Author Topic: Dreaded acid corrosion - Some success?????  (Read 1771 times)

0 Members and 10 Guests are viewing this topic.

Offline Retropin

  • Trade Count: (+4)
  • ******
  • Forum Posts:
Re: Dreaded acid corrosion - Some success?????
« Reply #30 on: December 19, 2010, 10:10:50 PM »
Well seeing as some of you are following this thread, ill continue the updates until ive got this sucker working again.
Im STILL working on the corrosion damage. Things i thought i had fixed are now problematic.. all sorts of cranky stuff is happening mainly due to the crappy makeup of these boards. When i replaced traces with single strand wire, some were very difficult to determine just where they went as they tend to disappear under IC's. I thought i was being smart by soldering directly to the trace and alleviating following the trace to its next logic point. Problem is that the traces really suck and have lifted where ive soldered. They are so thin that vibration tends to snap them clean off or create hairline cracks.
problem can be seen here circled in red... i did once have a wire soldered to this trace but noticed that it just didnt look right... it had snapped right at the end of the solder cutting its data path short. To find where this trace went to i had to remove the 4040 chip JUST to see if it went straight under it or deviated slightly... LOL - it went straight!




I also didnt like the look of the CPU 2650 socket... i had originally used the CLOSED type of socket here.. mainly cos im a tight arse and they are cheaper at Jaycar. When Mike "Homepin"came round he pointed me to the wire wound strip type as connections ABOVE the board can be easily tested as well as BELOW. I decided to change this socket and im glad i did... found ANOTHER floating pad.
The CPU tests SOLID now... i have all data stobes strobing as they should... im now 100% happy that this area is good


Ive also REtested the cct where the two RAMs interact.. this little area is essential to the CPU booting... Im happy that this is OK, but in all honesty there is so much damage up here and it gets so complicated that its hard to say im 100% happy with it... i have found a couple of mistakes here



I got to the point where 7 out of the 9 test points were good... this really is good progress and it does indicate that im taking steady steps towards getting this sucker running and its VERY staisfying to see good pulses on the CRO where they are meant to be

1 test point that is not working is POWER FAILURE... im not fussed about this as i can simulate power good by forcing a high onto pin3 at voltage input connector CN9 by shorting pins 3 & 4.
My main concern is test point 6.. OUT COUNTER.

This is a complicated little cct driven by 2 JK Flip Flop IC's 74ls156 and 4040... seemed to me that lack of signals are my main faults on this board, so i pulled the sockets and tested the traces... yet again.... found another couple of faults... these are impossible to detect without sanding back the board to reveal the copper traces... very fine lines can be seen on the solder pads... put your meter across these lines and you reveal an open cct.

CCT has been repaired and quality sockets installed

[
Unfortunately, this did not fix my cct.
 Some years back i did electrical engineering at TAFE in my spare time.. all we used in Digital Applications were NAND gates and bloody JK flip flops. Its been quite a while since ive worked with these and im finding it hard to get my head round... i once used to be able to read a truth table for these, but i look at it now and it might as well be in Swahili. Im going to need to pull my notes out and found out whats wrong... i SHOULD be getting a HIGH from pin 11 ( 74ls156) but im not... this is a Q output whos state is determined by active clock and inputs..... Its at this point i start looking for little ceramic caps etc... but no luck here.. just a pull up resistor.

So... where am I? Well its all good actually... just the one test point to sort out and im pretty certain i can get this board to boot...she aint too pretty though.
front looks like this



Back looks like this



I did start off doing the back with nice little twists in the wire trying to keep it neat, but i ended up with so much on the board that i gave up with these as it just added to the confusion. I had lots of little twists and it was difficult to see where a particular wire had its beginning and end. There is also no code to the colours chosen... i just got bored with yellow and moved to red etc.

Anyone with good knowledge of JK flip flops feel free to chime in... would save me a nights swotting from my old notes