Jump to content

Tales from Midibob's workshop. MPU3


midibob
 Share

Recommended Posts

1 hour ago, midibob said:

The nightmares continue...

Just finished a board that after the usual corroded components had been swapped out showed up a lamp fault with some lines not working. From cold though they were all working, I also noted that once the lamps stopped so did half a bank of switches. No amount of freezer anywhere or preheating the board made any difference to the time the fault took to show? With the switch test running on the diags the bad bank would work until a couple of minutes when if a switch was operated the board would continually bleep as if the switch was being operated on and off. Shortly after it would just stop and the whole bank was then dead. The other bank worked just fine??

Eventually I noticed that IC4 (6821) was losing an o/p which feeds to IC11 (74LS138) which controls the mux pulses for the lamps and switches. From cold it was there but then it would go intermittent and then stop. No amount of freezer on IC4 made any difference, nor did heating it up before switch on. So, in with a replacement and bingo all switches remain operational. Ready to wrap up when I noticed there was still one line of lamps out.:headache:

IC11 had a bad o/p. Once changed all was good.

Phew!!

A fantastic write up bob, it’s always interesting reading about your progress with repairing these old boards,,  😄

Link to comment
Share on other sites

18 minutes ago, Andrew96_ said:

bet that freezer spray cost you a small fortune as cans of the spray are not cheap these days!!

Yep, every time I use it I hear a 'kerching' in the background!

  • Like 1
Link to comment
Share on other sites

  • 2 weeks later...
Posted (edited)

Yet another board from hell arrived on the workbench! From initial looks anyone would think it's not too bad, just the usual stuff to swap out and off we go. How wrong was I!

005-1.jpeg.e7fbc78a76059a8c09689f567dc07aa3.jpeg

The list of faults won't be in chronological order as there were so many I've lost track. I think it originally was stuck in reset but as all the transistors and other components were going to be swapped out in the reset circuit no in depth testing was done. There were loads of crumbly solder joints and getting some of the old bits out was a challenge but no pads lifted in the process which was good. So with all new bits in and looking shiny underneath it was time for a power up.....and....nothing!

Not surprising really but things went downhill from here onwards, especially my sanity!

On switch on you can get a rough idea if it's going to boot or even initialise by the click it makes and the reels judder. In this case there was nothing and the first step is always to check the RAM, of course it was knackered. In with my test RAM daughterboard and ..... nothing!

Next one I always go for is the 555 ocillator as if this is out of limits then it won't run. Having said that with the diags it should show a triac output whether the osc is running high/low or not at all but in this case there was nothing. With the scope it measure around 450Hz which is far too low. All the associated components were swapped to no avail so a new 555 was put in and we had a good 500Hz. Still no boot though.

So we now move on to the support chips and see which one is stopping it from booting. I won't bore you with the details here but once the 6840 (IC2) the three 6821's (IC 3,4&5) were swapped it sprang into life.  However things weren't right and the reels were intermittent as were the lamps (there were more problems here) and the triacs.
What was happening was the connectors were covered with that brown corroded (patina) and weren't making good, if any, contact. No amount of switch cleaner made them any better so the best course of action was to get them all swapped out. Luckily I had a donor board to hand and this is the only way to get these sockets now especially with the right colours. So with 'new' clean sockets things were better but still not right.

Moving the ROM cart caused it to freeze and despite cleaning just tapping the cart caused it to freeze. On of the connectors was also broken which you might be able to see but I don't think that was an issue. Anyway back to the donor board and exchange the cart socket. Now flexing the cart it was solid but a tap and it would freeze!! Hmm thinks, I'll come back to that one.

Attention was now turned to the lamps. The new socket had brought a few more lamps into play but there still a couple of lines playing up and 2 BDX34's were found to be duff.

There were also switch issues which were caused by IC22 (CD4049).

OK, so it all appears to be running and testing fine, except for one thing, after about 10 mins on the test rig it freezes. Tapping around the cart and CPU also causes it to freeze? Hmm not removed the CPU yet so out it comes and back in with a socket and.... still the same.:angry:
By this time I'm thinking of jumping in the Thames again!

New day, new ideas. As the CPU is now in a socket it's easy to put a new one in. Well, what do you know solid as a rock. Back in with the old one and either leave it for about 10 mins or lightly tap it and it freezes. The pins are all clean so it must be intermittent inside. You couldn't make it up.

