EEVblog
ฝัง
- เผยแพร่เมื่อ 4 ต.ค. 2024
- A deep dive into how and why Apple's post 2016 MacBook Pro's fail after water ingress, and the poor design decisions that lead to it. Which engineer at Apple is to blame? Or is anyone actually to blame?
The PCB and product design process, the often siloed design jobs, and failure mode analysis
Includes a demo of water hydrolysis on a PCB.
Louis Rossmann on the Apple botched design: • Apple botches Macbook ...
Water ingress in a MacBook: • How do I fix a black s...
Paul Daniels and Open BoardView: pldaniels.com/
Forum: www.eevblog.co...
#Apple #MacBook #Design
Bitcoin Donations: 38y7DE8HEHNj8fGDtUr4PkCn9nWxiorvvy
Litecoin: ML7oQokTwB38bgzzjLDbRV97HKAHuwRfHA
Ethereum: 0x11AceA38DCA9DbFfB4F35f3F746af65F9dED28ce
EEVblog Main Web Site: www.eevblog.com
The 2nd EEVblog Channel: / eevblog2
Support the EEVblog through Patreon!
/ eevblog
AliExpress Affiliate: s.click.aliexpr...
Buy anything through that link and Dave gets a commission at no cost to you.
Stuff I recommend:
kit.com/EEVblog/
Donate With Bitcoin & Other Crypto Currencies!
www.eevblog.co...
T-Shirts: teespring.com/s...
Likecoin - Coins for Likes: likecoin.pro/@...
For those interested, this design began in 2016, and has persisted throughout 2017, 2018, and 2019 with no changes.
While the fast-blow design "technically" abides by the rules/requirement, it'd be nicer to see the return of the more defensive design of past. Imagine if I designed software like Apple designed these latest machines... mmm... oh wait.
Louis Rossmann
you da man Louis
Usual Apple BS
Then its a problem, looks like its "Blame the customer" style
Thanks for the mention there Dave - will have to send you a licence of FlexBV.
It's Paul Daniels!
@@posae86 no, it's a cat!
It's a cat called Paul Daniels!
Hey it's Mr. Daniels! Hihi's here somewhere too. The gang is all here on another channel!
@@A2000MHz I'm just being pulled around to where ever people are tagging me... like a leaf in the wind, see how I fly...
Putting ground pin next to BCKL was standard Apple practice ... until this model, but as Tim Cook wrote in a letter to investors this January: "People Bought Fewer New iPhones Because They Repaired Their Old Ones".
BTW Backlight circuit fuses in Apple devices almost never trigger, connectors melt/go up on fire first :)
This person did their homework on failure modes!
Louis Rossmann
Look who’s up at 2:13am.
Standard *deliberately engineered* practice, or just "standard practice" by accident?
Like I said in the video, even if it was deliberate choice before, a new designer could have come along and have not known, or it wasn't documented etc and made an innocent change without knowing.
Some people like to try and imply this was a deliberate design choice to enable premature failure, but I'm not buying that.
@@EEVblog Dave - the first model it can be just by accident, with the second model they should have known better, but a 3rd?
No, they can not be so dense that after having had issues with 2 generations of products, legal issues and forced to offer extended services, they didn't figure out that they might made an error and fix that for the 3rd generation.
(And for electrical engineering you normally do learn about things like leaving place between signal and power lines to protect the more delicate parts - and at least one person should have brought that up during the design phase)
@@ABaumstumpf I thought this *wasn't* an issue (the CPU dying thing) with previous designs?
Former Apple engineer here. I can well imagine how this happened. Young engineer joins the company, so the team gives him something simple to start with: draw the schematic for the display flex circuit--seems easy enough. He/she correctly puts the high current pin on the large contact, and then proceeds to label the signals starting from pin 1. There's no design rule forbidding this and nobody points out that this is sub-optimal, and the prototypes get built. The reliability team tests these units, and they have their hands full ironing out issues that happen in nominal use cases (e.g. butterfly keyboard). The product doesn't claim to be waterproof, so water ingress is low on their priority list. When they do test water, all sorts of things go wrong, and the display connector frying is just one of them, so no one pays this particular failure too much attention.
The boards eventually go to production, and now it's a pain to fix, because you'd have to change both the display and the motherboard, and go through all the testing again to make sure your fix didn't break something else. Also, you'd have to set up your supply chain to manage the different configurations, etc. Realistically, for relatively small issues like this, the next chance to fix it will be when they change the display, which would be a major updatel. Hopefully the engineer who drew that schematic is watching this, and will learn from it for the next design. But all that talk about planned obsolescence is just ignorant--we tried to make these as reliable as possible so we wouldn't get calls from the factory on a weekend complaining about failures.
Yep, good explanation. I don't buy the deliberate planned obsolescence thing either, especially when it comes to the location of this pin.
Don't subscribe to malice what can be adequately described by bureaucratic process and oversight.
Sure, Apple don't give a shit about the repairability of their products, but having engineers design in deliberate failure mechanisms is just fantasy.
@@EEVblog A big reason Apple doesn't like independent repair shops is because a lot of them ham-fist the repair and damage the battery. The majority of battery fires happen to devices that have been to an independent repair shop, and when batteries catch on fire, you get bad publicity and sued, whether its your fault or not. Also, third party components sometimes reverse engineer the firmware to make a compatible system, but breaks when you update the firmware (and then the customer complains that your iOS update bricked their device, and the engineers have to spend time analyzing the failure only to find a knock-off part). Everyone thinks trying to stop 3rd party repairs is a cash grab by Apple, but it's more about not having to deal with all the failures caused by shoddy repairs.
