Suspect Bad PS Apple AcBel API 1PC36

[Toasty]

Re: Suspect Bad PS Apple AcBel API 1PC36

That's UCC - the little shield emblem.

LTec links to PDF files and you need the Adobe Reader plugin installed for your browser.
Attached PDF's for your convenience.

C43 is an LTec TK.
C13, 35, 41, 59 are UCC KY.

Toast

[buz]

Re: Suspect Bad PS Apple AcBel API 1PC36

Well you're a champ Toasty .... thank you! Some fish giving going on on one hand but feel like I'm working for it on the other.

Seems an answer always produces questions. I've 'finished' as best I can. The L-Tec data sheets confused me, they're laid out differently than the Teapo and UCC data sheets. So I'm not real confident of the values I came up with for the L-Tec's.

Then, when I was getting the UCC caps done I noticed the different values depending on a temperature/impedance value. I believe all my values came from the 20ºC column - so either they're all wrong or 'mostly' right.

Thank you for the series correction on C43.

[buz]

Re: Suspect Bad PS Apple AcBel API 1PC36

So if my ripple and Z values are correct, then do I match those values when shopping for replacements? (Still am concerned w the L-Tec ripple and Z numbers)

Which brands are most respected?

I've seen here (in a mobo thread) that using a replacement with a higher V rating can be ok, is it true for PS's?

I've also read that larger diameter caps are ok as long as there is room physically??

[Toasty]

Re: Suspect Bad PS Apple AcBel API 1PC36

All that info is in here.......

Z = impedance = ESR as I said before (post 19).

Equal or better with ESR and Ripple.
ESR - lower is better: 0.02mΩ is better than 0.04mΩ
Ripple - higher is better: 1500mA is better than 1200mA
Voltage - one "step" up is considered acceptable: 2.5v->4v->6.3v->10v->16v->25v->35v->50v->63v
Sometimes the step up gives you better ratings too.

Diameter is the hardest to fit in a lot of units. You've plenty of height clearance here, so that should not be an issue.

Most replacements you need are available in the sizes you already have. One may need to nudge a coil a bit to fit a 12.5 in where a 10 was (C21).

Caps that would be good for this supply.
Look at each ones datasheets and figure what will work for you here. Obtaining them all from one source will save costs for shipping.

Nichicon - HE, PM, PW
UCC - KY, LXY
Panasonic - FC

Then you need to find who carries which and are in stock.

www.digikey.com
www.mouser.com

Always tick the "stocked" or "in stock" boxes when searching those sites.

Happy Hunting!

Toast

[buz]

Re: Suspect Bad PS Apple AcBel API 1PC36

Before I do any ordering ... wonder if you'd take a quick look at my selections. You didn't list the Panasonic FM series but it was listed elsewhere in the forum ... thinking there could be a reason you didn't list it.

I see that C21 is hard to come by in a 10mm and agree a 12.5 could be 'wedged' in. Also, C10 is drawn on the board much bigger than 10mm - any reason to upsize it?

Thanks for all your help & (patience)!!!

EDIT: From my spreadsheet it looks like I may as well order all Panasonic if I can.

[Toasty]

Re: Suspect Bad PS Apple AcBel API 1PC36

FM's are okay, but you run the risk of overkill on a PSU when it comes to ESR. Ripple handling IMO can never be too much, but too low an ESR can toss the balance off on the filtering as designed.

School's still out on that, and I've not had a bad experience with that either. But, there is caution in the rumblings about the forums regarding others who ran into a higher ripple when using "ultra-grade" caps.

That's why I did not give the FM's as a choice. FC is lower than the HE in one line, and equal to the PW in the other (you'll notice that they closely follow each other in the sheets {makes one wonder... } ). In both cases, sticking with the KY originals is likely a better choice.

Toast

[EDIT: C10 - whatever is at hand from the manufacturer is what goes. If 10mm are cheaper this week... ]

[Toasty]

Re: Suspect Bad PS Apple AcBel API 1PC36

Also, very nice job on the charting and cap worksheet. Good to see you're getting the hang of it.

T

[buz]

Re: Suspect Bad PS Apple AcBel API 1PC36

Getting the hang of it requires a good 'guide' - Seriously!!

