Let's repair an old 286 mainboard Jaton JAM-2301-V1

hello and welcome in my last video i shared with  you guys that huge pile of hardware which i got   from stevor a longtime viewer of my channel he  generously donated some really rare and exciting   hardware pieces and in the future i am going to  take the parts one by one and introduce or repair   some of them in my videos and i'm going to take a  closer look at this really exciting 286 main board   today as i already said in my last video this main  board has an amd 286 cpu with 12 megahertz and an   absolutely warping amount of memory on board four  megabytes as sim memory modules plus one megabytes   of onboard soldered memory making 5 megabytes  of total memory an unbelievable number for our   main board of that time which almost nobody could  afford back then what else can we find here the   two bios roms are here and on the left there are  some 16-bit iso slots but as you see on one slot   the second part is desoldered and a little bit  longer this board already has a lot of ram but   such longer slots were used for memory expansion  cards to get even more ram onto the system   such cards were usually very long with a lot of  dip sockets for the memory ics what else a useful   feature on this board a super i o controller  with a floppy and ide controller as well as   serial and parallel ports back then usually such  controllers were on a dedicated isocard such an   on-board controller was an absolutely unusual  feature for a board of this age an apropos h how   old is it actually here we see siemens 87 but this  is not a date code more a copyright message the   western digital floppy controller chip here also  says 87 this chip says 18 9 in the date code this   one 89 as well that one too so i would say that  this main board was made around 1989 maybe 1990. okay as i said these main boards suffered some  damage from a leaky battery and as far as i see   someone already scratched the oxide of the traces  a little bit but the damage doesn't look to be   too widespread i will have to remove some parts  clean everything and see if some traces are broken   i usually don't suggest to start the board before  the battery damage was cleaned completely since   current can amplify the chemical reaction and  already damaged traces can be damaged even further   but here someone already cleaned the damaged area  and looks like the battery electrolyte was also   already neutralized so let's give it a short  try first of all i will remove the memory to   reduce possible error sources and since i will  wash the board afterwards i'll need to remove   the memory anyway i really don't like this type  of sim sockets the memory modules are hard to   pull out of them and the old and brittle plastic  latches are snapping off really easily if possible   i always replace these slots by something more  steady a brief test for short on the power rails   5 volts and -5 volts look good this is ground -12  volts and 12 volts seem to have no shorts as well   very good let's connect the at power cable  never forget black to black and all of them   in the middle i will also add a post analyzer  card to see if the board is actually alive   and we are ready for the first try   and the board seems to be totally dead no  numbers are running on the post analyzer at all   at least when i turn on the system the reset  led lights up shortly and goes off again that   means that at least the reset circuitry is kind  of alive and the cpu should start the execution   anyway let's remove the bias chips and the cpu  from the sockets i'm going to wash the board   later and it is not absolutely necessary and i'm  often too lazy to remove the acs from the sockets   but since i have to look into the sockets and  clean them as well to exclude any problems with   contacts it's worth to remove the acs this  time as i already said luckily the damaged   area is not very large so i will have to remove  only some parts around the keyboard connector on the back side the main board looks as good  as new very nice a small desoldering notes   always add a little bit of fresh solder to the  old joints this simplifies the disoldering process   significantly if the pins don't want to come out  by themselves don't pull on the parts or you could   rip off the pads from the board instead wiggle  with a pair of tweezers a little bit on each leg   to loosen them gently afterwards you should be  able to easily remove the parts from the pcb so i removed all the parts which were in  the damaged area and well under the keyboard   connector it still looks really bad a lot of  corrosion and probably broken traces as well   and under the ics everything looks quite good  actually i don't see any damage there so far still   it is always worth it to remove surrounding ics  since the worst damage can hide underneath like   under the keyboard connector in this case  those parts do still look quite good though   so i'll not touch them to avoid  unneeded stress for the pcb   so it's time to neutralize the residues of the  battery electrolyte using some acid if you don't   know why i'm doing this i made a video about it  some time ago please feel free to take a look   however i'm always getting some questions on this  topic for example which acid i'm taking exactly   this is a normal white vinegar which you can buy  in any supermarket this one is highly concentrated   at 24.9 percent and is a bit too aggressive so i  usually mix it uh at about 50 with water to get   the concentration down to about 10 to 15 this  works really good and i'm usually using this   solution however it smells very unpleasant so  there are alternatives and i got some questions   indeed which other acid can be taken for this  procedure and the answer is simple you can take   any weak acid solution you have at hand it just  shouldn't be too aggressive and attack the copper   traces and for example today i am going to take  