SE30: C9 capacitor (positive pole), i've lost its trace.

[Macbuk]

Hi everyone.

I've a problem.

Reccaping my SE30 i encountered a problem with C9 capacitor.

Desoldering some capacitors (3) i broke their traces. But i've addressed the problem in any situation except the bastard C9.

I've lost its "positive" trace and i've some concerns about its origin.

Could anyone tell me where to find schematics?

Could anyone take a read with a multimeter to check resistivity between C9 positive pole and ground? there are only two choices in my opinion.

I'm suspecting two nearest points where it could start but i would be sure before trying to turn on. Even a multimeter read could be useful for me.

Regards.

 

[wally]

C9, 47 uf/16V is the bypass to ground for -12V coming in on J12-7. So the positive side goes directly to a ground feedthru exactly between C9 and C8 and nearly on the centerline of C9 and C8. On the bottom side artwork you will see it hit the corner of two perpendicular traces connecting to ground mesh artwork that surrounds and connects to C48. Ohmmeter reads to ground only the resistance of the test leads.

 

[Macbuk]

Greeeeeeeeeat.

The best explanation i've ever read.

I've drilled a little hole along the feedthru to reach those perpendicular traces connecting to ground mesh you were referring to, i've uncovered them and i soldered each arm to better secure the connection. Now multimeter says ground.

http://img141.imageshack.us/img141/563/motherboarddrilledmg6.jpg

http://img120.imageshack.us/img120/472/motherboarddrilled0mf9.jpg

Wally,i'm afraid but i would be glad to know if i can reach C5 and C6 positive trace destination on a chip or other component: C4 and C5 positive trace unglued from the motherboard (still connected to new capacitors) but i would like to reach their destination with a bridge to better secure the connection. Or if it isn't possibile what's the best way to secure unglued traces? Glue? silicon? cyanacrilate? paraffine?

Tomorrow when i'll try i hope it doesn't explode! [] ]'>

 

[wally]

Drilling a solder filled feedthru can cause the inner plane connection(s) to be lost if the drill removes the feedthru barrel. So if you have really tiny drills, accurately centered, ok. Better to suck it out with a antistatic solder sucker, solder braid, or melt it and punch thru with a stainless steel wire that the solder will not wet. However in this case I think there are other grounds and you will be ok. That feedthru probably went to the inner ground plane and probably did not ground any inner layer traces.

C5 positve bypasses two sound chip voltage reference pins to ground. So the positive end of C5 connects to UB10-15 and UB11-15. Once the pad is gone or even lifted, that really skinny trace is the devil to connect to even with the finest copper wire (flexible power cord is a source for fine bare wire).

C6 positive goes to UB11-7, a sound chip pinout that needs a 1 uf bypass to ground.

C4 positive goes to UB11-10. This is the right channel sound output on its way thru the blocking capacitor C4 , series resistor R12, EMC supressor L2 out to the phone jack J1.

I have had some success resecuring lifted pads with 5 minute quick setting epoxy. But when you hit a freshly glued pad with the hot iron, who knows what the chemicals emitted will do to you, so have some good cross ventilation going. Soldering first then sticking down the pad might be better if the arrangement of things allows this. Clean the board and pad surfaces with denatured alcohol before gluing. If the pad is corroded away you can make a new pad from copper foil or even a smashed down bare wire, epoxy it in place and then you have a somewhat secure place to anchor the replacement part. But when at solder temperature, both the epoxy glue and the epoxy in the PC board is practically liquid under the pad, so a gentle touch with the iron is necessary. No harm in applying a bit more epoxy after the soldering is complete if necessary.

 

[JDW]

I would suggest you hot-glue those replacement caps as soon as possible. From the look of your photo, the least bit of pressure on those caps could snap one or both of the traces off the logic board. In my own case, I used hot-glue to ensure that even if I bumped a cap while handling the board it would not snap the traces.

 

[Macbuk]

Thanks Wally, i've had to uncover and lift C5/C4 remaining traces (hair like diameter) and i had to solder them to capacitor. Annoying but not so difficult...

I checked:

-UB 10-15 and UB 11-15 grounded thru C5+,

-UB 11-10 has continuity with C4+

I've some concerns if it's better to make new wire bridges bypassing these original unglued thin traces or leave them as are and glue. I guess it would be better to bypass but i could do it in the future if any intervention will be necessary. I think epoxy the best choice, better than hot glue but who knows. That being said, I hope it will start, without simasi pattern and SOUNDING! [] ]'>

 

[wally]

Epoxy might be too strong, and hard to remove later should you have to rework anything in the bonded area. Good for pad repairs, maybe too good for parts support! Imagine trying to break an epoxy bond loose years from now and peeling off some traces underneath in the process...

