Jump to content
Andrew96_

JPM system 80 cpu board repair

Recommended Posts

So this board found it's way to me not working of course, It had some previous repair work done

IMG_2752.jpg.dbbd4f693700ad419800dbe28e10d718.jpg

IMG_2754.jpg.96f04689719ce4b727584f69bad2caac.jpg

IMG_2755.jpg.0252afcea92e65b93678597f36100424.jpg

where the chips had been out on the topside and replaced the legs had been soldered to what was left on the pad, I supose for quickness but made it very difficult to remove the chips this time round

 

IMG_2759.jpg.348f0bdbe19425c88b94f52943cabe9d.jpg

IMG_2760.jpg.b94f2f1349ef7da74497619a8418d057.jpg

 

this end pin the track was broken right where it met what was left of the pad, and a wire was put under the board to replace it

 

IMG_2753.jpg.8b4240586d523c1907d7a303d2d159d4.jpg

so this is right under the battery! place where most damage is likely to occur, but nothing had been done

IMG_2757.jpg.d56894c8b8ab700c14972ac475ce2633.jpg

BUT shine a light through the board and........ all becomes clear!

IMG_2756.jpg.e2afbcc8b581a428623cb644ba7a7596.jpg

yes a broken via track!

 

took me about 3 hours to remove everything and clean up the pads, a good few missing on the top of the board where someone had removed the chips previously, my removed ones that had never been out but still had the crystallized solder came out fine eventually. I have no idea what solder was used but it seemed to have a very high melting point and some really nasty flux in it! never smelt such horrid flux before! but the board is starting to look better now and cleaned up good

IMG_2761.jpg.1c12064185a84b5f9b52805ce4e14a58.jpg

IMG_2762.jpg.7269150de43c72f1e40f453af7e092bc.jpg

 

IMG_2763.jpg.d1250ff9be852e8e58b00526f687eff5.jpg

 

ok so now the track repairs can begin! starting with the ones under the battery! just the one was rotted through but all via hole tracks done for future reliability

IMG_2765.jpg.b2f1e811515188670651e9c7ad9039d6.jpg

 

 

  • Like 1

Share this post


Link to post
Share on other sites

Looks in a bit of a mess, will be good to see what you do to this one.

Share this post


Link to post
Share on other sites

Certainly a thorough job there Andrew. You'd never get that done with a commercial repairer that's for sure. I wouldn't even want to think what they might charge either.

I do hate boards that have been got at as you end up with more faults to find and usually ones that don't make any sense.:tuttut:  

Share this post


Link to post
Share on other sites

well when I removed the 22k resistor network legs fell off! now I can see there was rot and corrosion which WAS present on the previous repair and one resistor was not working, so instead of replacing the resistor network one resistor was added and a hole drilled through the board to fit it! ok not too bad really but the hole drilled had caught the edge of a track on the topside! well for me reliability counts so I cannot understand why this 23p resistor network was not just replaced as it was clearly faulty back then damaged by rot!

 

extra hole drilled close to track! circled in yellow

 

IMG_2765a.JPG.e6a348832955f298966d390bd2a2ad83.JPG

 

and in yellow the extra resistor fitted because the resistor network right next to the battery had one rotted/failed resistor!

IMG_2752a.JPG.a52458cb856e22aa852bd978b55959f3.JPG

 

yes these are the botches I could do without!!

 

 

Share this post


Link to post
Share on other sites

Mike, I decided to put wires on the top tracks so they can pass through the IC holes to the bottom side where the pads are intact and are folded over flat on the pads, this gives room for the ic legs to also pass through the holes and be soldered on the underside, this way there is room to desolder the chip legs free should they be needed to be removed in say 30 years time! I did think I might use the rivets (or eyelets as there known in the industry) but I am still waiting for the tool to arrive that folds them over nicely and this makes the holes smaller with less leg clearance and more difficult to remove chips if needs be! It appears the person that removed the chips in the first place managed to damage every 0v and 5v tracks on the topside to those removed chips! top pads just ripped off! I bet the old "screwdriver under the chips and a big push" trick was used to get them out! the ones that had never been out which I removed were fine! no track damage! as you say Bob more introduced faults to repair!

