Reply 7 of 40, by kenrouholo

User metadata
Rank Member
Rank
Member

yeah measure all your voltages that's a good starting point too

also try loosening any sort of timing you can... For example, take off jumper JP5 to switch from 0 wait state to 1 on the main memory.

Also... did you say you drove 5v into speaker, or you measured it? never drive voltages into outputs. Hopefully you mean you measured it.

I found the jumper setting for the ROM selection and you're set correctly; that's #5 on the DIP switch and on means 2x128Kb which'll be what you've got. ROMs are in the right order, too, I found a label in the document.

Possible problem: Bank 0 looks to be the unfilled sockets to me? Or is that the SIPPs? The PDF shows bank 0 as being at the corner of the board.

Edit: PDF also shows 6 rows (columns in your pic), 3 for each bank and I only see 5 in yours; the SIPPs must replace some of that but I'm not sure what...

Last edited by kenrouholo on 2017-02-16, 21:13. Edited 2 times in total.

Yes, I always ramble this much.

Reply 8 of 40, by Predator99

User metadata
Rank l33t
Rank
l33t

As written, the ROMs are OK and I added their image files in my initial post (maybe somebody else needs them too). They are in the right order as there are markings "H"/"L" under the chips.
I assume the large socket beside is for the optional 80287.

According to the manual of the similar board DIP5 controls wheather the ROMs are 27128 or 27256. As the images are 2 * 16k I assume they are 27128, I tried both options however.

No, the board is most probably not manufactured by SUNTAC, but its the famous ST62 Chipset 😉

Many thanks for your ideas anyway..I will try to fix the Dallas tomorrow... 😀

Reply 9 of 40, by creepingnet

User metadata
Rank Oldbie
Rank
Oldbie

Just a possible suggestion, but you might want to check in to see if the RAM configuration is controlled by DIP switch or Jumper somewhere, the Octek REV 5.1 in my GEM Computer Products 286 had that setup IIRC. When I bought that computer in mid 2005 it was populated with 2MB of RAM on DIP IC's and I topped it out at 4MB of RAM on 30 pin SIMMS. 4MB is a popular ceiling for 286's using SIP or 30 pin SIMM RAM.

If not that, if those are 1MB SIPS, it could be not booting because it does not support that capacity of RAM (especially if it's older). IIRC SIP RAM came in capacities as low as 256K per Module, it's been a long time, but I've had a 286, 386, and even a 486 laptop that used SIP RAM. The 286 had 512K Modules, the 386 had 4 256K and 4 1MB Modules (5MB of RAM).

I was having similar symptoms on my 486 this week when the DX4 100 CPU I just bought for it burned out. So it could also be a bad CPU too, 286 Chips should not be too terrible. I always vouch for keeping a spare around in case of this sort of thing. That could also be worth looking into, though I've never seen a 286 with a bad CPU before.

If you are really into this hobby in a big way like I am could get your hands on a cheap Motherboard tester like the one I just bought and use that to find out where the bootup sequence is hanging at, that can provide a clue......

Oh, and I took a look at the first photo, Bank 1 is populated with DIP Memory - THAT could be your problem too, as I tried to get more RAM out of the GEM 286 with the Octek board that way and then found a manual for a similar board and it mentioned it WILL NOT POST if you mix memory types, that you had to pick from SIMM (or in your case SIP) or DIP RAM and cannot mix both types.

PS - Just found a close match to your motherboard on Total Hardware 99 (UCL hosted version) - http://www.uncreativelabs.de/th99/m/U-Z/33180.htm - It looks like you need to configure some Jumpers and maybe some DIP switches to get the memory configuration right so that could be the problem too.

Last edited by creepingnet on 2017-02-16, 21:16. Edited 2 times in total.

~The Creeping Network~
My Youtube Channel - https://www.youtube.com/creepingnet
Creepingnet's World - https://creepingnet.neocities.org/
The Creeping Network Repo - https://www.geocities.ws/creepingnet2019/

Reply 10 of 40, by kenrouholo

User metadata
Rank Member
Rank
Member
Predator99 wrote:

According to the manual of the similar board DIP5 controls wheather the ROMs are 27128 or 27256. As the images are 2 * 16k I assume they are 27128, I tried both options however.