One approach is to get it to work first, with VERY careful handling of the unsupported parts. When it works, you can consider refinements such as securing stuff better, or more compact self supporting caps such as tantalum radial leaded or SMT parts.

 

[Macbuk]

Unfortunately something has gone wrong. It doesn't start at all, nor the monitor neither the power supply. I guess something on the main connector (12, 5 or -12) is shorted to ground leading powersupply to autoprotect. With the main connector disengaged PS starts; instead when i connect the mobo hd and fan don't start. Indeed isn't a big loss for me cause it was firmly stuck on simasi pattern: During latter times it hadn't been sounding, then overall conditions suddenly worsened, so before this recapping attempt i had no much hope.

That being said i would like to exclude evident and stupid mistakes with C8 and C9 positive and negative and their relation with main connector.

I suspect one of them the problem, i must have done an error. C9 positive seems well grounded but i'm unsure about the negative lead.

i 've also some concerns about C8 traces where do they connect? I wouldn't appear boring so if you could suggest a place to download schematics i would try a basic troubleshooting before i sell it for parts.

 

[register]

With the main connector disengaged PS starts; instead when I connect the mobo hd and fan don't start.

Please do not power up the switching power supply without any load connected to it's output. SPSs of older make sometimes toast themselves when powered without load, causing some "Sparks and Noxious Odours".

 

[tomlee59]

It's altogether too easy to short a supply to ground when doing the sorts of operations you've been doing. You may want to do a quick ohmmeter check to see if that happened somewhere along the way. Check +5 to ground; +12 to ground; and +5 to +12. I don't know if the SE/30 has -12V, but if so, check continuity between it and the others.

To avoid forward-biasing junctions that can confound the measurements, I recommend using the ohmmeter in "ordinary" mode (i.e., not "diode-check" mode).

 

[wally]

You might now have a parts unit, and a reference board at least for the non corroded areas.

register is right about the SE/30 power supply having a minimum load specification, but hard to say if any damage resulted from this one instance however.

Some might get the impression that replacing the capacitors is some kind of magic fix. The reality is that all the damage caused by capacitor fluid leakage must be located and reversed also, and some of this is shorts under connector bodies and ICs, as well as pads, feedthrus, and surface traces that have corroded away. The best candidates for cap jobs are those that show at most wetness but no corrosion of pads and traces. To do a good repair on a seriously corroded board, it's best to have a reference board, not necessarily working, so you can see where the traces once were that need to be put back. Even after doing all of this, there has been at least one case where the same screen symptoms were due to bad video ram, not caps.

C8 (and solder side ceramic C48) are bypass caps to ground for -5V from J12-6.

It is possible you made no wiring mistakes and one of the replacement caps is shorted. Temporarily lifting one leg of each might allow you to isolate it. Powering up with no caps is okay for shorts elimination purposes, but do not expect the computer logic to run consistently. But judging from the look of the C8, C9 area photo, there appears to be a possibly corroded area that might contain shorts. Where there is surface corrosion, the area between conductors needs to be mechanically scraped clean down to bare board to eliminate shorts that can be difficult to see even with a stereomicroscope. Toothbrushing and alcohol wipes are not enough because the short is plated thin film metal and other insoluble metal compounds. J12 itself might have corrosion underneath the nylon housing, look carefully with magnification from the sides.

+12V J12-14

+5V J12-12,13

-12V J12-7

-5V J12-6

GND J12-1,2,3,4,5,8

Video J12-9

Hsync J12-10

Vsync J12-11

 

[Macbuk]

Thanks everyone for support.

Wally your Se30 knowledge is amazing!

I've checked all capacitors four times: their bypass grounding seems fine.

Moreover i checked and rechecked those lines whose rework wasn't properly clean:

J12-6 -C8+ ground line is fine

J12-7 -C9+ ground line is fine

UB10-15/UB11-15 +C5- ground line is fine

UB11-10 +C4- ground line is fine

UB11-7 +C6- ground line is fine

Furthermore i checked with naked eye C1, C10, C12, C13, C7: their soldering was clean and troubleless. When i have some time i'll check better with a lens those areas inheriting audio and main connector.

Tomorrow i'll test resistance reading between voltage pins and ground and if necessary i'll check for shorted caps. I work as MD and this stubborn behaviour remembers me a sort of terminal patient overtreatment. :-(

I recapped succesfully 1 macintosh II, two nubus video boards: this was the first time i've met aluminum SMD capacitors with spot glued base. Moreover they seem more prone to leak than standard aluminum axial like those used on non SMD era. Anyway ...i'll let you know

Thanks

Greetings.

 

[Macbuk]

Thanks you for help even by mail! [] ]'> That's a nice community.

