The blanket is new. Without that it seems like this would have been a shot in the dark situation. I like how you figured out whether something was a reflection or not.
Definitely. I need a better one, or an actual "dark box" to put over everything ( a bit tricky with the tripod on the iR cam, I'll need to create something better perhaps ). Once you block out all the external influences you'll then start picking up reflections of the heat from the camera itself ).
As you can see you still do get some transmission through it, and then more amusingly, as things become darker, you start getting reflections of the iR from the camera itself.
Semi-simple repairs pays the rent and keeps the lights on. The capacitors seem to be the Achilles heel on these computers. Thanks for the video Paul. 🍻 PS. You could mark your working doner board and order more of those ICs; one for the board and a few extra for stock.
I could mark the donor, yes, but overall it's easier/saner to have another working board and accept the sacrafice of the one (I do at least have two other working units of that specific board).
Its good you are careful with your voltage injection. We aren't allowed to use voltage injection, probably because the instruments we are repairing reach well into the six figures. We aren't even allowed to use hot air unless it is BGA. The way I would have been required to find that short is to pull the old trusty Agilent 4338B Milliohm meter along with the 16338A Test lead kit which uses four wire kelvin connectors. Probably would have ended up pulling the cap closest to the IC as you did since it would have the lowest resistance, using JBC tweezers not hot air, unless I could get a needle tip probe on an outer ball which can be doable with the right high quality probes. Only then use hot air with the proper sized square nozzel and a preheater if nothing critical nearby, otherwise its off to the CNC machine. Then reball new chip with low melt. If you got it wrong and the chip is fine, expect to speak with the boss.
Definitely a different approach with such delicate/expensive gear. I have got a 4-wire meter here, one I designed & built many years ago (PLD-LOM7) but the biggest pain I had with it is finding decent 4-wire probes that didn't resemble clothes pegs; I've come up with some alternatives more akin to normal probes but to date I've not really *really* needed it... yet. Downside of my idea of suitable replacements is that you still ultimately have some offset that you need to actively null out proportionally to current ( even if it's 1mA )
I noticed that every repair technician blast every component with incredible amount of heat or soldering it really hot. Do you ever killed any component because of the heat or a repaired computer came back with the same problem after the repair? You removed a cap in 5 sec and put a new one in 12 sec, removed a controller chip in 8 sec and put a new back in 13 sec with a bit of preheating, don't you find it a bit fast? The steepest reflow curve 3°C/sec, yours is 29°C/sec for removal and 18°C/sec soldering back (calculated with 20°C room temp). Ones I killed a buck converter with too much heat, but I didn't know at the time, the symptoms was that the power supply always killed its schottky diode which seemed impossible because both the fuse and the power supply which fed the buck converter was rated lower than the diode. After the 3rd diode replacement I realized the controller will be the problem, preplaced it now on a preheated board and the problem is gone. After this accident I always preheat every board. People say with quick heating you don't stress that much the surrounding components, but I don't buy that. I watched I guy blowing up a capacitor while replacing a DrMOS, maybe 1 capacitor is dead, but other 5-6 lifespan shortened 50% because of the extreme heat. I don't say people should follow the refow curve, but what I see in these videos I call it extreme stress on the components. For smaller components I preheat the board from the other side about 1 minute with hot air station, for larger componets I preheat the board about 3 minutes with preheater. Maybe I just waste my time preheating, but I'm glad that non of the motherboards came back what I fixed.
I think better to check on live-videos where there's no editing. I remove a lot of the rework time on edited videos because it's just me sitting there waving the wand around waiting. In a lot of cases I tend to use the want to warm the board area up around first if it's a feasible option. There's also the situation where the board is already hot from prior rework so the next application will be faster. Definitely some parts do not tolerate the rapid ramp well ( hence I tend to cover a lot of things if I can, especially tantalum caps ). I would say it's a balance of risks, which is true of most of the things done with reworking that you see; sometimes it's effectively impractical from a time or setup situation to preheat; however if you do have the time then by all means. One risk with reworking and preheating is that you can inadvertedly induce solder-squeezage out of underfilled neighbours ( particularly with the NANDs and CD32xx chips on these boards ). Your concerns are merited, and at the end of the day it does come down to the balance of risks and what a shop is willing to accept & support.
