Macintosh Portable Recapping/Repair: Progress & Questions

[Dark Goob]

In the next post, I'll post photos of my progress with the recapping itself, since I'd like to know what the experts think I could improve upon in terms of my soldering.

Original condition prior to repairs, but after battery replacement:

• would only boot from a floppy or external drive
  
• would not start up or run from a battery, only from an external power supply (24W version from PowerBook 165C)
  
• speaker will not work
  
• battery will not charge

Repairs made so far:

• replaced innards of original battery with a PowerSonic PSG-650 (it fits perfectly inside the shell of the old battery)
  
• removed all radial through-hole caps and replaced with ones from Radio Shack (fatter and with higher voltage ratings/lower temperature ratings, but still fit OK)
  
• removed all aluminum electrolytic caps and replaced with tantalums ordered through trag (Jeff Walther)
  
• replaced the hard drive with a similar model (Dylan swapped the PCB from the original dead drive to a working other drive from a desktop Mac, and it worked... thanks Dylan!!)
  
• cleaned motherboard as much as possible with alcohol & swabs, also CyberClean putty (not sure if OK to use, but it did pick up a lot of crud from the board and didn't seem to damage it!)
  
• removed broken diodes from positive wire of 9V battery (see issues below)

Current condition after the repairs:

• after capacitor replacement, speaker works again!
  
• still won't boot up or run from a fresh battery, which tests at 6.45V (I may have botched part of the recapping, see next post)
  
• starts up from internal drive with the (7.5V @ 3A) PowerBook 165c adapter
  
• after booting, in the Battery desk accessory the indicator is about 1/6th of the way full (just past the initial little line indicating reserve power)

Question related to the diodes: I noticed that the positive wire leading from the battery-door switch to the positive terminal of the 9V backup battery had no continuity, so I cut off the black plastic that was wrapped around one section of the wire. Underneath that plastic I found two 1N5817 20V @ 1A diodes in series, but the connection between them was broken. I couldn't find 1N5817 diodes at Radio Shack so I temporarily picked up a couple of 1N4001 50V @ 1A diodes. I am guessing that the two 1N5817 diodes are simply there to prevent damage to the circuits in case the 9V battery is installed backwards -- is this correct? I see that the voltage drop from the two 1N4001's is higher, so if anything, it shouldn't damage the computer to have those on there instead of the 1N5817's, and they would offer more protection. But can anyone tell me if the 1N4001's are an acceptable replacement? I can easily order 1N5817's from Digi-Key but if the 1N4001's are just as good, then I'll just run with those.

Question related to the battery wiring and switch: I am a bit puzzled by the wiring for battery power connections in this machine.

• a thin black wire goes from the negative terminal of the 9V backup battery, then merges with a thick black wire that connects to the negative terminal of the 6V main battery; then both wires run down to the four-pin motherboard power plug (the thin one goes on the left and the thick one goes on the right, but they both test as being at the same voltage in all situations)
  
• a thin red positive wire runs from the 9V battery's positive terminal through two diodes and into the first terminal on the three-terminal switch that is activated by the placement of the battery door; meanwhile a thick red wire runs from the positive terminal of the 6V main battery into the second terminal on the same switch; lastly the third terminal on the switch feeds the thick/thin red wire pair that runs down to the four-pin motherboard power plug
  
• when the battery door is in place, the voltage across any red-black wire pair at the four-pin motherboard plug is 6.45V (same voltage as main battery)
  
• when the battery door is removed, the voltage across any red-black wire pair at the four-pin motherboard plug is 8.45V (same voltage as 9V battery minus the voltage drop from the new diodes I put in)

Now, I have read that you must have a fresh 9V battery installed for the machine to boot. However, given that the 9V battery's positive terminal is not connected when the battery door is in place, then I fail to understand why the 9V battery being in place would make any difference as to whether or not the machine will boot! It seems to me that the ONLY function the 9V battery could possibly have is to simply act as a backup so the motherboard still gets some juice when you remove the main internal battery. Otherwise it's not even connected. Thoughts?

-=DG=-

PS -- I will post the photos later today

 

[mcdermd]

I thought it might be you. I was going to email you back and suggest you ping the MLA here but you're one step ahead of me.

Dark Goob is another Oregon user.

 

[Dark Goob]

Photos of the (probably crappy) soldering job are below.

Right next to C11 there appears to be some possible damage to a trace. The darker green area actually extends pretty far down, so I don't think it happened while I was working on the board (possible damage from previous owner, the machine has apparently been dropped once or twice based on cracks and broken parts in the outer case of the computer). Either way, If indeed that is a damaged trace, what can be done about it?

These are the two tiny 10µF guys:

These are the ones along the bottom edge of the board:

These are the small 1µF guys next to the reset buttons:

These are the small 1µF caps up by the rear ports of the machine:

If you look next to the voltage regulator on the bottom right-hand corner of the following image, there looks to be some crud coming out of it. Might these need replacement as well?

-=DG=-

 

[Mk.558]

You could run a small wire from the source of the broken trace to the destination.