I see now that because I find an option on the data sheets, doesn't mean Digi-key or Mouser will offer it. I've gone through both vendors and made a list of replacements. The new lists are a 'cry' from those I had listed from the data sheets but similar in stats. . Anyway, I don't know whether I need a 'rubber stamp' or not ... as in, "oh those will work" or "oops, better swap out ..... ". It appears to me that perhaps those from Digi-key may be the better of the two.

C10 is the loner for finding the right physical size replacement. Its a 10mm and the only caps I could find with the right uF/V are 12.5mm. It's spot on the board is drawn as a 12.5mm.

Funny how across brands and series the stats are identical.

[Toasty]

Re: Suspect Bad PS Apple AcBel API 1PC36

C10 is okay at 12.5mm.
C21 - you'll need to "adjust" the coil next to it to get a good fit.

C8 - 47/25 - KZE ; Specs on DK don't match your sheet for PW {180/600} looks like it was HE.
KZE is 250/300 and 2¢ cheaper

C35 - 470/35 - KZE and 12¢ cheaper --> FC is below original spec

C21 - 3300/10 - KZE too ...and it's 20¢ cheaper ...sensing a pattern yet ?

Toast

[Double check me on these. It's easy to go bug-eyed on those sites. ]

[buz]

Re: Suspect Bad PS Apple AcBel API 1PC36

Did you use one or the other of the two sites? (Digi-key or Mouser) .... maybe it doesn't matter. I'll make the changes in a few mins.

[buz]

Re: Suspect Bad PS Apple AcBel API 1PC36

Pretty much answered my own question (above) - DK had everyone you pointed out. Much better stats on those KZEs.

Not real sure about the pattern other than the UCC's are original equipment with seriously better stats.

Thanks for that on C10/21. Guess I'll get my order in (may check it again) - am anxious to see if the ol' girl will start again. Any bets?

[Toasty]

Re: Suspect Bad PS Apple AcBel API 1PC36

No bets. It -will- work.

[buz]

Re: Suspect Bad PS Apple AcBel API 1PC36

Hats off to Digi-key. Order placed the 14th and delivered the 16th from MN to OR, surprised me!!! Really didn't expect the order till next week.

So any last min tips on going back the other way? Think I'm gonna look some more for that 1/32nd inch (.031") solder ... other than that idk, just put it back like it was.

[buz]

Re: Suspect Bad PS Apple AcBel API 1PC36

Found the solder, checked the order - looks like it's time. Wish me luck.

[buz]

Re: Suspect Bad PS Apple AcBel API 1PC36

- well there's no joy. Same symptoms as before. Fans pulse, mobo LED flashes on and .... nothing. Dang it!!

Any thoughts?

[Toasty]

Re: Suspect Bad PS Apple AcBel API 1PC36

Jumpered on, does the PSU run by itself?

[buz]

Re: Suspect Bad PS Apple AcBel API 1PC36

Jumpered on?

[buz]

Re: Suspect Bad PS Apple AcBel API 1PC36

I'm bummed ... sort of. I was getting ready to check the pin out and tried a start up. Recalled a kb article about the power buttons on these machines being off center from the switch behind and 'may' need to press the pwr at the '11 oclock' position - I did and it start up.

Why bummed? Well, wondering .... only came across this after committing to the PS rebuild. In the bigger picture, I certainly don't regret 'freshening' the PS and the learning involved.

How do you jumper the PSU?

[buz]

Re: Suspect Bad PS Apple AcBel API 1PC36

Toast ... realized I may have sounded ungrateful for all your help and that is NOT the case. It's running and thank you.

[Toasty]

Re: Suspect Bad PS Apple AcBel API 1PC36

Glad you got it working! The "freshening" certainly won't hurt in the long run. It might actually be running 5 years from now.

Look up the other threads in here. There is a pinout diagram in one of them.

Also, coming off the PSU board, the wire was marked ON/OFF or some such. I think it's the Green wire.

Disconnected from computer and plugged into wall outlet, jumper from that to any black(ground) and it should come on. I don't know for sure if this requires a load to run. Some do.

Toast