Hey Paul, great repair. I'm having similar issue with leakage on a smd qfn switch. I think you partially answered my question but do you think that the root cause here is internal to the silicon of the ic, not some corrosion under the chip (like between the balls) where it's impossible to see? You mentioned that heat is the main enemy of these chips and I agree, I think that during initial board fab, the lead free temp requirements comes too close to the death from heat range and it causes latent damage to the silicon so that it fails eventually for leakage/resistive short.
Unfortunately in this instance I didn't check the resistance before/after the cleaning process. I think it's fair to say *sometimes* it's the material on the PCB causing conductivity between pads/balls. In this video the junk was between PPBUS_AON, 3V8_AON and GND so the 17R could have likely been purely the liquid agent; if there was enough conduction between PP and 3V8 it could have triggered the sensing. I could go back and check that part, since I keep those for each job.
Hello from America! Just subscribed. Always a big fan of Louis Rossman, he spoke of you often in his repair vids. Amazing how you are his polar opposite on flux usage ! lol 😂 he uses all of the flux, and you use it like it costs money ! 😅 anyway appreciate your video, love watching things fixed the correct way.
Thanks for the sub ( now to get the other million+ subs from Rossmann ;) ) We certainly have our opposite matters; * Northern --- Southern hemispheres * Loud --- Quiet * Lots of flux --- not a lot of flux * Hounded by NY Govt --- Not. * Refuses to update software --- writes software updates Of course, we've got a lot of similarities, most important we're both carers of cats :D
I can remember when Louis was doing repair work he sometimes said '1 micropaul of flux' and flooded the whole area with flux 🤣 Now I see where it came from
@@ZakHooiTM Interestingly that comes from "Paul Stanislawski" who worked with Rossmann (the military dude with nearly bald head who Louis gave the difficult jobs), but we both have the same first name and we both prefer to use as little flux as possible :D
Almost never have jobs like that coming in any more. Even at substantially reduced repair costs the Magsafe1/2 series rarely shows up now, maybe the occasional A1466 and those mostly now seem to either be a dead cap or a failed CPU IMC :(
Hi Paul, thank very much for this very intersting video i learn more...can u explain how do you create the dark dome for the thermal camera? i don't undertand.. i have a TTI HT-18+ suggest to implente with this.. thanks for sharing with us your great experience bye Francesco Timpano from Flornce Italy
Hi paul, what wick was that you are using at 15:00, it looks very sturdy , when i try with the hot air it just spreads it around , i think its my wick, also could you tell me which leaded solder you use, thankyou.
Try this stuff - I bought a 100' roll, initially expensive but it'll take a while before I run out. www.amazon.com/No-Clean-Blue-4-Braid-AS/dp/B001UHQRD6
I've written a set of functions for the task within the schematic engine. You can use "classic text search" that can help too but there's a lot of noise in the runes of those schematics.
Hello John, The software is FlexBV5 from pldaniels.com/flexbv5 Yes, genuine Australian software grown right here in this lab. pldaniels.com/flexbv5/manual/flexbv-manual.html#part-finder
19:22 Could you elaborate as to why Lead-Free solder is 'bad' \ 'worse' then Leaded solder? is there no Lead-Free solder brand that is of the same quality and strength?
The nice thing about leaded solder is that it has the lower melting point ( but still high enough for practical purposes ) and more importantly it's quite a lot more durable & ductile for this purpose than a lot of lead free solders. Lead free SAC305 is a fairly nice replacement though still has a higher temperature ( ~220'C vs 183'C for 63/37 leaded ).