a pure lemon juice of course without sugar this   works just as fine it is not too aggressive  and smells nicely of lemons instead of vinegar   after 10 minutes you already can see that it  dissolves most of the oxidation and what you can't   see is that it neutralizes the battery electrolyte  residues stopping it from further reaction   now what i have to do is to scrub  it a little bit with a toothbrush and after some scrubbing the main board has to be  washed to get all the dirt dust and the residues   of the chemical reactions from the board and  this is another common question which i get   how to wash the boards properly as i already said  i often don't even bother to remove the socketed   ics in this case i removed most of them but i  didn't remove the memory parity chips from the   sockets for example you should always remove the  batteries however everything else doesn't matter   so why do you need to wash such hardware well you  will need the hardware the soap a brush it should   be quite soft to avoid any damage on the board  and water that's all the secret first you have   to use the brush to get rid of the most dirt then  add some soap and use the brush again to wash the   pcb properly especially in the place where you  used the acid many people are afraid of even   saying the words water and electronic parts in  one sentence but this fear is not really justified   water doesn't kill the electronics on its  own however salt in the water are kind of   what you have to think about if the water is  distilled or contains no salt in other words   you can easily wash all your electronics with it  no problem but if you use the normal tap water   there are souls in it and when water evaporates  you see these typical lime stains left behind   this is salt and this is what can be dangerous but  it's usually not a big deal you just have to drive   the hardware properly but first let's leave the  board soaking in the water for 10 minutes or so now the board has to be washed thoroughly with  clean water to get all the soap off the board   take your time and use the brush again to wash  all the rest dirt and soap away and again don't be   afraid also don't forget to wash the backside once  you are done with washing put the board upright on   the edge like this and leave the board for some  minutes to let the water drip off the pcb as much   as possible the less water remains on the board  the less probably you will get the lime stains on   it pay attention to have the slot vertically  like this so the water can run out of them   if you are unsure you can use a can of compressed  air to blow most of the water from the board   and from under the ics and  slots i usually don't bother after another 15 minutes or so  i put the board again in such   an upright position onto a slightly warm radiator   in a dry warm room and leave it there for another  24 hours to get it really dry if you are in hurry   probably even 12 hours would be enough but  to be on the safe side just take your time so about 20 hours passed and the main board  seems to be completely dry this is how the   corroded area looks like now the rest of the  corrosion has to be polished away from the pcb   therefore i'm using such green hard  rubber tips for the dremel you can buy   such a set of tips from china for about  10 euros including shipping and taxes you just have to set the dremel to low speed and  gently go over the traces with a tip don't apply   any pressure take your time and be patient or you  will polish away the copper in a blink of an eye   polish only as much as needed but remove all the  oxide until you see the clean and shiny copper   and clean everything with some alcohol  afterwards that is important for the next step   the problem now is that the oxide was actually  a copper which was eaten away by the electrolyte   and turned it into salt now where it is gone the  traces are a lot thinner and weaker they are not   in spec anymore to transfer the currents  which they were originally designed for   if i would try to use the board now i would  simply risk to burn the traces therefore i   have to tin them again to regain some thickness  and therefore i put some flux all over the place   add a drop of solder to the tip of my soldering  iron and go over all polished and exposed traces   add some more solder when needed but  don't add too much solder at a time   or you will breach the traces so pay  attention to that but also don't be too afraid   this procedure is really easy thanks to the flux  the solder will tend to flow into the direction   of the exposed copper automatically for me this  is always a magical and very satisfying moment   after everything is tint and all the traces  are reinforced use again some alcohol to clean   the pcb also double check now that none of the  traces got shorted if you put too much solder   somewhere use again some flux and maybe a  shoulder wick to remove excessive solder after cleaning it's time to check for  broken traces by such a massive damage   there are usually some and they have to be fixed  again but in this case there are magically no   issues and i can measure continuity on every  trace unbelievable but true all the removed   parts seem to be not damaged and the values are  alright so i can solder everything back in place so i will not put the keyboard connector  yet since it is not necessary for the   test and it could be in the way should  i need to follow the traces underneath   i will add some deoxide to all the  sockets clean them and insert the ics in case of the cpu think of orientation it has  no protective key but in the socket there is   such an arrow or triangle if you want that  must be aligned with this dot on the cpu it is also important to remember which  bias i see goes into which socket   if you are interested i can make a short video  about why there are some main boards with two   bias ics and some with only one but for now  