I thought this was a louis rossmann video. I got bamboozled. I am subscribed to dave as well, so I'm not mad.
Yeah same
Richard Smith I’m only subbed to DJ, and was avoiding Rossman for a while so the suggestion bot would give me other things to watch. When the bot picks up on this, I’m mad enough for us both.
I also thought this was a louis rossmann video
We're all mad here ;)
Same, I'm getting slow and inattentive as I get older.
If anyone is to blame, it's the customer that refuse to hold the company accountable. I know someone who, after TWO Apple laptops burned up on them, bought a third. No warranty stuff here; they bought three.
Insane stupidity is a disease.
The dark compound appears like "carbon", but actually, most of them are copper oxide or other copper compounds.
Anyone can make a mistake, happens to all of us. What happens next is important, here Apple has not addressed the issue that is the real problem.
Actually you ought to watch the videos on the events of 3mi. Island’s nuclear reactor meltdown, then talk about human error.
4 generations of broken Butterfly keyboards and how many 5+ years for a new, overpriced MacPro that few people asked for in such a way (aka. normal modular Mac for normal people). But yeah, changing the pinout causing more problems smells a bit like: "we do did not have enough sales and water damages, Let's make sure stuff blows up more often", ..! :-/
Bits & more by René Rebe
I wasn’t planning on going that far downhill with the OP’s little red wagon, but if it must be done, you must.
they will just ignore it pretend they didn't do anything wrong and tis all the fault of their users. then promptly fix their mistake in the next iteration.
@SteveQ Well said.
Dave and Louis in one video? I've never clicked so fast.
I never hesitated watching an EEVBlog video for THAT long. Or any video to be honest
That's what she said.
@@juweinert Sometimes they both seem to ramble or get really repetitive and I find that annoying. My first thought upon seeing Rossman's face on here was "Are we watching a video about a video and both of the videos feature repetitive ramblers?" That's some ramblingception right there!
I thought they were going to talk to each other. Two ramblers. Who would win?
The problem with Rossmann is not that he finds various mistakes. The problem is his ferocity and hatred.
I think I would have the nearest pin be a sense pin. If it sees any voltage it immediately shuts off the backlight, and keeps it off until the computer is shut off. The next nearest pin would be a ground pin to shunt any actual power to where it won't damage anything else.
Just wanted to write the same comment.
Condensation: Often missed as a source of water damage. Take something that's been outside in the cold and put it in a warm (and therefore more humid) room and you get water everywhere. Keep a slab of aluminium in the fridge and then take it out to show someone condensation damage, quite an eye opener
Or just look at your bathroom mirror after a shower.
Full aluminium body + terrible (passive) cooling design, cold metal + hot board = moisture. Sounds plausible.
Actually, condensation is strictly a cooling phenomena, during which the warmer water vapor upon cooling down “solidifies” so to speak, thus becoming liquid. But perhaps you mean it exactly that way, in which case my comment is wholly superfluous.
@@orthodium yes, it would be the cooling of the warmer ambient air on the cold laptop, if not turned on.
You two should have chat about it, but other laptops have their boards wrapped in foil and case/keyboard are designed in a way to route the liquid trough the holes in the bottom of device, where in Apple your sneeze can shoot trough keys and land directly on tracks.
It's a bit of a tangent, but it'd be cool if you could test the effectiveness of various conformal coatings against moisture with your simple 50v test.
It is quite a tangent...to a tangent.The 50V test done here is to approximate the conditions in a particular connector, yet the conductor materials were themselves approximated, supposing that FoxConn had used pure copper in them.
What may have happened here is that the LCD module changed (eg. due to an updated CoF driver) and the connector pin assignments along with it. The PCB layout person may have then been handed a completed LCD module and interconnect pin assignments by another design team and have had to work with what they were given...
You can reorder connections in the flex cable...
@@kilrahvp Agreed, however the LCD design team may not have considered pin reassignment on the flex or Apple's design rules may not have permitted reordering on a flex. In either of these scenarios, the PCB designer's hands would have been tied.
Dave's point about design by committee (ie. multiple teams and inter-team communication) may have had deeper and more subtle implications for the finished design...
I just watched Paul replace that mux chip. I'm in awe of anyone who can herd those tiny little balls into place and actually fix something. Bravo!
Thanks! Anyone can do this soldering stuff if you have patience and attention to detail. And a good microscope and set of tweezers!
@@A2000MHz and limited supply of coffee is advisable.
Any other Pauls want to chip in their 2¢ here?
@@HighestRank So many... I bet we can find some more!
And a very small and appropriate amount of flux.
I have seen the "Silo" structure in software development - and the idea of kicking things "over the wall" so it becomes someone else's responsibility is all too easy. I've never liked it.
Thats what dependencies are.
Sometimes you just have no choice as a dev. You do the task you're assigned and have to rush off to the next massive emergency and have to forget about the last thing.
@@Tomasu82 Been there. Done that. What I'm talking about is the formal establishment of such a structure so that EVERYTHING gets done that way.
