Reball Challenge #2 - Louis Rossmann CD3215 Suffering
HTML-код
- Опубликовано: 13 окт 2024
- Part way through watching another great Rossmann live stream, it was clear after 4 failed reball attempts that some experimentation was required from the community to help save the day.
The chip in question is the CD3215 USB controller, now present in a lot of the new Macbooks, and reballing them is a common requirement, but there's no dedicated stencil for them yet (unlike most SMC chips) and it seems to be a painful process.
What I've found here is that you can use the iPhone 5S Power module stencil from UnionRepair ( www.unionrepai... ) and it has a perfect fit available.
Use the dry paste and combined with the slow-cook method and you should be able to reball these chips with a lot less dramas
When I saw Louis's video I also knew that you would make this fun and enjoyable response.
You didn't let me down Paul. :) Cheers
It was too tempting, had to do it, and since I had to learn myself, figured it was two birds with one stone.
Hey Paul! Yep, that was a long video from Louis, so glad to see you reball that chip! Love the fact you found the stencil for it and did a great job of it! Flattery aside, you did a great job of showing up Louis when he so many times makes fun of you! Bullocks to him! Ha ha!
Wholesome fun, with a chance to also work out what is causing some potential issues with the process. Definitely I can see (with wet paste) why it would be such a nightmare, short of a perfectly flat stencil applied perfectly to a very clean chip, the wicking of the paste under the stencil is going to be an issue every time.
Awaiting Louis to teach me how to code next... "How to make the 'fault is here' button"
I haven’t watched this video yet but I can tell you that I prefer to use the dried up paste on my blade for larger ICs rather than fresh, wet paste from my syringe.
The footprint is called "96-Pin BGA MicroStar Junior." Hint: CD3215xxx == TPS65982 or TPS65986 or one of the TPS6598x family of chips from TI. If it has MOSFETs wired to it in the schematic and the model supports Thunderbolt+DisplayPort on USB-C ports then its likely TPS65982. If there are no MOSFETs attached to the CD3215x then it is likely TPS65986 or similar. Some variants in this family, like the TPS65986 have no connection for external MOSFETs and are for lower power devices. The TPS65982 and TPS65983 have internal MOSFETs and also have pins for controlling external MOSFETs.
Louis doesn't use enough flux
You must be from the Amtech flux factory ;)
The bigger the gob the better the job!
@@almightysoul4475 I thought that was just in the Porn industry 🤣
Some people say the key to being successful at doing microsoldering is to use flux. That's WRONG. The most important thing is to be called Paul. I predict that Louis will soon change his name to Paul Rossmann.
It does have a nice ring to it.
Nothing wrong with a little bit of banter ;) And we all learnt something!
I think Luis meant that problem is not in contacts which are getting loose with PCB, rather than cracks which occurs inside the chip. When you heating chip, materials get wider and temporary fix cracks.
He says that you can heat chip with temperature below solder melting like 180C and chip starts working again, but no solder actually melted anyway. Thefore he suggest to replace chip rather than reballing.
If cracks occurs between chip and PCB, then reballing should work, if cracks occurs inside chip, then chip should be replaced, but it could be that it will be manufactured in same way and will crack later on as well.
Excellent job. How do you do an excellent reballing? When I do it, solder balls solder to the stencil. I need your advice
good idea on scissors, I never thought about it but subconsciously did what said.
Another great video Paul, love watching your stuff. I always watch to the end, I am a fan of your theme music 🙈
You can thank Cody's Lab for that, I noticed his videos had that sort of thing happening, and one of the pieces he had I really enjoyed the tune. I never found the source for it but that did at least get me on my way to finding my own.
Imho putting a capton tape on the back of the chip to keep it tightly attached to stencil will not only help you avoid those leftover solder paste that requires once more air job, but also avoid the chip moving away from stencil.
I should show my efforts learning to reball an SMC for the first time... I spent hours doing the first successful one, but then I stuffed it up when putting it on the board! I did it twice more, both took less than 15 minutes in the end.
Yep, we all go through that sort of process, no matter what the task is, at least you kept on forging ahead so you would get to the point where you can do it routinely.
There obviously needs to be a repair off between you two. Go Aussie!