J12-12 and J12-13 +5v lines are shorted to ground: it could be everywhere along the board because C11 bypass seems fine.

I'll try to remove C9 capacitor wire lead placed along the feedthru i drilled: i could have met a sort of 5V line or mesh on intermediate layers even if the hole was centered. Nevertheless short could be everywhere over the surface and its really difficult to find: +5 is widespread along the board.

The search is discouraging, requires time, acute sight and nevertheless a lot of patience! Perhaps the board demonstrates multiple signs of corrosion. I'll try removing C9 but I guess it's better to find another MB. I've a curiosity, are mac classic Moboards SE compatible?

 

[gobabushka]

no, you cant plug a mac classic board into an se.

 

[wally]

...J12-12 and J12-13 +5v lines are shorted to ground: it could be everywhere along the board because C11 bypass seems fine...

Check also C1, C7, C12, C13 which are also +5V bypasses to ground that you probably recapped (47uf/16V). There are also a ton of .1 and .01 SMT caps on the underside that bypass +5v, however these rarely short unless mechanically disrupted.

The recapping process involves a lot of board handling and soldering which can scatter bits of solder, broken pad and trace copper, capacitor lead clippings and such into bad places, like between IC leads. When power is applied the shorts can weld or solder themselves in place and be resistant to being dislodged by compressed air, vacuum, or shaking. On the underside there are lots of metal stubs that can get bent over, possibly shorting to the ground grid artwork that surrounds a lot of them.

Also, take a close look at the sheetmetal deck and card guide metal for metal whiskering, a kind of velvet fuzz growing out of the metal surface consisting of pure metal whiskers that can break off, drop onto the MB below and short out stuff. I had one SE/30 chassis that was growing this stuff. I cleaned it off with steel wool but never really trusted it again and eventually it became a parts unit.

If you decide to blow or vacuum the PC board be aware that the nozzle can develop a static charge from the airstream friction. If you hold the board ground and the nozzle near the tip with your two hands the static will pass harmlessly thru you rather than jumping from the tip to burn out the components.

 

[Macbuk]

Check also C1, C7, C12, C13 which are also +5V bypasses to ground that you probably recapped (47uf/16V). There are also a ton of .1 and .01 SMT caps on the underside that bypass +5v, however these rarely short unless mechanically disrupted.

Wally you'r so precise ! This evening i tested all capacitors you suggested: it seems that C13 had been ground shorted. There was a lifted and broken plate so when i tried to connect C13 to the remaining "underlevel" ground point i guess i've intercepted a +5 line on the upper surface. I rerouted C13 ground lead to C11, the nearest ground point known and i obtained 233 ohm reading between J12-12,13 and ground. No more ground short.I still have some concerns about the continuity of this 5V line, we'll see.

http://img61.imageshack.us/img61/4899/627lh1.jpg

Then i tested some lines i thought fine

UB10-15/UB11-15 +C5- ground line is fineUB11-10 +C4- ground line is fine

UB11-7 +C6- ground line is fine

In particular UB10-15/UB11-15 +C5 line, whose reading wasn't properly 0 ohm but something between 1 and 2 ohm. The trace was lifted (not broken) but i guess neither this nor cold joint was the problem. May be it was corroded somewhere increasing resistance value: this leaded me to rework these audio circuits with some bypasses.

http://img89.imageshack.us/my.php?image=633zd6.jpg

Now overall +5 line to ground resistance has misteriously increased to 290 ohm.

http://img207.imageshack.us/img207/2342/637nr5.jpg

Tomorrow or the day after tomorrow when i have some time i'll try if it starts (i've Se30 remaining parts in my old house) and i'll let you know.

I'm really excited cause i'm waiting my fourth apple IIgs (SCSI card equipped) deliver from germany. Vintage apple computing is taking a lot of time in my life, i would avoid not to become haunted. [] ]'>

 

[JDW]

Macbuk, first of all, congratulations on making the time to do things the right way. There is often a lot of "talk" in the online forums, but its refreshing to see the "walk" -- especially with nice clear photo, as you have so kindly provided.

If I can suggest one thing more it would be this. I see you placed your logic board on a carpeted surface. Beware of static because it will destroy chips, especially RAM. And even if you don't directly place your board on a rug, scooting your feet across the floor will generate enough static charge to zip something to oblivion. If you must work on a carpeted surface, I strongly suggest a ground wire around one of your wrists. Indeed, such a ground would be good when working even on an anti-static floor.

 

[register]

A damp towel might do as a cheap antistatic mat. It also catches dropped parts, preventing them from vanishing.

 

[tomlee59]

I prefer something less aqueous. Aluminum foil works well. It's best to remove the battery, though, or otherwise prevent any possibility of short-circuiting its terminals.