At one point it was very common to have two connectors going to a laptop display, one for power, and a separate one for video. That's something they definitely could have done and as long as the connectors were physically separated even if they shared the same flat flex cable I don't think it would be this much of a problem
Having done layout with similar connectors I can offer another explanation for the numbering system. In my experience the larger pins are considered to be for mounting the connector and are to be electrically GND’s or non-connects. The narrower pins are considered to be the signal pins and numbered accordingly. The larger pins are numbered last in the series with the highest numbers so they will appear last on the pin signal assignment list. This way they can be grouped together and specified as physical pads on the footprint without an electrical connection.
In large organisations, designs normally have many review stages with detail analysis by teams of qualified people, some assured to be qualified but not directly on the same project , so less likely to overlook errors
I just don't understand how you can design a proper schematic without knowing how you are going to lay out the board(s).
Similarly, I can't understand how someone could place and route a board effectively without understanding the functionality of the circuit. At least for me, I couldn't deal with other people's bad habits or lack of foresight. Thankfully, I am able to have control over my entire design from start to finish. It's a lot more work, but I like it that way.
For the price and attitude, Apple should not be having these issues.
Usually it's a co-operative process and the various actors peak over each others cubical walls.
If doing it properly, you would train your staff to have the same, or similar mindset (set of design rules) when coding, drawing, testing etc. Only small companies who can't or don't want to afford training cost, just throw their staff in the big pool.
IIRC older Thinkpads had liquid egress from topcase (keyboard) through to exit bottom case. Apple, instead, is a TRILLION dollar company because they do not waterproof their products. They sell Applecare+. They sell parts and service. They sell replacements. Now, let me point out that this is "Accidental damage". This is NOT normal design behavior during normal use. The designs after they sealed up the battery access now prevent a user from shutting down, then removing all power sources. You cannot do this anymore. You can argue: It's a portable and will be exposed to areas where the user consumes beverages, etc. Apple will rebut: it was damaged. This is proof. Would you like to apply your Applecare + Insurance coverage premium? Is your data backed up? (it better be as there is NO excuse for cloud syncing or using external HDD with TimeMachine.
If Apple designed and sold waterproof products, they would lose billions. Customer solution: stop buying Apple products.
I even have a 25$ Logitech wireless keyboard that has liquid drainage holes threw it!
My current gen one does too
In general, separating power and signal lines as good as possible is one of the most basic things anyone working with electronics should know.
PERILEX there is not one good, no, not one -JC.
@@HighestRank What a shame...
PERILEX Actually, with limited pin count, using a quiet DC line as shielding may reduce crosstalk. Though in this case it's an outermost pin, so no benefit.
@@johnfrancisdoe1563 Crosstalk in form of interference is only one of the reasons why power and signal lines should be separated as good as possible.
Preventing high(er) voltage damage to low(er) voltage components is actually the main reason.
16:22, at the left edge of the water droplet, the black solid conduction filament begins to grow, really fascinating to watch.
I klicked on this because I thought it was a new Louis Rossman video. Turns out it was something even better. Got both of my favorite channels in one =)
Would be interesting to repeat the 50v water test and actually put a ground rail in the middle - does it prevent the other pin from corroding?
The adjacent ground pin would corrode. The metal from the +50 volt rail would literally dissociate and be deposited on the ground pin. The system would still fail as Dave mentioned, but it would at least protect the low voltage logic chip and allow for a repair.
If you actually read the thermodynamics insert, you’d realize that ionization due to electrolysis only requires an anode, cathode, and a little under 2 volt potential difference.
@@gavincurtis Yeah but I'm not confident all of the metal would be deposited on the ground. If the other pin was at say 1.8V, there's still -48.2V to the high voltage rail. I agree it would definitely slow down corrosion but would it definitely stop it? Not sure
Can someone explain to me, why any electronics engineer should consider liquid on his board a valid influence instead of blaming the case manufacturer for not having goot IP ratings?
I am a systems designer and as a designer I have to take responsbility for operation of my designs in a real world environment. I design critical systems and a design like the one in this video would never pass our QC. The case is part of the Apple design so they have control over that as well. Also a 60V rated connector for a 50v backlight is no where near sufficient as there is almost no factor of safety. It does not allow for board contamination or higher himidity. It is a bad design but apple really do not care once they sell something they wash their hands of responsibility.
I think Apple cares as much as customers care. The same applies to fast fashion and other stuff. As much as we care in a small group we are in minority. People just switch to a newer stuff and don't care about quality or sustainability :(
Actually it's not a design flaw, it's a feature.
As Apple says: "Think different!"
think different was jobs quote ... now the crooks at apple thinks profits only..
Think Different, Think Broken!
Well, the solution was done with the predecessor: Have a ground pin next to a power connector.
And that's what I'd do as well. Can't have too much ground!
Less coupling inductance too because the ground is right next to the power so the field loop is smaller.
it is likely cupric oxide / Copper (II) Oxide on the positive terminal
Knight Sun I agree. Liquid water keeps temp below. 100C. That is not really hot enough for carbonization. And the “spring water” minerals probably helped the process along.
Fizzy soft drinks usually ad phosphoric acid to help with oxide layers.
Actually, copper is forming as a fine powder near the cathode (upper left copper strip) and thus it doesn't have the typical metallic shine. At the anode Cu is oxidized and dissolves into the water. There is a faint green/blue coloration visible in the droplet. The shiny pure copper underneath is exposed at the anode. At the same time, Oxygen is formed at the anode, happily bubbling away.
