Two Sick SE/30 Logic Boards

[CC_333]

Another Update:

I redid my work with C8 using smaller wire, checked everything with a continuity tester, and it still isn't working.

Would UC6, UE6, UG6, UI6, UC7, UE7 and UG7 have anything to do with video? If so, could a failure in any of those cause the symptoms I'm experiencing?

Or maybe there are other possibilities?

If anything, this is a good learning experience.

c

[phreakout]

Can you verify with us that C6 and C8 are connected right? C8 (-) goes to Pin #6 of J12, and C6 (+) connects with Pin #7 of UB11. UB10 and UB11 are the two sound chips on that board.

Position the board so that the notches on the top of those chips are facing to the right. The leg/contact on the top right-hand side of UB11 and UB10 is Pin #1, next to that leg (moving to the left) is Pin #2, next to that leg (moving to the left) is Pin #3, and so on. The last pin on the top left-hand side is Pin #8, then move down to the bottom row starting with the bottom left-hand side. The first pin on the left is #9, next to that is Pin #10, next to that is Pin #11, and so on. All chips, whether they are square or rectangle, surface mounted or not, have their legs/contacts/pins numbered counter-clockwise.

Regarding UC6, UE6, UG6, UI6, UC7, UE7 and UG7, if I recall correctly, those are part of the video circuitry. There are other smaller chips that sit in-between that group, C2 and C7, that might be part of that circuit as well. The problem is that there are at least 3 different revisions of the SE/30 logic boards in its nearly 2 year existence and each may be slightly changed. One revision is in regard to boards made in Ireland; those were in fact built more reliable through using some more stable parts and better quality control.

73s de Phreakout.

[CC_333]

Hi, phreakout,

C6 and C8 are verified. I checked them time and time again with a continuity tester, probing each capacitor and the appropriate pin(s) on the components they're supposed to connect to. The tester beeped, which I assume means that there's good contact among all the parts in question. I doubt the capacitors are the problem at this point.

As for the UC6, UE6, etc., I tried reseating them with no change. Could a failed one of these cause this problem? Is there a way to test them?

Also, how do I tell which revision this board is? It has a socketed CPU, if that means anything.

I feel that we're getting somewhat closer to the solution (I think?).

c

[phreakout]

The boards made in Ireland are lighter in green color, have a white colored PDS slot and the Bourns network filters (RP2, RP3 and RP10) are white or thin black-white colored covers as well. Plus, obviously, there is a sticker near the ROM SIMM that says "Assembled in Ireland" plain as day.

The remaining 2 other board revisions are the darker color green, have a black colored PDS slot and the Bourns network filters (RP2, RP3 and RP10) are either dark tan/beige or brown colored. I think these boards were either assembled in Mexico, USA, China, Korea, maybe even Thailand, Philippines or Japan; it's hard to say.

Those video chips can be a real pain to extract, at least from my experience. Part of the problem I'm encountering are my soldering irons and the solder Apple used on the original parts. That solder was formulated for low temperature soldering and has silicon or not enough flux in the mixture; it turns grainy and cloudy after hardening and aging. I could really use a soldering iron tip designed to heat up and desolder those type of chips, without any damage, and fits 20 to 40 watt Weller and Radio Shack irons. After I get the chips removed and the salvaged replacements ready to go, I'll be installing chip sockets FIRST on the board, then plug the chips in their sockets. I can't stress it any further how much of a pet peeve it is for me to have chips without their sockets!!

73s de Phreakout.

[CC_333]

Well,

As I was reseating the socketed ones, I noticed that one of them didn't mach the rest (UE6), it's thicker looking and the markings are different from the others.

I swapped it with the equivalent on the other dead board without change. By the way, the other dead board has the same exact problem.

What are the odds that both could've failed with the same problem?

Also, I actually have one of those Made in Ireland boards, except it needs to be bent a certain way in order to work (the symptoms are similar, except it's completely dead; no sound, no video). I assume it's a bad trace somewhere.

c

[Mk.558]

I'm working on Rev 2.0 of the capacitor diagram right now.

Note to the iBook: It's not a big deal if I spend 45 minutes trying to set up a nice drawing of a Sony Stereo Sound Chip and I go press Delete and you somehow confuse that with logging out. Of course there is a autosave function in AppleWorks but apparently that was on vacation. Bullocks -- this time we redraw the whole thing scaled up 3x (Yes I measured the sound chip and it's 6x19mm) and put nice "SONY 03E03G ..." on it.

It'll be better this time because apparently I forgot to include the half circle indentation on the last drawing, so the reference to Pin 1 was unobtainable because there was no reference mark. That's okay I look at it now and it looks all fugly anyways, polluting JDW's nice socketed motherboard image with terrible graphics. Steve Jobs would have told me off in front of a crowd for that.

[CC_333]

OK. I await your Rev. 2.0!

Meanwhile, I've given up on the one SE/30 board (the one I've been talking about) for now, and am instead concentrating my efforts on this apparently made in Ireland board that will only work if I bend it a certain way. I've concluded that it's either a bad trace somewhere or a bad solder joint somewhere (I noticed that it gets somewhat better when it's warm, so it could be heat/age related).

I would like to get these boards working, but I'm not sure if I'm competent enough at this point to do it myself (although I'm learning fast!)

Until next time...

c

[CC_333]

Does anybody want to tackle one of these boards themselves? Let me know and we can go from there.