So almost time to wrap up until I checked out the AUX port PIO. I always run software with an alpha as if that runs OK then the chip is usually good. Well guess what the alpha just came up with a load of rubbish! In with yet another 6821 and thats all my spares gone.  It all looks like it's working so now it's off to the Cabaret machine for a soak test.

Half the 7 segs aren't working!! The diags don't check these which is a pain. This was due to IC16 (ULN2803).

At last it's a goer and passed the soak test with flying colours.

Edited by midibob
  • Like 3
Link to comment
Share on other sites

  • 6 months later...

It's been a while since I wrote an MPU3 fault up. Just when you think you've seen it all yet another one comes along to bite you in the bum?

This one came in as not working/untested and after the usual suspects had been swapped out I noticed that IC23 (74LS08) was burnt! Swapping this out and then powering up on the test rig I was pleased to see the reels moving and the lamps flashing. The triacs were all pulsing correctly too. So moving on to the switch test and ....nothing??

Now IC23 is fed from IC11 (74LS138) and here there were no outputs because there were no inputs! Hmm what's going on here? The inputs to this chip come from 3 of the 6821's IC3,4 & 5 and there was nothing leaving on all three on pin 39. A quick resistance check to ground showed 2.1 Ohms 1k Ohms and open circuit (should be approx 5K) so although all 3 PIO's were working in their other functions all the pin 39's had popped!! What a bummer.

IC21 (9602) is also connected to pin 1 of IC11 and this had also popped.

With all parts replaced things were back to normal thank goodness.

  • Like 7
  • Thanks 1
Link to comment
Share on other sites

  • 2 weeks later...

Some more boards giving me the run around or beyond economic repair.

First one was a real pain, it wouldn't initialise and for the life of me I couldn't see why? Someone had been here before which is always a worry but they work seemed of a good standard.

So the nitty gritty....on switch on there was just a short pulse to the RAM and then a few address lines died and it just sat there. IC9 (74LS138) and IC24 (74LS12) had been changed so I knew all was good in that dept and all the other usual stuff had been done. The first suspect was the CPU which was swapped but no difference. Next one was the 6840 as I've seen this stop boards booting before, but again nope! Not wanting to continue the wild goose chase I thought I'd try and think this one out.

Next step was to tap out all the data and address lines to make sure there were no shorts or open circuits, which there weren't. Then it was the turn of all the other functions eg E, NMI, VMA etc. When I got to the R/W I found there was no connection from the RAM (pin 10) to the R/W line feeding the rest of the board? What a relief I'd actually found something.

On the previous repair the two RAM sockets had been changed but because they were using the RAM daughterboard they had to be the stamped pin type as you can see in the pic. On removal of the socket in IC8 I found this...

P1040638.thumb.JPG.7b78be0bbc9a6ff9c2455efbd1520bc2.JPG

The via at pin 10 of IC8 had been ripped apart and consequently lost connection. So a quick rivet insert, a tack wire and a bit of solder mask to finish off and she's as good new.

P1040639.thumb.JPG.3001c097fa20b383f81c7efa30a1c7c1.JPG

When all put back together it booted and ran fine.:D

....and the next one.

This had already been deemed beyond economic repair as you will see from the pic....

007-1.jpeg.05db4d51ae99a23e97cce881f47cf5b7.jpeg

The board was toasted right through just above the power diodes and various components further up the board varied in incinerated, well done to medium rare depending whether they caught the flame or not.:lol:

As I was head scratching yet another board I thought, why not, and stuck it on the bench. First thing is to give it a clean and try and remove as much carbon as possible. Once done this left a sizeable hole in the board. This has to be done though as any carbon left will just track and start the whole process over again.

P1040640.thumb.JPG.b8eb3bfb552f073ccc1d989e183e8009.JPG

I must admit it does look scarey but I've done this quite a few times now so it's pretty much business as usual. The next step is to make up some resin, seal the bottom with cellotape and then fill the hole.

P1040641.thumb.JPG.1c953d5a9c728602917101ea15a84a86.JPG

Once this has set the track needs to be re-made, holes drilled, rivets fitted and solder mask applied.

P1040642.thumb.JPG.efc525066a9b2e58c8601fb9aadd5e93.JPG

New track cut and fitted and riveted. They do look close together but it's a trick of the light! I still need to drill and fit another rivet for the white power lead but that's tomorrows job. Then a bit of solder mask to finish off.