it's just important to know that this one goes   into the socket u9 as it is written on  the sticker the other one goes into u14 so once again the post analyzer  the power cable and let's turn on nope still the same error the board is still  dead and the fixed corrosion damage didn't   change anything so far still that was  absolutely necessary and now it is also   not dangerous to let the system run for a bit  longer what can we do to analyze this issue   well i'd like to measure if there is some activity  on the address lines that would give us a hint   that if the cpu is working at all therefore i will  use such a prototyping board which i designed some   time ago i will put the link into the description  where you can download the kycat project   i use this board to design and test my prototypes  and if you find this interesting as well feel free   to produce some of these boards for you as well  as all of my projects this one is also open source so the board can be plugged into the iso slot  and now using the oscilloscope we can measure   some signals on the address lines thanks to  the notation on the prototyping board i always   exactly know which signal i am measuring  let's try a0 and there is something now a1 yeah definitely there is proper square wave  so the cpu is definitely alive that's a good sign all the address lines seem to be okay  let's see what's on the oscillator pin and we have 14.3 megahertz that's fine and the isobus lock is at six megahertz  interesting it's half the speed of this   cpu actually this clock can be up to  eight megahertz so upgrading the cpu to   16 megahertz without overclocking the iso bars  would may be a nice option for this board later okay before i dive deeper let's remove the bios  roms once again and try to read them with the   tl8662 plus i didn't find this board in the  ultimate hardware database so i will have   to dump the bios roms anyway the ics are nmc 27c  256bq the tl 8662 plus software knows either b or   q so let's go with b reading the rom seems to work  but when talking about such old ics my suggestion   is to try to read it couple of times in a row  this seems to work okay so i saved the first   rom into a file and made the same procedure for  the second ic i read the content and saved it into   a file and then i clicked couple of times on the  read button and look what happened ah you see this   reading the ic has failed and the operation  hangs now this didn't happen with the first   ic but i will take two new old eproms  and copy the content they are onto so the replacement ics are ready let's test  them zero means even and one means odd i got   that reading problem only with one eeprom but  i will replace both of them for now this board   seems to be very rare and i didn't find any bios  images online for this model i could use another   bios from a similar board but i prefer to go with  original now just to avoid ghost hunting due to   some bios incompatibility in the future that's why  these eproms are a simple copy of the roms which   i pulled out of these sockets so everything  is already connected let's turn on and see   if we see some numbers on the post analyzer and  yay something is happening so the main board is   back to life at some extent seems like at least  one of the original roms could be bad indeed   luckily i still could read both of them and back  up the bios since it seems to be hard to find it   for this main board the question is how could that  happen well it's hard to tell now but my theory is   some kind of short circuit or some kind of current  peak look at those three blue decoupling caps   another one is here for example and another one  of them is here here and here interesting is that   they are all blue however those two guys are red  all other caps on this board are blue except those   two my guess is that there was a short on this  board once in the past and someone replaced these   two caps these are tantalum caps and they tend  to short sometimes and maybe they produce some   current peak which slightly damage the eeproms but  this is just a vog theory and is maybe completely   wrong whatever happened to the eeproms the board  seems to be alive now and for further tests   i will need a keyboard so now the keyboard  connector can be finally soldered in place so the next try now we can connect the keyboards  avga graphics cards which i currently have at hand   it is an oti o77 but that doesn't matter now the  important thing is to see a video signal at all   and furthermore i will add a compact flash  to ide adapter with a compact flash card   as a hard disk replacement luckily this main  board has all the i o ports directly on board   as i already said very unusual for a 286  board of that time everybody who watched   my channel regularly will recognize the compact  flash card where i have dos 622 pre-installed   with all the required tools drivers and even some  games so everything is in place oh one important   thing is the pc speaker so we can hear if there  are any post error beeps that can help a lot   now i think everything is ready for  the next test let's give it a try and we have a signal jayton super 286 system  2 and we get some check some errors which   are definitely true since we have no battery  attached and all the bios settings were reset   and there is a beep strike f1 key to continue  f2 to run setup we want to get into setup so f2   here we have all the errors related to the  missing battery that's okay and we are in bios   let's do some settings i have no floppy attached okay i played with the hard drive type ids and  looks like we can use type 48 to set up custom   drive and not 47 like it is usually well that  doesn't matter the compact flash cards geometry is   987 cylinders 16 heads pre-compensation and  landing zone shouldn't make any difference   and 63 sectors that makes 485 megabytes of  total size that looks right to me what else i can press