Also, I appreciate all the wonderful help I've gotten here.

Thank you very much!

c

[techknight]

I can take a look at them. I have an oscilloscope and working machine to make comparison charts with.

[CC_333]

Great!

That would be nice. I have two (hence the topic name), but I will deal with the one which I tried to rework (and therefore should technically function properly, if not for this latent and as of yet unknown problem).

I guess I'll mail it??

It'll figure itself out, I suppose...

c

[techknight]

I would send them both. What you may think is one way, but could be different than the boards actual condition. I am in the states, and send me a PM.

Reason I say this, if it comes down to being a rare or impossible chip to replace thats actually bad, it may be a different chip on the opposite board, and i could make 1 board work from 2.

Then again, i could fix both of them. Never know.

[CC_333]

OK. I will send both, then.

I'll also send you a PM (private message?) with more information.

c

[techknight]

yup , Plus I can swap out all the HAL/PAL/GAL ICs from your board, to my working one and rule those out.

[CC_333]

OK. That sounds good!

c

[Mk.558]

Ta Da!



It's a US Legal size image (try a Super A4 ISO size at 22.7x35.6cm if you're outside the US) , so it should print out fine if you have that capability. It was designed originally to fit on 8.5x11 US Letter but there simply wasn't enough room (and that was before I added the Procedure part, and the Polarities section). Then I tried A4, and nope. The Legal size is just barely enough to contain practically everything you need. I did it all with AppleWorks 6.2.9...I really should move on and discover proper image and drawing applications

Notice: All functions of the capacitors are not set in stone. Still waiting on some expert advice based on detailed circuit analysis. I have a copy of the BOMARC and Apple diagrams if anybody wants to tackle that (shouldn't take long).

[CC_333]

Thank you, Mk.558!

c

[techknight]

Yes, i would like to see those just to take a peek.

[phreakout]

Mk,

In the words of Peter Griffin, "That's phreakin' sweet!" Brilliant!

73s de Phreakout.

[techknight]

I had received the boards today.

I realize you were trying to learn, and i cannot knock on anyone for trying and learning. However constructive criticism says this:

Need to be a lot less sloppy with the soldering iron. You have done some extreme damage to the boards. Im sure i can fix it, but the thing that bothers me are the vias. its like you drilled out the vias? you cannot do this. ever.... you will break connections between layers of the PCB. and possibly short layer-to-layer. Trust me, i know....

These are 4 layer boards, and traces run from layer to layer. Each traces is connected between layers if routed between layers, through a Via. If the via is broken/drilled/etc you can damage the the inter-layer connection. Anyway, I have schematics so its not difficult for me to fix. I just have to hope layers arnt shorted, or i will have to cut traces, and route around it to clear the shorts.

Also on one board, I noticed you have a bare copper wire run from C8 all the way around to the bottom of the board. I am not going to explain the reasons why this is a bad idea

On the flip side though, it seems besides the ones with broken pads, you got the capacitors soldered in place perfectly fine. No problems there, good job

So... What I am going to do, is fix these boards, I will document everything that I do, on my youtube channel, etc... Therefore, I recommend you and everyone who has a modern computer, watch youtube, and use this as a learning experience in case the same thing happens to anyone else.

Stuff happens, no one can predict it. All we can do as human beings is learn from it. So we shall

I am heading up to parts express tomorrow because I have to restock on my caps so I can do the other 2 boards. I use radials. I dont use tantalum because I dont like the spontaneous shorting that could happen when they age. When tantalum caps get old, they start to become noisy when DC leakage starts to creep in. Then sometimes they will just short. Which in high current circuits result in a BANG.

P.S. I am sorry, I am trying to not be offensive to anyone. I am simply stating my findings, and mention ways to work around it and learn from it, as we have to fix this before I can begin further troubleshooting of which IC could have taken a hit.

[JDW]

Wow. Techknight, that sounds so bad that I simply MUST see a photo! I think we can keep it anonymous so the original owner won't get embarrassed or offended. But if my board photos can serve as an example of what to do, it would also be good to have photos of "what NOT to do!"

[techknight]

Well like i said, I am trying not to offend the rightful owner. I am trying to be constructive about this and not hurt anyone's feelings. He tried, and he actually did pretty good on the install of the new caps. its just that some errors were made that shouldnt have been.

[JDW]

Well, then take a photo "with the owner's consent" of course. My think is that a photo showing "what not to do" might help others avoid the same mistakes in the future. It seems like such is needed in light of the fact many of these folks have seen my "BEFORE/AFTER photos" but still are making these kind of mistakes.

[techknight]

Ya, ill take photos. But it will be up to the owner whether he wants them shared in the public. BTW, check your PM.

[CC_333]

Hi, all,

Well, all I can say is I'm learning.

This is a very important learning experience.

Hopefully, techknight can help me out of my hole, and I can then proceed to learn more so these mistakes don't happen again.

Thank you all for the invaluable information and wonderful help.

c

[techknight]

Thats why i will video/photo document my proceedures. this way, you and everyone can learn from it, and when the next capacitor replacement comes on a new conquest, You can perform it with confidence

Oh, also, I want to teach you all this tidbit:

Routing through a via is often easier than taking a route around the entire board. Ratshack sells 30 guage wire, which is fine wire. This wire is small enough when stripped, will fit through an existing via to reconnect a broken via trace segment.

Also flux and tin the via with fresh solder BEFORE soldering anything to it, in it, or on it.