P1040643.thumb.JPG.aa7ca7e0cb7e1f40a7422298042e4bbf.JPG

All that's left is to swap out all the burnt components and the usual bits and pieces. I just hope after all this it doesn't have any obscure faults!!:headache:

  • Like 5
Link to comment
Share on other sites

32 minutes ago, midibob said:

Some more boards giving me the run around or beyond economic repair.

First one was a real pain, it wouldn't initialise and for the life of me I couldn't see why? Someone had been here before which is always a worry but they work seemed of a good standard.

So the nitty gritty....on switch on there was just a short pulse to the RAM and then a few address lines died and it just sat there. IC9 (74LS138) and IC24 (74LS12) had been changed so I knew all was good in that dept and all the other usual stuff had been done. The first suspect was the CPU which was swapped but no difference. Next one was the 6840 as I've seen this stop boards booting before, but again nope! Not wanting to continue the wild goose chase I thought I'd try and think this one out.

Next step was to tap out all the data and address lines to make sure there were no shorts or open circuits, which there weren't. Then it was the turn of all the other functions eg E, NMI, VMA etc. When I got to the R/W I found there was no connection from the RAM (pin 10) to the R/W line feeding the rest of the board? What a relief I'd actually found something.

On the previous repair the two RAM sockets had been changed but because they were using the RAM daughterboard they had to be the stamped pin type as you can see in the pic. On removal of the socket in IC8 I found this...

P1040638.thumb.JPG.7b78be0bbc9a6ff9c2455efbd1520bc2.JPG

The via at pin 10 of IC8 had been ripped apart and consequently lost connection. So a quick rivet insert, a tack wire and a bit of solder mask to finish off and she's as good new.

P1040639.thumb.JPG.3001c097fa20b383f81c7efa30a1c7c1.JPG

When all put back together it booted and ran fine.:D

....and the next one.

This had already been deemed beyond economic repair as you will see from the pic....

007-1.jpeg.05db4d51ae99a23e97cce881f47cf5b7.jpeg

The board was toasted right through just above the power diodes and various components further up the board varied in incinerated, well done to medium rare depending whether they caught the flame or not.:lol:

As I was head scratching yet another board I thought, why not, and stuck it on the bench. First thing is to give it a clean and try and remove as much carbon as possible. Once done this left a sizeable hole in the board. This has to be done though as any carbon left will just track and start the whole process over again.

P1040640.thumb.JPG.b8eb3bfb552f073ccc1d989e183e8009.JPG

I must admit it does look scarey but I've done this quite a few times now so it's pretty much business as usual. The next step is to make up some resin, seal the bottom with cellotape and then fill the hole.

P1040641.thumb.JPG.1c953d5a9c728602917101ea15a84a86.JPG

Once this has set the track needs to be re-made, holes drilled, rivets fitted and solder mask applied.

P1040642.thumb.JPG.efc525066a9b2e58c8601fb9aadd5e93.JPG

New track cut and fitted and riveted. They do look close together but it's a trick of the light! I still need to drill and fit another rivet for the white power lead but that's tomorrows job. Then a bit of solder mask to finish off.

P1040643.thumb.JPG.aa7ca7e0cb7e1f40a7422298042e4bbf.JPG

All that's left is to swap out all the burnt components and the usual bits and pieces. I just hope after all this it doesn't have any obscure faults!!:headache:

Wow sorry bob didn’t think they’d be so much trouble  

thank you you genius.

Link to comment
Share on other sites

Omg, that burnt board looks amazing, what a fantastic job, is there any Mpu3 that you can’t repair Bob? If that one fires up, that will be brilliant 👌👌

Link to comment
Share on other sites

Hi Ronnie, I used an Isopon body repair kit. If the hole is large enough I chop up some glass fibre strands to give it more strength. The only thing you need to be careful of is the temperature. It can go soft with heat so if you need to drill rivets through it with soldering a bit of care is needed. Not had any problems it has to be said.

  • Like 1
Link to comment
Share on other sites

20 hours ago, MikeB said:

Omg, that burnt board looks amazing, what a fantastic job, is there any Mpu3 that you can’t repair Bob? If that one fires up, that will be brilliant 👌👌

Fired up first time and ran without a hitch. Don't you just love it when that happens.:D

  • Like 4
Link to comment
Share on other sites

Haha,just catching up on Bob's jobs.

You are saviour to the cause Bob.

Well done mate on everything you do in the name

of saving the boards to keep the old girls going.

Legend🤗

 

PS...stay away from the Thames😂😂😂

Edited by Line Up Chris
Link to comment
Share on other sites

 Share

×
×
  • Create New...