Anyways this is how the board looks tonight on the top so far, more still to do and a lot of track metering to do too! but that's for another night......

IMG_2766.jpg.8b3ae2e06334888f0ba6ae5af323a8fd.jpg

IMG_2768.jpg.5d2955368cd3e8079b15e70b9ee39ac7.jpg

  • Like 1

Share this post


Link to post
Share on other sites
3 hours ago, Andrew96_ said:

 

extra hole drilled close to track! circled in yellow

 

IMG_2765a.JPG.e6a348832955f298966d390bd2a2ad83.JPG

 

and in yellow the extra resistor fitted because the resistor network right next to the battery had one rotted/failed resistor!

IMG_2752a.JPG.a52458cb856e22aa852bd978b55959f3.JPG

 

yes these are the botches I could do without!!

 

 

Both 6K CPU cards I fixed a while ago also had the extra resistor next to D1. Maybe a factory fix as they didn't look like any work had been done over the years.

 

Look again at the resistor... it's going to ram pin 8 which has nothing to do with the network resistor 😉

Share this post


Link to post
Share on other sites

hmmm, I have 4 boards here in total and none of the others have this fitted!

this shows where the resistor comes through the board and it is attached to a ram pin ad the pad is definitely lifted and turned! not straight! it also had that horrid foul solder and flux rework on it! so to me it did not look original one bit! At this point I had not looked at the diagram as there is too many induced track breaks to sort out first! the resistor network is definitely knackered!

IMG_2755a.JPG.d7e3bd8ec5e500146fd4eadf76fb9133.JPG

Thanks for the heads up Louie, looking at the diagram I can see it IS fitted on later boards so was a 'add on' but due to the way it has been replaced / fitted it really was not done well making it look bad! Looking at the diagram it is to pull the ram chip selects high to the ram supply voltage so as to stop the ram being written to and corrupted during powerdown. so YES a legit mod but it could have been done better!

230231430_sys80R23.thumb.jpg.d9cabd71fd3efbe85e5dfe001b0a0d06.jpg

 

as you say it has nothing to do with the bad resistor network which really would have been good to have been changed on its previous repair.

 

Share this post


Link to post
Share on other sites

Just imagen if you had to fix a few thousand boards on the JPM production line back in the early 1980's and given a few minutes to do it. Drill hole... solder in resistor... next. Don’t think it would be very inspiring if you were doing it all day.

As for the other fixes things could be better, but that’s part of the learning process? Maybe this is the person’s first ever attempt at fixing a board. It's not exactly easy to find someone to help.

Maybe the person had a go at fixing it, which should be encouraged as there are not enough people rolling up their sleeves to try their luck.

Please showcase your best work here and be mindful those who may have spent hours giving it their best shot may be reading rather negative commentary.

  • Like 2

Share this post


Link to post
Share on other sites

threads would be boring if they was all the same, that's why when I get something different I post it

I know the person that repaired this board charged and was one that had a go at me for taking money for repairs, but will leave that subject there.

this board lasted less than 2 years since repair.

 

this is just about how I rectify the problems I find with boards, just the same as I have always done. can I not even do that now?

 

you could always spend time doing your own threads on repairs as that would also help the community too wouldn't it?

Edited by Andrew96_

Share this post


Link to post
Share on other sites

ok so repair is coming along, not really more pictures of it just yet, but thought I would bring up the subject of counterfeit ram I have found in these sys 80 boards!

so here is apparently the same make of ram on 2 sets of chips...

IMG_2782.thumb.jpg.a4ef53ba32e9bb4c9c9e0ceb192624b2.jpg

left set is counterfeit! things to look out for...

 top surface has been sanded! original is shiny! there are still sanding marks left in the top surface!!!

counterfeit is laser etched, original is solvent proof white paint type screen printed markings, font..... well totally different! not even close!!! no explanation needed here!!

date code, original is 85, either 3rd week or third month of 1985, that would put the ram about the right date, counterfeit is 2011! 29th week!! they were never produced in 2011 so far far wrong!!!!

so does being counterfeit matter i her you all ask!!! most would say as long as it works it is ok! but counterfeit is rarely fit to work at the correct speed it was designed for, indeed with all this remarking to cover up the chips true identity you don't even know what chip it was in the first place!! so can it be trusted??

this is one of the OLD chips on the left, old but it is good!