Side-note: 128Kbit, 256Kbit, factor of 8. I messed up when I said KB, should be Kb. (Not sure if you were confused by that, just making sure.)

Yes, I always ramble this much.

Reply 11 of 40, by Predator99

User metadata
Rank l33t
Rank
l33t

As mentioned here
https://allpinouts.org/pinouts/connectors/mot … ard-pc-speaker/
The speaker has +5v on PIN4 and GND on PIN3 .. therefore when connecting a LED to both it lights up and keeps lighting. Therefore is assume the 5V is stable, but I have not measured it.
The speaker itself connects to Pin1 and Pin4 of the connector.

Yes, the memory layout is my main concern as the pdf does not match my board.

When you look at the photo: There are two brackets [ ] around the populated banks and they are marked "bank 0" and "bank 1". Therefore these banks should be populated. Which should be sufficient for powering up. However, I am not sure if these are the correct chips...

Reply 12 of 40, by kenrouholo

User metadata
Rank Member
Rank
Member

OK gotcha I did completely misunderstand what you meant with the speaker thing. That said, you won't necessarily see high amounts of ripple unless it's quite extreme, though I'm also not exactly saying I think your PSU is bad (but it could be).

I wonder if there should be a jumper on the RAS setting. I imagine those jumpers are all 3-way because they should be set (left or right), unless it's got 3 values (left, right and off). Maybe take that wait state jumper and put it on RAS temporarily and see if that changes. RAS and CAS are RAM settings (row access strobe and column access strobe).

Yes, I always ramble this much.

Reply 13 of 40, by Predator99

User metadata
Rank l33t
Rank
l33t

@creepingnet: Thanks for that link...!!!!! This seems to be indeed the board with the settings for the JP21/JP22/JP23 which I was not able to find! 😀

I dont think the CPU is the issue as it is getting really hot after a few minutes...

I will try testing tomorrow...thanks to you all for the moment!! 😀

Reply 14 of 40, by kenrouholo

User metadata
Rank Member
Rank
Member
Predator99 wrote:

I dont think the CPU is the issue as it is getting really hot after a few minutes...

How hot is "really hot"? Do you mean like uncomfortable to touch? Or is this an exaggeration of simply "warm enough to know it's getting some sort of power"?

Yes, I always ramble this much.

Reply 15 of 40, by Jo22

User metadata
Rank l33t++
Rank
l33t++

I would try a diagnostics card and see how far it gets. Even if there are shorts or power issues, such a card can be helpful.
In case of shorts, the LED for the affected power line is either dim or off.

kenrouholo wrote:
Predator99 wrote:

I dont think the CPU is the issue as it is getting really hot after a few minutes...

How hot is "really hot"? Do you mean like uncomfortable to touch? Or is this an exaggeration of simply "warm enough to know it's getting some sort of power"?

Early 286 chips in NMOS design can get pretty hot at times. That's perfectly normal - I'd be worse if they kept that heat inside.
That's why they were also made in ceramic package with heat sink (like LCC and PGA).
https://en.wikipedia.org/wiki/Chip_carrier#Leadless

"Time, it seems, doesn't flow. For some it's fast, for some it's slow.
In what to one race is no time at all, another race can rise and fall..." - The Minstrel

//My video channel//

Reply 16 of 40, by Predator99

User metadata
Rank l33t
Rank
l33t

So..I did some new tests...

The abbreviations RAS and CAS (row access strobe and column access strobe) have been new to me 😉 I expected these 3 Jumpers deal with the memory configuration because of their location.

As I received the board the RAM was populated as shown in the picture below

DSC_0217.JPG
Filename
DSC_0217.JPG
File size
1.36 MiB
Views
3287 views
File license
Fair use/fair dealing exception

According to the manual (thanks again creepingnet 😀
http://www.uncreativelabs.de/th99/m/U-Z/33180.htm

This belongs to configuration "1MB (A)" with (8) 44256 & (4) 41256

JP 21, 22 & 23 have to be open in this configuration.

--> Nothing happens

My SIPPs are populated with 9 x 511000 (1 MBit), making 1 Mbyte per SIPP:

DSC_0218.JPG
Filename
DSC_0218.JPG
File size
362.7 KiB
Views
3287 views
File license
Fair use/fair dealing exception

They should be supported by this board. 4 of them should match the "4MB (A)" configuration "(2) 1M x 9 + (2) 1M x 9".
I removed the DIPs for this test:

DSC_0219.JPG
Filename
DSC_0219.JPG
File size
1.26 MiB
Views
3287 views
File license
Fair use/fair dealing exception

--> Also nothing happens.

