You're thinking about this backwards. If three different symptoms occur but there's one underlying problem, the cause is related to all of them. So it's not a waste of time to be diagnosing any of them: so start with the symptom that is most regular.
-
Hello MLAers! We've re-enabled auto-approval for accounts. If you are still waiting on account approval, please check this thread for more information.
Macintosh SE/30 logicboard recreation (thread revival)
- Thread starter Bolle
- Start date
You're thinking about this backwards. If three different symptoms occur but there's one underlying problem, the cause is related to all of them. So it's not a waste of time to be diagnosing any of them: so start with the symptom that is most regular.
I’m not sure if I understand your question correctly - but the behaviour is such that the machine can run for a while and not show any corruption on those columns. At some point (sometimes hours, but more often minutes after powering on) the little horizontal lines start to generate in those columns, gradually building up whenever there is movement on-screen and usually following the mouse pointer.
Mouser, but it’s moot. One of my boards uses all original donor parts.
I used pick and place for the backside passives. Any chance these could have been incorrectly selected? Is it worth desoldering and checking all values?
The first reloaded board doesn’t have, but in case it’s relevant the donor parts are a mixture of components from two different SE/30 boards.
The other two reloaded boards have new F253/LS30/LS166/LS393 etc. All larger chips - the DIPs and PLCCs - are 100% original.
Wait, are you saying it doesn't do it for hours? Are you running them outside the case?
Does it do it if you point a fan at it?
It’s rare that the fault takes hours to show, it usually happens faster than that. But it does vary. There is also a difference in terms of behaviour between my three builds - one of them appears to exhibit the symptoms more frequently and more severely than the other two.
I’ve run the boards properly fitted inside of a case and also outside.
Will try the fan.
Out of interest (and wanting to know more about how such things are fabricated), why do you say that?
Because they are a professional business that make boards, all day, every day. It is not in their interest to get things wrong, and if they were, they'd have already invested money and time to sort it. These boards are trivial to make compared to some of the stuff they'd be asked to make.
We all can't understand why things aren't working sometimes, but it is almost never the parts we use, but decisions, misunderstandings or mistakes we have made along the way. You're talking like the three boards are three independent cases, but there are multiple common factors - some components, some parts selection, and the methods you've used. You might have cooked a chip on all three, or solderbridged a via on all three, or used an LS where you should have used an F series logic chip. Did you make an error on the first and copy the first board for the other two? Have you used the same PSU on all three?
Its a probability thing - has a company that has a good reputation made an error on three boards in a row, or, have you? Not meaning to be mean, but unless you are a professional batch assembly person and these are exceptionally wild circumstances, you're really looking to work out where you went wrong.
Really sorry, it sounds like I'm questioning your ability, but I'm not - I'm just saying that manufacturers are on a different level to us hobbyists. My stuff doesn't work all the time, but I've been in that situation enough to say while technically possible that a manufacturer could have messed up, it has always being something I have done.
The board itself is created using lithography and is tested automatically using probes to ensure that there aren't unexpected shorts. Components are placed by a pick and place machine (a robot) - components are lined up on big reels with labels on them. Cameras are used to inspect the PCB and fitted components to ensure that they visually look like they should and aren't in the wrong place. A human inspection is also done.
There are videos showing the Seeed Studio and JLCPCB factories on youtube... let me see...
Where I have selected new ICs, they are the following. Do these all seem OK?
UB12, UC12: SN75175DE4
UD1: SN74F240DW
UA9: TL071CDR
UF8, UG8: SN74LS393DR
UE8: SN74LS166ADR
UJ6: SN74LS30DR
RP9: 4816P-T02-221LF
RP1, RP4, RP5, RP6: 4816P-T01-220LF
RP7, RP8: 4816P-T02-102LF
UI2, UI3, UI4, UJ2, UJ3, UJ4: SN74F258D
UA8, UB8, UC8, UD8: 595-SN74F253DR
RP2, RP3, RP10: 4120R-601-250/201L
UB12, UC12: SN75175DE4
UD1: SN74F240DW
UA9: TL071CDR
UF8, UG8: SN74LS393DR
UE8: SN74LS166ADR
UJ6: SN74LS30DR
RP9: 4816P-T02-221LF
RP1, RP4, RP5, RP6: 4816P-T01-220LF
RP7, RP8: 4816P-T02-102LF
UI2, UI3, UI4, UJ2, UJ3, UJ4: SN74F258D
UA8, UB8, UC8, UD8: 595-SN74F253DR
RP2, RP3, RP10: 4120R-601-250/201L
Much appreciated all your input so far today guys. Even just a gentle nudge in the right direction is a huge boost.
For the most part, the letters at the end can be ignored because they usually relate to the package (i.e. DIP) and packing (paper tape etc.).
So what you're concentrating on is are the letters in the middle the same as the original parts - "LS", "F" etc. Similarly, if any parts have a speed grade.
Thousands of 64 pin sockets available here:
https://www.connectorpeople.com/Connector/TYCO-AMP-TE_CONNECTIVITY/8/822019-2
Drop it in a tray of white vinegar for a couple of hours and see what it comes out like.
Ok. Thanks. It was scrubbed with white vinegar but I'll let it soak.
@cchiro I'm curious as to how you get on with this. I have three battery-bombed ROM SIMMs, all different physical versions, that don't work despite my best efforts to save them. I initially thought they were just simple two-layer boards but it turns out that at least two of them are four-layer. I'll probably end up designing replacement PCBs and then discovering that the chips themselves are toast...
@CircuitBored - LOL - Sounds like Murphy’s law. I’ll keep you posted on my progress.
I’ve found out that if I make contact with pin 6 of UE7, the issue clears up immediately.
What I mean is that I’m poking at it with a thin pointy metal file.
This pin is labelled ‘Nubus’ and connects to pin 33 of the Glue chip, and pin 82 of the PDS slot. I can make contact at any of these locations and it fixes the problem for as long as I hold down on it.
With this file wedged here, I don’t even need to touch it for the issues to go away. With a smaller metal item like a paperclip, I do need to touch it.
What is touching it doing to the circuit - am I adding a resistive path to ground?
Other thoughts?
Similar threads