Got to rub them really quickly for that friction to set fire on thing - watch "Primative Technology" dude for an excellent demonstration of setting fire to things with lots of rapid friction. ( www.youtube.com/@primitivetechnology9550 )
I can appreciate why they do it, though I do wish there was at least *some* sort of pathway to remove the lock; albiet I suspect Apple go along with the concept of "Any valid pathway is a pathway for future abuse". At least for now it does mean there's a supply of quality donors at lower prices.
Another very nice precise power short circuit diagnostic and clearing the issue with replacement parts. Great work Paul
Thanks for that. Hoping to get more edited videos out too.
You are such a legend Paul. Thanks for teaching me things I couldn’t learn on my own.
Appreciate you watching!
Low V and Low I injection is the safest way to go, Nice work Paul.
Thanks 👍 Goes along with my preferred mantra of repairs - "First do no harm", Hippocratic as it were.
You know that you are a good tech when your common sense overrides your ocd for cleanliness. Another great fix.
Always a fine line between meticulous and "just get it done".
The blanket is new. Without that it seems like this would have been a shot in the dark situation. I like how you figured out whether something was a reflection or not.
Definitely. I need a better one, or an actual "dark box" to put over everything ( a bit tricky with the tripod on the iR cam, I'll need to create something better perhaps ). Once you block out all the external influences you'll then start picking up reflections of the heat from the camera itself ).
What a craftsman. Incredible work.
Thank you :)
I didn't know that trick with the thermal camera covering the light so the thermal sensor can find this soft short circuit. thanks for sharing
Thanks for the video from the future. Mac booted to 12am on the 27th of August and I am watching from 10am on 26th of October.
Aaah, but the year was 1980 ;)
Hmm have to try the blanket method sometime. Thanks for the tip Paul!
As you can see you still do get some transmission through it, and then more amusingly, as things become darker, you start getting reflections of the iR from the camera itself.
That fast forward effect was cooler than shirt! Nice!
Excellent and thank you for sharing. Informative learning for me.
Glad it was helpful!
Another board saved not going to a landfill Paul saves the day :-)
Correct, I want my bank balance being filled up, not the waste dumps :)
@@pldaniels haha 😂
Thanks for the amazing video Paul.
Appreciate the comment (helps with YT too).
Sweet. Thanks Paul 👍
Semi-simple repairs pays the rent and keeps the lights on. The capacitors seem to be the Achilles heel on these computers. Thanks for the video Paul. 🍻
PS. You could mark your working doner board and order more of those ICs; one for the board and a few extra for stock.
I could mark the donor, yes, but overall it's easier/saner to have another working board and accept the sacrafice of the one (I do at least have two other working units of that specific board).
Thank you very much for this very interesting video, kind regards Franco
Cheers, and thanks for watching.
Its good you are careful with your voltage injection. We aren't allowed to use voltage injection, probably because the instruments we are repairing reach well into the six figures. We aren't even allowed to use hot air unless it is BGA. The way I would have been required to find that short is to pull the old trusty Agilent 4338B Milliohm meter along with the 16338A Test lead kit which uses four wire kelvin connectors. Probably would have ended up pulling the cap closest to the IC as you did since it would have the lowest resistance, using JBC tweezers not hot air, unless I could get a needle tip probe on an outer ball which can be doable with the right high quality probes. Only then use hot air with the proper sized square nozzel and a preheater if nothing critical nearby, otherwise its off to the CNC machine. Then reball new chip with low melt. If you got it wrong and the chip is fine, expect to speak with the boss.
Definitely a different approach with such delicate/expensive gear.
I have got a 4-wire meter here, one I designed & built many years ago (PLD-LOM7) but the biggest pain I had with it is finding decent 4-wire probes that didn't resemble clothes pegs; I've come up with some alternatives more akin to normal probes but to date I've not really *really* needed it... yet. Downside of my idea of suitable replacements is that you still ultimately have some offset that you need to actively null out proportionally to current ( even if it's 1mA )
I noticed that every repair technician blast every component with incredible amount of heat or soldering it really hot. Do you ever killed any component because of the heat or a repaired computer came back with the same problem after the repair? You removed a cap in 5 sec and put a new one in 12 sec, removed a controller chip in 8 sec and put a new back in 13 sec with a bit of preheating, don't you find it a bit fast? The steepest reflow curve 3°C/sec, yours is 29°C/sec for removal and 18°C/sec soldering back (calculated with 20°C room temp).