Can you give more detail on temps by using a temp gun and measure at chip surface during the process then people can match up by calibration.
Louis doesn't need a tutorial, he knows how to do it, he just doesn't have the damn patience for it! He has plenty of chips to send you... or Paul S
Good stuff Paul. Do you have a link to the dry paste? Would like to buy some to experiment.
Take your normal paste, hit it with the hot air at about 180~200'C until the flux has mostly boiled away. It should just start to 'crumble', at least that's how I like mine.
Some people put the paste on a napkin but I find it a little difficult to get the resulting mass off afterwards, maybe they have a trick.
Other option is just to find an old tub of solder paste and the surface/rim area will have plenty of dried out paste for you to use.
I have no paste at all. So far have not had a need for it.
Your paste in the unpreferred method even had less flux that Louis's multiple attempts had last night. Many people in RUclips chat told Louis to use drier paste. Even taking some out of the jar he had and drying it a bit with low temp from the rework station probably would of helped him.
It's interesting, when you're in the middle of a live-stream it can be exceptionally hard to get your brain to shift gears to a new thing, but hopefully afterwards you can try.
Gidday Paul, I’m going to be hopefully sending my computer up next year,
All good - hope you're making good progress with your recovery.
Paul Daniels thanks mate, yeah 8m getting there, , thank bro
#ReBallWars come Soon 🤣
Nice ReBall
Slow and steady Wins the Race
Every time 😀
keep them coming Paul.
That 0.3mm ball ?
Cheers for the cutting tips!
Best of luck with it if you try it.
@@pldaniels cheers! Been trying dry paste and has been working well for me :-)
Good to hear, definitely makes it easier for me :)
8:20 I think they call it annealing.
Great job
Aussie balls are the best balls
And they're made of soLder, none of that strange 'soder' stuff :D
You didn’t think to clean the board?
It's a donor board.
Paul Daniels I missed that fact...
Both methods work for me. If the paste is wet, I make my adjustments. Everyone has their preferred methods. As long as the outcome is a nice uniformed reballed chip, the density of the paste, dry or wet should not be of concern in my opinion.
Definitely if you can work within the margins of allowable error, then no dramas. For some jobs fresh liquid paste isn't a problem, when I'm stencilling for PCBs to be pick & placed / reflowed I'm using normal paste.
I'll be joining you on that journey this weeked.. got a few 820-00164's to play with
Good job!!!
Hi Paul, we still haven't taught Louis to pronounce solder, for some reason most Americans leave out the L, in English it is not a silent letter, if he wants to pronounce it that way then spell it Soder.
keep up the good work young man. Regards from Ray in North Wales UK.
I don't think there's much chance of a language shift alas, we'll just have to persist to be bilingual :)
12:15 filthy chip. Did you clean it up ?
If you're so inclined, you can just put them in the ultrasonic to clean them up, but not a lot of reason to at this early stage. The "dirty" look is mostly due to impurities from the flux/IPA rinse. Realistically, once you put that chip back on with more flux and reflowing, it'll clear out perfectly fine and the whole board will go through an USC at the end of the job.
hahaha thats the same stencil I'm using for the cd3215
What can’t you do ?
100,000 subs and $1 million... for now ;)
what works for you works for you and no one can really say what works for you lol pratice also helps a lot in knowing what works for you ..now this is a lot of work lol
Agreed, if you find a method that works for you, then all good.
Yes i think mr rossmann needs training from you he he :-D
I wonder if a hot plate for the chip would help at all, slightly less heat from the top perhaps?.
I could be talking crap, i often do, so why should i stop now ha ha :-D
Oh, I'd say we could have some interesting collaboration events, mutual sharing of info, hopefully no freeze-spray attacks :D
I suppose the long-heat process I prefer does in effect do what you're alluding to, gets everything up to a uniform temperature before the reflowing, helps the balls bond to the pads rather than getting wanderlust for other balls.
It took me litterly 5 minutes to do this job fir the first time. Only because of the experience with iphone cpu’s 😂
Lol, watched the same stream.
You need to use balls and hand place them on the pads
first :)
Did you camp out in the street to be the first poster? That's some serious dedication - kudos.
I actually got the notification of the upload straight away for once. :)
Ima call this before 3 mins in...u just served louis!