IMG_2780.thumb.jpg.8d73a21a4bd135bdfb08ae612a73439a.jpg

you can hopefully see it is one of the old ones in the socket, and it passes just fine!

now compare this to the counterfeit chip

IMG_2778.thumb.jpg.bf7dd33ed104f2d8ad9f560233bd2042.jpg

 

oh dear! data pins stuck high!! not a good sign! clearly bad!

so would you now trust ram chips off ebay? and it is getting worse!!

 

just make sure any sellers you are thinking of buying from will send you pictures of the actual chips they will send! the giveaway is most will say they don't have the time to do this, also ask to see the manufactures bill of sale to prove they are original, again if counterfeit you won't even get  a answer!

 

just take care as it is a right faff if you buy non working counterfeits, it will be on you to prove they are bad to ebay protection which is now getting more difficult! s just don't buy them in the first place!

hope this helps people buying ram in the future!

 

Share this post


Link to post
Share on other sites

The term 'Caveat emptor' has never been more appropriate!

Share this post


Link to post
Share on other sites

well the counterfeit ones must have worked for a couple of years before failing, I am sure it is the refurb process they go through that makes them fail, due to the timescale I would think they was subject to static charge during the refurb which has lead to junction damage  within the chip and it's very reduced life!

Share this post


Link to post
Share on other sites
6 hours ago, Andrew96_ said:

well the counterfeit ones must have worked for a couple of years before failing, I am sure it is the refurb process they go through that makes them fail, due to the timescale I would think they was subject to static charge during the refurb which has lead to junction damage  within the chip and it's very reduced life!

Didnt realise this andrew .always thought counters wouldnt work at all or at best until they warm up !im sure there was a batch of 9980 that would work until warmed up after say 30 mins then fail .they came from ut source i think .but intersting stuff .

Share this post


Link to post
Share on other sites

yes that's why protecting against ESD is actually a big thing in electronics, old cmos or mos technology is much more susceptible to static damage especially as little as 200v (or less) which is easily accomplished charge by the human body just moving about! so old cpu like the 9980 and old ram can easily be damaged and you would not even know its happened as at that voltage you would not feel or hear a spark!.

I don't think people realize in general how or why they should do anything to protect against ESD when handling electronic chips / boards etc!

this makes interesting reading on ESD even if you only read the first bit!

https://www.electronics-notes.com/articles/constructional_techniques/electrostatic-discharge/esd-effects-how-affects-electronics.php

Share this post


Link to post
Share on other sites

scroll down to the section on 'Latent ESD damage' as well, this I think is what the refurb process does to chips from what I have read

Share this post


Link to post
Share on other sites

Thanks andrew a interesting read .glad i dont use plastic cups !

Share this post


Link to post
Share on other sites

OK so back to the board! been slowly putting it back together!, solder resist done and components mounted back, good ram from 1985 tested and good put in, luckily the 9980 cpu was good as that was a ami type that had been replaced by the previous repairer and was fine (as you now know the ram was not so good) board now looks like this

IMG_2785.jpg.f41a58614d00173cc5b3507fc50af975.jpg

ok not really looking much different on the topside! just all the broken topside tracks done! where the top pads had been pulled off have been left if there was no traces connected to them as the underside was fine to solder the chip in. the resistor pack changed along with the diodes near the battery and a few resistors and a decoupling capacitor next to one of the chips as it had the top of it damaged over the years and one 74ls138 ic

underside

IMG_2787.jpg.4e18315cd5cdfef9db3fdd2df23690c2.jpg

Pleased with that! the wire coming through the battery hole has been left like that at the moment as have been waiting for the bigger size pcb eyelets (rivets) to arrive which I am still waiting for! but this will be tidied up soon!!

the 22k resistor on closer inspection DOES look like it was put on the board from new as did have slightly green on the legs from battery rot! so apologies to whoever repaired this last time! I have put it back but drilled a  hole slightly over so it is now away from touching any tracks, does the job!!

