Search the Community
Showing results for tags 'Tantalum'.
Found 2 results
Subject: Power Macintosh 7300 stops working after tantalum recap Hi Everyone I decided to replace the electrolytic caps on my favourite Power Macintosh 7300 computer and unfortunately it hasn’t turned on since! I’m writing for help on what to check or do, as I exhausted all checks that I can think of. The problem is power supply fan spins up, power LED on but nothing else - no video output, no chime or sad Mac sounds etc. I did make a mistake that has never happened before but was due to the multi platform nature of this logic board - I added two extra capacitors and subsequently removed them after one power on. I was able to see my mistake when I did a ‘cap count’ for the top layer of the board, there should be 9 47uF 16v caps but I had 11. I removed all the top side 47uFs with tantalums, and the two unpopulated spots fit and looked like ones I had removed. Troubleshooting steps that made no change - Swap the existing Sonnet Crescendo card for a stock Apple CPU card from my PM7600. - Reset CUDA by pressing with button with computer both on and off - Disconnecting power and PRAM battery for 10 mins and powering on - Replacing PRAM battery - Triple checking continuity of one repair I had to made on the left side of the board. The capacitor pad lifted at C189 so I toned out where the positive lead went to and added a kynar patch. I later compared the continuity of my patch to the other working PM7600 I have and it matches - so I am confident it’s got to stay. I did reverse my patch just in case and no change - Remove and reseat all chips such as VRAM, RAM, Cache - Went down to one VRAM and one RAM chip, then back up to two. - Double check of C56 labelled values and measured capacitance. On my Atlas meter the original cap measured 56uf with 10.0ESR (high). The silk screen of the cap is 47uf at 4v, not the typical 16v. This is the first time I’ve come across a 47uF 4v so I replaced it with a 47uf 16v. - Check for shorted capacitors by putting black lead on ground then testing the positive and negative leads of all capacitors I installed - twice! No shorts to ground due to soldering… - Spraying a fine mist of isopropyl alcohol on chips to see if anything was heating up quickly. The P74BA DS8925Ms heated up the quickest but still were not hot to touch. Heat is a great tool in finding shorted chips but I couldn’t detect anything. - No parts went up in smoke or were cracked after I turned it on. - I inspected the possible ROM chip on the rear side of the board and it’s perfectly clean / good traces. - I know it’s appropriate to use a tantalum on the cache card becuase my PM7600 has one from factory. The 7300 in question was electrolytic from factory. Interesting find that I verified against a PM 7600: Two capacitors are installed reversed from the factory. Initially I thought I had made mistake as the continuity test showed the positive was going to ground the bottom side of the board C392 and top side of the board C103. The board silk screening shows the + symbol and my tantalums match the board AND the polarity of the original caps. I have no answer why Apple reversed the caps (meaning negative is where the + symbol is labelled). I have not reversed my caps, they match the originals and the PM7600 for comparison. Board condition: No rotted traces, no battery leak, inspected with magnifying lens and no corrosion found anywhere. Working before tantalums put in. Soldering: I am experienced and have done probably 50 Macintosh boards thus far and this is my first time having a working board stop after recap. Continuity checks to vias against my patch done. No excessive flux or low temp solder running under caps. Conclusion questions: Attached to this post or linked are photos from my past two days of getting nowhere! Could it have been the two extra caps I added at the beginning that killed the board? OR Could it be the very low ESR of the replacement caps that changing the boards boot up? Considering there is no reference for caps online, I have typed the values into my spreadsheet and made a screenshot so the formatting is retained. Jeremy