Now I did not exercise excellent soldering skills on my tantalums inside the SE/30 right now, but yours look overheated. Almost burnt. xx(

It is not a crime to use electrolytic replacements instead of tantalums :approve: in fact on my spare SE/30 logic board I skipped the tants and had a MUCH easier time soldering. (Aside from the lifted pad on C8)

 

[mcdermd]

Are you using a nice, fine tip on your pencil? I'm wondering if you've got it on there too long, too hot with a tip that's too big? If you want to come down here one of these days, I can see about helping you clean it up and give some advice.

I've always skipped tantalums and replaced my aluminum electrolytic cans with like.

 

[shred]

The 1N5817 is a schottky diode. It has a much lower forward voltage drop than the 1N4001s that you are using, although it may not matter much for the backup battery. I don't know why there would be two diodes unless something was originally connected to the junction between the two diodes. In any case, as you've found: the 9v battery is disconnected when the battery door is in place. I have no idea where the mis-information about needing a good 9v battery came from, but I've seen it reproduced on a number of web sites.

Some of the solder joints look "cold": they have a frosted appearance. This often happens if the component was inadvertently moved while the solder was cooling. I'd re-solder those joints: you are looking for a nice shiny soldered joint. You should heat the component and board trace with the soldering iron, then apply the solder. If you try to melt the solder on the iron, then blob it in to place, it doesn't work too well.

The damaged trace on the board may be the result of an electrolytic capacitor spewing its corrosive guts onto the board. You should be able to very carefully scrape back the green solder mask and then solder a small wire over the damaged track.

 

[Dark Goob]

{snip}

Some of the solder joints look "cold": they have a frosted appearance. This often happens if the component was inadvertently moved while the solder was cooling. I'd re-solder those joints: you are looking for a nice shiny soldered joint. You should heat the component and board trace with the soldering iron, then apply the solder. If you try to melt the solder on the iron, then blob it in to place, it doesn't work too well.

The damaged trace on the board may be the result of an electrolytic capacitor spewing its corrosive guts onto the board. You should be able to very carefully scrape back the green solder mask and then solder a small wire over the damaged track.

OK. What about the "toasted" looking caps, do you think that I should make sure they're not damaged? Or are tantalums pretty durable? They shouldn't have gotten any hotter than 320C max when I applied them,nbut I have no qualms about replacing them if that is bad for them.

As for the damaged trace, what about using a trace pen? I am a bit scared to try soldering a wire on a trace that small. This picture makes it look huge but it's super tiny.

BTW, anyone know of a relatively inexpensive capacitance meter that can go up to at least 500μF? My meter won't test past 10μF.

 

[Dark Goob]

Ok I made the solder connections alot better and it was still not booting from the batt. Then finally I checked the motherboard fuse and... no continuity. There's the culprit! Bridged it with a 2A automotive fuse and presto! Booted right off the battery. Wish I would've checked that fuse first!! Oh well, I'm just happy she's up and running again.

Battery charging terminals now show a full 7.5V, so I'm going to let the battery charge for a few hours. According to http://www.power-sonic.com/images/powersonic/sla_batteries/ps_psg_series/6volt/PSG-650_11_Jan_12.pdf the battery needs to be around 7.2V to be fully charged. At 6.45V the Portable thinks it's near-empty.

So I'll let her charge overnight and see how it is tomorrow. But... very encouraging :-D

-=DG=-

 

[mcdermd]

 

[trag]

Some of the solder joints look "cold": they have a frosted appearance. This often happens if the component was inadvertently moved while the solder was cooling. I'd re-solder those joints: you are looking for a nice shiny soldered joint. You should heat the component and board trace with the soldering iron, then apply the solder. If you try to melt the solder on the iron, then blob it in to place, it doesn't work too well.

I'm too late to this party, but I agree with what Shred wrote. While the problem turned out to be a fuse, the photos show 'cold' solder joint. Either the solder was not hot enough to form good joint when the parts were soldered, or it was heated too much and started to oxidize. Apply liquid flux to the joints and resolder with a soldering pencil.

Ah, I see that you've already fixed that. Excellent. With cold solder joints, there is a chance that the components aren't making electrical contact at all.

There's no reason to use a hot air gun when soldering the parts back into place. In fact, for little two terminal parts like these, I wouldn't use a hot air gun at all (for removal either). That heats up a lot more of the board than is necessary, can damage surrounding components, and doesn't do the board a lot of good. There's a limit on how many times the components should be heated up to soldering temperature.

Excellent that you got it working.

 

[Dark Goob]

Thanks for your help, trag (sold me the tantalums I needed).

I did take the advice on the solder joints and added more fresh solder, they look bulbous and shiny now.

The hot air gun seemed easier since I didn't have another solder pencil small enough to get in there with dual tips to tweeze the caps off. However, in retrospect, I would just get the dual pencils because indeed, the hot air gun can easily melt stuff. I've read posts elsewhere that recommend insulating any sensitive parts of the board with aluminum foil, etc. But I am not sure how one ought to attach such insulation securely to the board -- and t seems like more trouble than tweezing with dual pencils.