IMG_2789.jpg.b55784a4579842c49b8eb1a3ac3558e0.jpg

and underside, 

IMG_2788.jpg.b5abe3e7bcd9ea5581fb9ea8236cb605.jpg

C1 had been changed in the crystal oscillator section and a 0.1uf (100nf) capacitor had been put in, I have now changed it back to the correct 0.01uf (10nf) capacitor, also I had a bit of lacing string lying about so thought it a good idea to tie down the crystal as it does seem very vulnerable perched on top of 3 resistors!

IMG_2790.jpg.368f0c5726b7cdb937f2c3277058c3d8.jpg

 

hopefully that will help protect it!

Edited by Andrew96_
  • Like 1

Share this post


Link to post
Share on other sites

over the past couple of weeks I have seen the posts on making a sys80 ram adapter which got me thinking.......so now I have a working board I  thought I would try my plan out! managed to find the first mpu3 ram board I made for testing,

IMG_2791.jpg.0773711a135cbf8b5b9618e7881d0278.jpg

this will be my base board, the 2 chips it replaces are side by side, sys 80 is end to end ram chips! 

 

so off came the pins and turned pins put in place of the left side which will plug into one of the ram sockets, then after a bit of scrambling around found the SMD ram chips and soldered one on. then to make this universal for different pin spacing I put a small length of cable soldered to round pin into the place where this would plug into the other ram chip on MPU3, so it looks like this...

IMG_2793.jpg.c089a9245062db87a24194afd625d0a5.jpg

 

next plugged into the ram sockets on the sys 80 board

 

IMG_2795.jpg.cdd63a10b20e8b840b0b8137a407fe3c.jpg

 

switched on and it works perfectly!! Brilliant I now have a known working and good sys 80 test ram adapter!!

 

Edited by Andrew96_
  • Like 1

Share this post


Link to post
Share on other sites

awesome stuff :) 

 

well done .. 

Share this post


Link to post
Share on other sites

That makes for a universal conversion both Tec + maybe more. Can see how the PCB can be shrunk or made single sided, which will drive down cost so good for everyone.

The CPU card now looks good. Looked an easier gig than the one I was working in the same timeframe. Could post up some pictures for a critical eye if you like?

Share this post


Link to post
Share on other sites

yes sure post away louie!

when I first made those mpu3 ram boards I was going to try exactly that on sys80 boards then design a adapter to fit sys80, but nothing happened towards it until last week as I had never had a working sys80 board with the smaller type of ram and of course no machine to try it on, also back then ram was easier to come by so it really was not worth the bother, but I am glad it works and least I now have some known good test ram as I don't have any of the type of ram used in these lying around!

Share this post


Link to post
Share on other sites

when I did have the MPU3 ram board I got them down to £12 each, but the chips they replaced was around £5 a chip at that time. but that was 6 years ago now! how times fly and how counterfeit chips have risen in price and common place!

Share this post


Link to post
Share on other sites

just looked on ebay for the 5514 ram,EVERY listing on there looks fake ram to me!

look at these!! no proper date code and quite clearly have sanded tops!! not even blacktoped these days!

what has the world come to when they fake ram chips so blatantly!!

https://www.ebay.co.uk/itm/5PCS-TC5514P-TOSHIBA-IC-SRAM-1KX4-CMOS-DIP-18PIN-PLASTIC/322738465285?epid=1918439840#

and these... well easily spotted sanded top, incorrect markings, and the chip ident marking is far to big, that's not even a Toshiba chip!

https://www.ebay.co.uk/itm/INTEGRATED-CIRCUIT-TC5514P-1024-word-x-4-STATIC-RAM-CMOS-PLASTIC-TOSHIBA/183490879568

 

how blatantly fake can you get!!

Edited by Andrew96_

Share this post


Link to post
Share on other sites

×