Jump to content

Repair log for an MPU4 with a hidden short .


MAGIK
 Share

Recommended Posts

Check this out.  This might go to help if you have a MPU that wont boot and you cant find the issue.  Sometimes you can find hidden shorts on these things...

 

So I thought I would put a little repair log on here for you lads.

MPU4 MOD2, fails to boot, owner cant find the issue.

So at first I coudn't find the issue either.  No dead pins on CPU or RAM, sometimes CPU starting sometimes not, not stuck in reset, no interrupts stuck low.  I was a little surprised at the IRQs stuck HI at first because that crystal card I am using is a total git (they are multilayered things and the pads fail) and more often or not it locks up with the IRQs low. 

So a little investigation started, first thing was to find out if there was a read /write or enable issue.  Ram voltage was fine, a little low at 4.6v ,  Output enable , pin 20 on the 6116 all pulsing away nicely, but pin 18 was stuck hi.  Weird thinks me because the 138 at IC16 is pulsing away, and as far as I could see I had a good working address bus (well I thought I had).

Then I smelled something burning.  And then I saw something burning.  C37 became a roman candle.  Switched off and checked nothing shorted on my "everything" power supply . No apparent cause for the blowup.  The cap had blown open , schemys show its just a bypass cap at the 5v line input so just left it alone and powered back up.

Nothing else went bang so continued on.

Checking up at the 138 a little further I noticed that pin 1 never went low. Which was odd because it was address line A8 it was tied to.  Checked at the CPU, noticed it was always HI as well. Pulled the game card, it was still HI.   Didnt think I would get a buzz, but checked for a short between pin 1 at the 138 at 5v. BEEP!

Cut it and short was still there.   Next target looked like the 541 tied to those address lines from the CPU, so removed that, short was still there.  A8 on the CPU socket was dead shorted to 5v. How?

It didnt make a great deal of sense because all A8 tied to directly was that 541.  And that was no longer there.  Its when I turned the board over,  and used the zoom on the phone ( the cameras on these things make the digital cameras of old look like fischer price toys!!) that I spotted the teeniest of breaks in the trace pushed over onto the track for 5v. You can see on the video how its happened the two traces run next to each other and its very easy to short them if the solder mask is worn or flaky. So fixed the short , replaced the 138 and 541, reinstalled working CPU and RAM and fired up.

Nothing else smelling bad, nothing else smoking,  after fiddling with the PAL chip as usual on the program card to get good contact IRQs started running and I had boot.  But no alpha.  It was definetly working and display should have been up because IC3 and 8 were being selected and it was trying to do something with the reel outputs. 

Checking the pins for the alpha display on IC3 and 8 (cant remember the numbers off hand) Showed that everything was stuck low.  followed the outputs down to the display pins, and saw the three zener diodes on the clock, data and display lines looked burnt. Cut them out, and pin 39 on IC3 ( alpha data out) started running. So replaced the diodes, and voila!

Got "CASH METER FAULT" flashing away at me.  Easy bypass , clip lead across the top and bottom leg of T11 . Then a quick bit of trickery with a clip lead on the reel photo pins and it was in attract.

Hope you liked the video, got more repair logs coming up in the near future, I decided Id put them up here, and although I havent got all too technical on this one, I will try to be more accurate with pin and IC numbers and locations next time.

Hope you all had a good christmas and new year!

  • Like 6
  • Thanks 1
Link to comment
Share on other sites

 Share

×
×
  • Create New...