I also tried the wait state selector JP5 without success.

Yes, the CPU is quite hot, I can keep my finger on it for around 5 seconds...I assume 70°C. This is nothing unusual and it shows that the CPU is working. Its also not the first time I see that a 80286-10 is overclocked to 12 MHz. I assume this is the case for this board, too.

Most irritating is for me that I never heard a beep code during my tests. A RAM configuration error or RTC failure should result in one, however...

I think I will order a cheap diagnostic card from Hong Kong, I assume it will take some weeks to arrive then...

Thanks again for your support!! 😀

Reply 17 of 40, by .legaCy

User metadata
Rank Oldbie
Rank
Oldbie
Predator99 wrote:
So..I did some new tests... […]
Show full quote

So..I did some new tests...

The abbreviations RAS and CAS (row access strobe and column access strobe) have been new to me 😉 I expected these 3 Jumpers deal with the memory configuration because of their location.

As I received the board the RAM was populated as shown in the picture below

DSC_0217.JPG

According to the manual (thanks again creepingnet 😀
http://www.uncreativelabs.de/th99/m/U-Z/33180.htm

This belongs to configuration "1MB (A)" with (8) 44256 & (4) 41256

JP 21, 22 & 23 have to be open in this configuration.

--> Nothing happens

My SIPPs are populated with 9 x 511000 (1 MBit), making 1 Mbyte per SIPP:

DSC_0218.JPG

They should be supported by this board. 4 of them should match the "4MB (A)" configuration "(2) 1M x 9 + (2) 1M x 9".
I removed the DIPs for this test:

DSC_0219.JPG

--> Also nothing happens.

I also tried the wait state selector JP5 without success.

Yes, the CPU is quite hot, I can keep my finger on it for around 5 seconds...I assume 70°C. This is nothing unusual and it shows that the CPU is working. Its also not the first time I see that a 80286-10 is overclocked to 12 MHz. I assume this is the case for this board, too.

Most irritating is for me that I never heard a beep code during my tests. A RAM configuration error or RTC failure should result in one, however...

I think I will order a cheap diagnostic card from Hong Kong, I assume it will take some weeks to arrive then...

Thanks again for your support!! 😀

I don't really know but it might not be the case of bad capacitors?
I have 0 experience with 286 boards but i have experience with electronics, and once i had a similar issue with one modern motherboard(asus m2n-x) and it would initialize the keyboard and recapping the board solved it.
I also had a problem with one asus a7v8x-x that didn't booted, and the keyboard would initialize and the board made some crackling noises(like a geiger counter) and it turned out to be a faulty psu)

Reply 18 of 40, by kenrouholo

User metadata
Rank Member
Rank
Member
.legaCy wrote:

I don't really know but it might not be the case of bad capacitors?
I have 0 experience with 286 boards but i have experience with electronics, and once i had a similar issue with one modern motherboard(asus m2n-x) and it would initialize the keyboard and recapping the board solved it.
I also had a problem with one asus a7v8x-x that didn't booted, and the keyboard would initialize and the board made some crackling noises(like a geiger counter) and it turned out to be a faulty psu)

Well, perhaps. Tantalum capacitors which is what this board uses can go bad. They are fairly similar to electrolytic caps. They aren't vented, though, so they actually tend to literally explode when they get bad enough. (There are also ceramic caps here, but I don't think it's as likely that they'd go bad. They don't use electrolytic fluid.)

Yes, I always ramble this much.

Reply 19 of 40, by Predator99

User metadata
Rank l33t
Rank
l33t

OK, I checked most of the capacitors with this method as I only have a simple Multimeter:

https://www.reference.com/science/test-capaci … 83b105dd3ea036e#

All capacitors act like desribed. With exception of this one

cap.jpg
Filename
cap.jpg
File size
280.59 KiB
Views
3200 views
File license
Fair use/fair dealing exception

Unfortunately it is the only capacitor of this kind on the whole board so that I am not able to compare this behaviour with another one. Additionally, its the - in terms of size - biggest one - so maybe there is another explanation that both pins are not connected to each other?