The whole process is basically electrorefining of copper, but with an over voltage (typical voltage for copper refining is 0.2 - 0.3 V in sulfuric acid). This excess voltage speeds up the forming of copper and doesn't allow it to crystallize in an orderly fashion, i.e. to form proper crystals. It also allows for the formation of molecular oxygen at the anode which normally doesn't happen at voltages used in electrorefining. At some 50 V you're easily overcoming the resistance of water due to impurities.
@@bonham1981 I thought pure water was non conductive. But maybe it can ionise over short distances due to the polar bonding?
I checked and found the distances are VERY short: pure water electrolysis has been achieved, by using nanogap electrochemical cells, where the distance between two electrodes is less than Debye-length of water (which is less than a micron, which is 0.0001mm)
I have a large Dell monitor with pretty much the exact same problem. There's a wire cluster going from the main board to the screen, and it even has twisted pairs, separated by grounds. The problem is that somehow, this particular wire cluster got fried, the insulation melted off, shorted out, and half the caps on the mainboard are bulging now.
The electronics design firms that I worked for all had environmental chambers to test products for temperature and humidity sensitivity at different air pressures.
They also had shaker tables that could dislodge the large heavy components on power supply boards and toss them into the air.
Testing is fun, Apple should try it.
but apple thinks different.
Shaker tables are fun, I rode one at an adult sleep over one time.
mechanical engineer here: why not just a small dab of di-electric grease in the connector to keep out the water in the first place? It's pretty common in cars, and industrial equipment that needs to be rain and/or submersion proof.
Can do, I've seen that before. They'd be slashing every cent in assembly time on these.
Might be a good option for those repairing these.
Louis is a great chap, seriously. I recommend him too for his words of wisdom on subjects unrelated to Mac Books.
Spring water always has plenty of dissolved ions, particularly Na(+), Mg(2+), Cl(-) and SO4(2-) so you're always going to get this electrolysis behaviour. It's not utterly pure distilled and deionised water. The issue with Apple is where the connector is located on the board - right next to the edge. If it's positioned away from the edge there's less chance of a tiny amount of of liquid will utterly bugger it up. Solution of course is to not run a long string of LEDs in series, instead run them more in parallel. Modern LCD backlights have more individually addressable LEDs anyway to light up smaller sections to improve zonal contrast.
They should sell bottled I-water (or perhaps, de-I water) in Apple stores, perhaps?
If there was a pin to the ground between the backlight 50V and data 1V, then all the current from 50V line would go to ground and not to cpu/mux. Of course, the connector would still be destroyed, but it's about saving whatever is possible to save in case of disaster. Like in earlier MacBooks, where there're pins for backlight, ground pin, pins for data. There was a separator ground pin (macbook air 2013-2015 boards 820-3437 and 820-00165 as an example of such design). But as we all know, Macbooks are not supposed to be fixed. They're supposed to be recycled to buy a new one.
But as per the question I raised in the video, was this a deliberate design decision previously? and/or a deliberate decision in the new design? or was it just an oversight?
@@EEVblog I guess calling it a deliberate design flaw would be a conspiracy theory, at least without any proof. Presumption of innocence should always apply imo, despite some people wishing to believe in the opposite when it comes to large corporations. Probably an oversight, or a careless wish to physically shrink the connector without giving a thought about such a problem possibly happening.
Errrrr is it just me or "dont chuck water in your 3k fondle slab" I would bet that most laptops will suffer from similar water ingress issues.
the problem with "failing safe" is that if you go to apple to "repair" it, they will just replace the entire board, and with their laptops being designed the way they are with everything soldered on this essentially equates to an entirely new computer. for them it makes no difference if a 10 cent capacitor fails that craps out the entire computer, or a CPU fails... it is all the same procedure. so why design something that can "fail safe"?
The #1 reason for Macbook failures is water damage. So yes, Apple designers should have introduced mitigation measures several versions ago. Especially for a device of this price/premium quality. Then again, given how much they charge for repair, and how tenacious they are about thwarting right to repair, they were probably told not to prevent water damage issues.
It's like when you discover that two of your friends actually know each other :)
I didn't look too closely, I just assumed we had EEV as a guest on the Rossman Repair channel.
I'd also like to point out that a board with a fried CPU = a board with lost customer data since the dipshits soldered the SSD to the boards on the vast majority of these type C macs. The older type C ones have a PCIE breakout for the SSD, newer ones don't. Yeah, you should make backups, especially with how well time machine works, but still, its really an AWFUL idea to solder the SSD.
If you have valuable files on your computer and don’t backup your hard drive you’re gonna have a bad time.
0:25 You missed the opportunity to mention that Louis also makes cat videos.
Actually, Clinton makes great Louis videos.
Thos black gunk is copper hydroxide
I like Louis and respect his work on "right to repair" but I had to stop watching his Apple vids because it's like having to listen to someone talk about their ex-wife.
He gets to extreme at times.
Apple don't like different departments discussing information to each other I think that's why you have all these inconsistencies
No its "planned obsolescence" and its new "Think different"...
I have 5 mbp sitting with those connector issues, the connectors are on backorder.. thanks apple for putting food on my table!!
If there is a guard pin, you replace the connector and perhaps a trace. Without the guard pin you fry the CPU.
A ground pin in between WOULD INDEED protect the CPU data line from +50V in case of water ingress! Better if pixies flow to GND than to a data pin. ...but more damage is in apple's best interest, since it gives independent repair a bad day.