page up ah and here i  can see all the shadowed memory and   other tweaking i will set bios and video  rom directly to shadow that should give   some performance boost if you don't know what is  shadowing watch my video about xt ide performance   which i made a while ago everything else  can be left unchanged for now let's reboot the system hangs and refuses to  boot from the compact flash card   i can see that the led on the compact flash to  ide adapter is lit up all the time means that the   main board is trying to access the compact flash  cards all the time but somehow that doesn't work   well at least we have some activity on the ide  port that is a good sign unfortunately the board   doesn't show an error message anymore so i can't   even enter the bios let's remove the compac  flash card from the slot and restart once again now i get hard disk controller failure  and i can enter the bios once again   let's see type 48 987 1663 all the values  seem to be right that is unfortunate so couple of minutes later i tried to plug the  compact flashcards in and out and also tried to   use the type 49 instead of 48 like before but  that didn't make any different as expected so   let's try something completely different  first of all i will set the hard drive to   not installed now the compact flash card can be  inserted back into the slot unfortunately i still   can only enter the bios if the compact flash card  is not installed and now i would like to reference   my series of videos about xd ide once again so  if what i am doing now looks for you as a magic   please feel free to take a look at that  series and now i will take this network   adapter with pre-installed xd ide bios rom and  hopefully this will help to get the compact   flash card in this machine up and running it  was very important to set the hard disk to   not installed so the xd ide can take over the  detection completely and would you look at that   xt ide found the compact flash card indeed so  the onboard ide controller is probably alright   well high mam sees complains about insufficient  memory that's right we have only one megabyte   from which we took some for the shadow let's  start the check it and see what it says a 286 with only 640k of base memory that's  not much indeed and if i try to start   commander key in game it complains about  insufficient memory too well that is possible   to reconfigure and optimize the memory  layout to get commander kin running over   with the one megabytes of memory but as you  remember on top of the one megabytes on board   memory this main board supports additional seam  memory i have here four times one megabyte memory   modules which were previously in this board as i  got it so let's push them into the slots and see   where we get unfortunately the slots  in this board are really bad and i   will probably exchange them in the future  but for now let's just hope that they work and yay it counts up to four  megabytes of extended memory the buyers complains about invalid configuration and indeed it detected that the amount of memory  has changed there we see memory 640k and extended   memory four megabytes actually the board has one  megabyte memory on board but 384 kilobytes are   reserved and not shown here one more thing  i would like to do i will enable shadowing   for the xd ide bios at d8000 address this should  drastically improve the hard drive io performance and as you see this system boots and  heim sees doesn't complain anymore about   insufficient memory let's see what mem tool says yes now we have 628 kilobytes of free conventional  memory that should be enough to run any real mode   those game you want it seems to be happy as well  and reports four megabytes of extended memory   and finally let's see a game running on this  great piece of hardware a time correct example   would be commander keen and now it should have  more than enough memory to run and it does i'm so happy to have this super nice and  rare 286 board up and running again the   corrosion looked fortunately worse than it was  and none of the traces were completely damaged   a little bit of polish and tinning did the trick  however you hopefully saw once again how important   it is to look under every part the corrosion  under the keyboard connector was really bad   and would continue to damage the traces if  remained undiscovered also never forget that   xd ide bios is a great helper if you want to  see if an ide controller is working at all   many old bios versions seem to have problems  with more modern drives or compact flash cards   and such a network card with additional rom can  do some magic off off-camera i also tested the   original eeproms once again and the odd one in  the socket u14 seems to fail indeed the main   board didn't boot reliably if i use that luckily  i could still read that one in my programmer   so what left to say for today this is a very  nice board and it can not only be used in a   nice retro build but also heavily upgraded  the fpu socket is still empty as you can see   the cpu is a 12 megahertz one and luckily  it is socketed so it should be possible   to upgrade it to something beefier therefore  i'd also need to replace the quartz oscillator   but all that is a topic for another day  and today i hope you enjoyed this repair   and maybe you could take something with you  as well and even if it was just a little fun   just as always please leave your feedback below  likes dislikes whatever you feel is right for you   and if you like to watch such repairs  please keep an eye on the channel   since i'm going to make more of such videos in  the near future and we'll take a look at the   other main boards which were lately donated to my  channel and for now i say thank you and goodbye

Channel: Necroware

Views: 16,781

Keywords: retro, hardware, soldering, repair, review, nerd

Id: hc8Zut9DDXk

Length: 31min 40sec (1900 seconds)

Published: Sat Dec 18 2021