Better design if there is no fluid in first place. Owners fault
Extra trace and pin would take space and affect the rest of the design. They don'
t claim these can take spills.
@@jhoughjr1 Yes
That doesn't even require an extra ground pin. Just swap a few pins around. Apple LCDs aren't even interchangeable between models, are they?
21:40 Yeah it blew a fuse - the connector IS the fuse. Polyfuses always intact!
I saw a LinusTechTips video where they left a Dell XPS 15 in the rain for HOURS, and nothing happened to it, just worked like normal. A macbook pro? dead in seconds.
I remember that video, it was pretty impressive.
If you're interested, here's Louis trying to kill a thinkpad P50: th-cam.com/video/ig3xI8dUdm0/w-d-xo.html&t=1700
~$1000 laptop = I'm not scared some dumb water.
$3000 macbook = HEY WATCH IT! You almost dropped a single tear on my keyboard.
Cody and how do you know it died in seconds or that the Dell kept working after the water had a chance to corrode the PCB
I'm glad to see Louis make it to the EEVblog. Noble shit commentary for the win!
its louis, not lewis
Deilwynna this OP actually holds a record for fewest grammatical mistakes by Mike. Give him a break for once.
Oh! I though he said "no bullshit commentary", like straight to the point, no sugarcoated.
He was on the amphour once.
I was taking a break from Louis, but I’ll be damned if that wasn’t in the cards.
Couldn't the carbonization be helped by the connector being black? This means a lot more conductive Carbon Black is already there.
Good point
Was waiting for this one since the last AmpHour!
26:06 fail - they are not just "ground pins they could use to shift everything down". They are pads for shield part of connector
It's fun to deep dive with hind sight but I think most likely it ended up being a business decision. You touched on it in the end there, Apple has no intention of trying to solve every problem and there's always the issue of not just fixing the problem but testing and validating any collateral effects that change may have. No doubt they have done the analysis to figure out what percentage of devices experience this specific failure and have opted to go the route of letting those be handled through the service process. I love Rossman but he interacts with an extremely tiny percentage of total device sales.. even the noisy forums still constitute just a small fraction of sales -- most of which go through life without ever experiencing a single problem.
To flip the perspective with an exaggerated analogy, go drive your car into a lake and then see how quickly the automobile industry switches all vehicles to amphibious... ain't gonna happen.. it's a failure mode that is intended to be handled by other methods.. not engineering solutions.
agreed, sometimes you just have to say "well if the user does XYZ, they break it." as a more related example, should apple make the iphone tough enough to be used as a hammer? I mean i've seen trade show demos of people doing just that without breaking things, so why isn't apple doing that? ohh, right because it's a dumb failure mode, that isn't worth designing for.
Your analogy is thought provoking. Especially given that (in the UK) more people die drowning in cars than in boats. Not only has the automobile industry (understandably) not switched to amphibians, but they have actively switched from windows with winders to fully electric. The end user prioritises convenience over cost, and a combination of the two is prioritised over a subset of risks.
The subset of risks is heavily cherry-picked, though:
Other risks (like the discretionary risk of omitting to fasten your seat belt) are, perversely, prioritised over almost everything, witness the mandating of air bags, even though these can kill you if they go off inopportunely. I'm sure the knowledge that they save people who are too disaffected or disengaged to buckle up is a great comfort to those who are in the act of driving of a cliff, with their seat belts buckled up and their faulty air bag deployed.
The engineer may not be able to fix this issue on the next revision, it depend on what that connector connects to. If the board is supposed to be fit form and function identical they would not be allowed. The possibility of connecting a rev 2 board into a rev 1 system or visa versa would be too high. Especially if the board is available as a spare part
I'd say it's 100% apples fault. Thinkpads for decades have included drainage from the keyboard tray to protect the laptop. You could pour a full glass of water on the keyboard and just keep typing. Bad design is their problem.
Apple keyboards kinda have drainage, the backlight panel is kinda like a gasket between the KB and the logic board, it has some holes in, but it leads straight to the logic board!
26:10 The large group of ground pins in the bottom of the LCD connector is actually an outer shrouding. Still there are pins 18, 20, 22 that can be moved closer to pin 43 and make a good protective ground.
What’s the fuss about a +50 volt solder pad 0.35mm adjacent to a serial 1.8 volt line directly off the CPU? I did stuff like that when I was making my very first PCBs when 15 years old.
15 is way too old to get started on PCB design! I was significantly younger when I designed and made my first board.
@@flex209 Cool.
You have much better tools at your disposal than when I was 15 I bet. We didn’t even have a laser printer with transfers like today. I had to use rub on transfer stencils that took way too much time, but worth it for a professional look. 😊
Everything else was perfboard.
@@gavincurtis I don't think so, I started with the Sharpie method and could never get the toner transfer method to work properly. This used to take a lot of time and the results didn't look very good...
Same here. Sharpie pens just couldn’t get a thick enough ink layer for me. Only the rub on transfers produced decent results for me. The first PCB I did was a MTS stereo TV decoder with stereo amplifier.
not trying to justify Apples mistake but if they would change pinout of that connector on subsequent revisions they would also need to change the connector itself to prevent people plugging in older display connector into newer motherboard and frying things anyway. soo yeah a ground patch surrounding the higher voltage pin is best idea probably..
Was the Macbook designed to work under water ???
Why would water get into the Macbook ???
If I drive my car into a lake.. I'm sure it would not function properly and start corroding.. yet is weatherproof.
Exactly what I was thinking. How many MacBook Pro’s out there are still working fine because the user wasn’t a clumsy idiot who poured water onto it? Thousands! I would think there are plenty of PC laptops that fail exactly the same way if you pour water into them too.
@@portreathbeach Of course there are. But Apple "sheep" don't hunt around for PC repair videos to comment "HA! Dell? More like CrapDell!". ON.EACH.AND.EVERY.ONE.OF.THEM. °_°
This is like saying your TV/hifi/bluray player has a design flaw because it's not water proof. They're designed for indoor use and therefore there's no call for waterproofing them.
Apparently today idiots demand their shit is idiot proof. Of course, nothing ever really is.
I find it hard to believe that they wouldn't check the location of The power traces and connectors in relation to those carrying data. It's not even an alien concept as many of the data traces are separated with ground to reduce crosstalk, this is ESPECIALLY true on cables/ connectors.
This screams planned obsolescence to me.
They couldnt fit an extra ground on that side of the connector. Why they decided to put one between 5V and data, and not between 50V and data instead though...
Gareth Evans crosstalk is reduced generally where it is needed to. Video mux signals may still be easy to discriminate if the parallel runs are kept to a minimum.
I guess if they where to fix this in the next revision, it would mean that they can't use the display assembly from the last revision.
Good point.
It could be that if they fixed the flaw, that would be admitting a mistake was made in the board design. That could open the door to a lawsuit and replacing a bunch of boards for free.
That would only be true if they also claimed the MacBook was water resistant.
dothemathright 1111 was it an iterative design? You’re inferring they put them together on purpose. Unlikely.
dothemathright 1111 so you’re saying the pcb is an iteration on the old design. Are you sure???
dothemathright 1111 that’s the most absurd comparison I’ve ever heard.
dothemathright 1111 and before we start arguing over semantics name me a computer company that fixes laptop water damage no questions asked... I’ll give you a hint, not one single manufacturer
Maybe don't get liquid in it?
You could be sitting in your backyard on your laptop, and suddenly it starts raining. Even if you run inside with your laptop, there should be SOME level of water resistance.
Most people having this issue, are morons using the laptop and spill shit on it, then get pissed when their shit expensive laptop blows up after they continue using it. It's their own fault, and cry foul at apple instead of taking responsibility of their own mistakes.
@@JosephArata I agree a little bit except here's the thing, you can find equally expensive or even cheaper laptops that aren't waterproof but will easily survive being drenched in water, now is it the customers fault for expecting an ounce of quality from a $3000 device or Apple's fault for not designing their products to survive at least small spills of liquid?
For example here is Louis attempting to murder his Thinkpad P50: th-cam.com/video/ig3xI8dUdm0/w-d-xo.html Now after watching that, both devices are expensive but do you think Apple's devices are anywhere near the same level of quality? And should a trillion dollar company do better or should they just tell their customers to quit whining over split milk?
@@JosephArata ya, my last apple purchase was a 2015 $90 non woking MacBook Air,,,, after I cleaned all the coke syrup off the mother board and under the wifi card. it magically worked, and still works. not one part replace.....thank you for clumsy people I saved $1,200
@@vgamesx1 still a consumer problem, if they would stop buying apple's products, the company might take more accountability for manufacturing products that are overpriced medium-tier hardware. Can't fix insane stupidity, when apple takes full advantage of it.
I am feeling a little nicer about the apple engineers but, apple has way more money than the rest of us so it seems like they should be able to shield from water electrically or more likely from case designs. It is much more gray than I imagined.
easy fix, just have a separate connectors for power and data.
Three years ago, I've accidentally spilled coffee all over my Dell laptop's keyboard. It absolutely soaked it internally and I had to disassemble it for cleaning. Fast forward to today, and all that failed were some keys of the keyboard. It still works fine.
A colleague just did that with a $6k Thinkpad and it was instant junk.
"post 2016 MacBook Pro's fail after water ingress"
As much as I hate Apple, this is all starting off as pure PEBKAC. Water should never be going inside a laptop computer or any of its connectors. I'm sure poor design choices on the part of Apple could exacerbate the results, but in the end it's the user's job to keep their devices separate from their morning coffee/beer/milk/water/etc.
Maybe it's time to invent the iCondom?
There's still better choices than putting data pins next to power pins when there is an option, and that connector has a lot of pins that could likely be shuffled around for a better result in the event the user is stupid (which they always are) and allows liquid inside the machine. Certainly, a failure is likely but the kind of failure and its severity can be mitigated through design... like not putting high voltage pins next to pins that connect to high value, very sensitive components without having considered that possibility. Having a ground pin between these pins would have been nice, I think. Some real effort into their rather expensive products to resist customer stupidity would be nice, but that would mean you didn't buy a new Macbook every time you spilled coffee on it. Let's be real, laptops are going to have shit spilled on them and not planning for that is simple laziness.
Ask for a miracle.
While you are absolutely right that end users shouldn't be spilling liquids onto their devices, this isn't what the video is talking about. It is asking who should be held accountable when in the past Apple had shown and did design their hardware to mitigate this issue. From what I have seen from watching Louis's videos, Apples quality has went down the drain and the end user who buys their crap won't hold the company responsible.
So lets ask this question a different way. Would you buy a laptop where if you spilt water on it in the same area that it would short out a cheap fuse and was easy to repair, or would you buy the laptop that when water was spilt in that area shorted out $180 or more GPU. If you say the first one, well Apple use to make that laptop but now they don't in my opinion due to laziness or because they can charge more money for the repair or trick you into buying another laptop.
Just like how their phones still have the exact same flexion based damage in the exact same spot on all phones since Iphone 6.
Accident do happen.
Matthew Bailey nailed it. Of course water in the notebook is the users fault, no one questions that.
@16:22
BigCliveDotcom mentioned the problems with electric induced corossion for ships,
and said to prevent it when grounding the ships while powering in the harbour they use large full bridge rectifiers.
Not as a rectifier, but wired so the diodes are in series and AC can pass both directions, but passing two diodes while doing so.
this means with your 1.23V and a Diode voltage drop of around 0.7V you are getting up to 1.4V, before anything happens
which hopefully blocks electrolysis for the most part.
In BigClive's example the Dock and the ship with different metals formed a battery, with sea water as conductive electrolyte.
Now drinking / spring / mineral water is often much worse than tap water, but i guess the horror is coke or coffee...
Apple does no conformal coating even where there are no connectors, so its really built to fail,
since many producers integrate protection for spilling drinks on the keyboard for years now...
Cheapest way to sell a new crapple since its clearly a "no warranty" issue.
They could sell iWater and iCoffee with anticorrosive in it
My guess as someone who only took AP chemistry: At one terminal you are cleaving the water into hydrogen gas and hydroxide ions (OH-) and at the other terminal the copper ionizes (Cu2+) and migrates to the other side. The positive and negative ions meet at some point and form copper hydroxide which is neutral and falls out of solution. I might be getting it wrong and it's actually a copper oxide instead, can't find any sources on this particular reaction.
In fact due to how heavy the copper atoms are I'm going to guess that most of the current flow is in the form of the OH- which is just grabbing a copper atom once it gets to the other side.
"If liquid gets in"... I'm hesitant to fault Apple considering protection against stupidity can be an endless endeavor. So the design flaw is not so much layout as it is lack of protection against water ingress.
@CK Lim Or live in a sufficiently humid environment, like a town in a hot climate with lots of pools.
Whattabout the problem of condensation from the atmosphere when one takes the Mac from a humid hot enviroment to a cooler airconditioned situation. One doesn't need the Mac on for condensation to form and when you turn it on ........... Spillage by the user isn't a requirement.
The metal adds ions to the water and probably there is a bunch to begin with unless you are using distilled water. So the water gets quite conductive.
Apple has a bad history with liquid. Their keyboards are very water unfriendly for no good reason. The membrane with the contacts is not sealed so basically any spill is a keyboard killer.
Got a real chuckle out of Dave calling Paul Daniels' software "brilliant" after all of the hundreds of times Louis has called it every four letter word in the book.
Friendly ball-busting is a fun working atmosphere indeed :D
how about like after fixing that damage, put dielectric grease insie around that conectro/
Giving away all our mechanic’s secrets, eh zzz?
And cause an even larger reliability problem once that grease gets into the connector. I guess you don't realize how tiny those connectors are and how fast and sensitive those MIPI signals are.
Water damage can be at least blamed on the customer but if the laptop has intermittent display issues because of your grease in the connectors, that's going to be a straight recall.
@@JanCiger Sounds like you're unfamiliar with the practice. It's a common, occasionally factory recommended automotive repair thing (including on ECU connectors loaded with analog and digital signal pins), the idea is to totally pack the connector with insulating silicone grease and then make the connection, pin tension wipes away the grease on the contacts and forms a reliable watertight connection. Whether or not it'd work on what are likely more delicate signals in this case I don't know, but it's not an idea without merit.
@@Broken_Yugo haven't heard about this method being used on what are impedance controlled several GHz MIPI signals and these connectors. Automotive signals are orders of magnitude slower and the connectors much larger and typically more robust than this.
29:22 Apple have started putting humidity triggers (non-electrical indicators), so it kind of hints their attitude
Just make the 50V supply a separate cable. Would this silicone coating stuff prevent this,too? I used to work with hearing aid PCB's exposed to rain, sweat and other nasty stuff.. we had to make this reliable. Reputation damage would be too big. We even replaced whole devices under warranty after people drove their cars over it. Customer satisfaction by company culture. The engineers also had enough time to do their job. We even had the chance to talk to customers to see what we could improve. The only thing i ever changed was protection against people put in batteries the wrong way
I imagine with hearing devices there is only so many people you can screw with... With phones, laptops... Well, you've got more than half of the world's population as potential customers.
Why I like you more than Louis? Because you can say these problems without ranting about apple between every word.
Dave im not sure but polymer coating is not soldermask, im pretty sure ipc says soldermask has no insulation value and you cannot rely on it for any insulation! polymer coating is another thing.
To be honest, I was Macbook user.
But when Steve Jobs passed away, the Apple magic become magical smoke
With respect, it's more like the Apple smoke dissipated and the mirror shattered. To me, Jobs was a con man selling high-priced snake oil.
The thing I liked about Steve Jobs is that he understood the important of functionality/reliability in addition to form/design.
@Bobby Brady Actually, I'm quite familiar with Apples "products". There's nothing about them or in them that reflects the "magical fairy dust" so lovingly "infused" by Mr. Jobs and his ephemeral choir of holy engineers. If there is indeed a cult, it would be of him.
Too bad cancer cut his snake oil sermons short, it'd be most interesting to see the church of Jobs that had arisen from it. I can only hope that he's living out his Buddhist wheel of life as a Chinese factory worker pumping out modern crap Apple "products".
By the way, my response to you is with odious disrespect, unlike my respectful response to Dhavid Setiawan KilluaDhavid
. If you can't debate with a civil tongue, this is the response you should expect.
For small water damage, like a drop of rain or a drop of sweat, moving them far apart would work because a single drop can only stretch so far. Of course if you spill your entire water bottle on it, then the distance won't matter much as you've demonstrated. But considering things like a single drop of sweat or single rain drop is enough to bridge the connections if they're directly right next to each other, that's still a bad design.
Greetings from melbourne. As a retired aircraft tech I look in horror at the things apple do (and probably others as well)
We used an anti humidity coating (Humiseal) on all boards used in all the black boxes. Wouldn't want an aircraft to come down because some moisture got in. LOL
If they conformally coated it people would complain that It was done to deliberately make repair harder.
Apple does not want anything to be too reliable, nor do they accept responsibility for their mistakes. I like Rossman, he's a technician that does not pretend to be an engineer.
He practises engineering.
"It isn't the piece of paper on the wall, it is the things you DO that define you"
~Bruce Wayne 2008
"This would never happen in a smaller company where the design is owned by one person from start to fab..." I think this is a false assumption, even an experienced engineer can easily make a similar mistake (missing a non-trivial case like this water damage).
There's a nice talk by Richard Cook, titled "How Complex Systems Fail", I think it's worth mentioning, might be relevant to failures like this.
Whilst you can't it will never happen. It's true that a single designer with the authority to do whatever they need to do does have a better chance to get it right, even if that doesn't often play out in practice.
But would a smaller company ignore a weakness like this once it was found? Would they keep using the same design for year after year without either a redesign of the connector or somehow mitigating the problem?
Not an easy questions to answer as there is no one correct answer. In the end I suspect it comes down to doing a cost analysis of the problem. What will it cost to redesign the solution? What will it cost in RMA over the years? What will the cost be in terms of current customers and potential future customers perception of the company or product line? What if the money it would cost to implement a fix was funneled into marketing instead?
I have a feeling that the larger the company get the more the bean counters take control of decisions like these.
Quick side note: That connector is made to order... 3 piece design.. so they know exactly what the spacing is.. they could also provide liquid protection.. in addition to the liquid detection bits...
Why do people continue to use the computer after a spillage you should always switch off the power and allow it to dry out. In the IT room that I was in charge off no drinks were allowed near the computers. Prevention is better than cure.
@Bobby Brady I'm wondering what kind of abuse some people throw at their stuff, though. I've never owned a waterproof phone, and the only one that died from water damage was a Sony W810 after I replied to an urgent call under a storm. My old iPhone 4s survived rain, moisture, sand, being abandoned under the sun for hours. I've never had the chance of pouring water on my several computers, and I regularly eat/drink/sleep at the keyboard. Should I start BATHING with my laptop?
@Bobby Brady I am no fan of Apple, but if you wish to take the risk with your own machine that is fine with me.
@Bobby Brady I'd think the rain would count as "spilling" .-D But, yes, I guess being at the computer and drinking is a common situation. I do that too and sometimes a few drops get on the keyboard or screen, but how clumsy does one need to be to pour enough liquid on a laptop to cause water damage? I never happened to actually drench any of my electronics.
Apart from flux. I spill flux all over my electronics XD
@Bobby Brady Not over a computer, they are too expensive to take chances with.
Bobby Brady Not **on** a computer, no. It's been 40 years since my first time handling a computer, so I guess I have sort of a reflex of being extra careful around electronics. I've had other people spill water near my macbook, and once it was A LOT of water and it went under and around; but it didn't reach the internals, which makes it water-safe enough for me.
A more interesting and representative test would have been to leave a trace between the 2 the psu was connected to, and use a dummy load to see what voltage/current would leak to the sensitive part from the middle one
@EEVblog thanks for Sharing your knowledge
Thanks to @Paul Daniels....
makes work a lot easier on Mac and PC motherboard troubleshooting
Flex board view software is awesome 👏
Earlier designs had a ground connection next to the backlight as I remember...That ground would have protected the CPU or mux...
I've had some macbooks come through the shop with literally melted connectors and craters in the PCB. And it was on the trackpad connector on one and the daughterboard on another. Lemme know if you want photos!
The problem is not the electrolysis. It's to electrolysis towards a dataline. If you would put ground pin in between the electrolysis would not jump over to the dataline.
Yay, two of my favorite technocritical vloggers in a single video :)
I discovered Louis a few years ago when his video explaining the bogus contract he turned down for a TV show they were gonna do based on his repair shop. I think it was Vice or some other crappy network. He’s a good guy and takes no shit, and is also a leading figure in the right-to-repair movement.
Where is that video. I must find it.
I'm Louis' new editor... I need to know how you do your clickbait thumbnails... :D
Hi Hihi
I would tend to wonder whether Apple used a standard LCD panel with a pre-defined pinout, or perhaps another group/engineer was in charge of defining the connector pinout on the LCD side, forcing the motherboard designer to use a matching connector/pinout. Not much you can do in that situation except report it as a potential issue.
flomojo2u See Alexey Lavrov's comment about what the standard connector is.
How do you milk a sheep?
Bring out a new iphone :P
RavenLuni 😂