Ones I killed a buck converter with too much heat, but I didn't know at the time, the symptoms was that the power supply always killed its schottky diode which seemed impossible because both the fuse and the power supply which fed the buck converter was rated lower than the diode. After the 3rd diode replacement I realized the controller will be the problem, preplaced it now on a preheated board and the problem is gone. After this accident I always preheat every board.
People say with quick heating you don't stress that much the surrounding components, but I don't buy that. I watched I guy blowing up a capacitor while replacing a DrMOS, maybe 1 capacitor is dead, but other 5-6 lifespan shortened 50% because of the extreme heat. I don't say people should follow the refow curve, but what I see in these videos I call it extreme stress on the components.
For smaller components I preheat the board from the other side about 1 minute with hot air station, for larger componets I preheat the board about 3 minutes with preheater. Maybe I just waste my time preheating, but I'm glad that non of the motherboards came back what I fixed.
I think better to check on live-videos where there's no editing. I remove a lot of the rework time on edited videos because it's just me sitting there waving the wand around waiting. In a lot of cases I tend to use the want to warm the board area up around first if it's a feasible option. There's also the situation where the board is already hot from prior rework so the next application will be faster.
Definitely some parts do not tolerate the rapid ramp well ( hence I tend to cover a lot of things if I can, especially tantalum caps ).
I would say it's a balance of risks, which is true of most of the things done with reworking that you see; sometimes it's effectively impractical from a time or setup situation to preheat; however if you do have the time then by all means.
One risk with reworking and preheating is that you can inadvertedly induce solder-squeezage out of underfilled neighbours ( particularly with the NANDs and CD32xx chips on these boards ).
Your concerns are merited, and at the end of the day it does come down to the balance of risks and what a shop is willing to accept & support.
Hey Paul, great repair. I'm having similar issue with leakage on a smd qfn switch. I think you partially answered my question but do you think that the root cause here is internal to the silicon of the ic, not some corrosion under the chip (like between the balls) where it's impossible to see? You mentioned that heat is the main enemy of these chips and I agree, I think that during initial board fab, the lead free temp requirements comes too close to the death from heat range and it causes latent damage to the silicon so that it fails eventually for leakage/resistive short.
Unfortunately in this instance I didn't check the resistance before/after the cleaning process. I think it's fair to say *sometimes* it's the material on the PCB causing conductivity between pads/balls. In this video the junk was between PPBUS_AON, 3V8_AON and GND so the 17R could have likely been purely the liquid agent; if there was enough conduction between PP and 3V8 it could have triggered the sensing.
I could go back and check that part, since I keep those for each job.
Hello from America! Just subscribed. Always a big fan of Louis Rossman, he spoke of you often in his repair vids. Amazing how you are his polar opposite on flux usage ! lol 😂 he uses all of the flux, and you use it like it costs money ! 😅 anyway appreciate your video, love watching things fixed the correct way.
Thanks for the sub ( now to get the other million+ subs from Rossmann ;) )
We certainly have our opposite matters;
* Northern --- Southern hemispheres
* Loud --- Quiet
* Lots of flux --- not a lot of flux
* Hounded by NY Govt --- Not.
* Refuses to update software --- writes software updates
Of course, we've got a lot of similarities, most important we're both carers of cats :D
@@pldaniels 😂 👍🏻👍🏻
I can remember when Louis was doing repair work he sometimes said '1 micropaul of flux' and flooded the whole area with flux 🤣
Now I see where it came from
@@ZakHooiTM Interestingly that comes from "Paul Stanislawski" who worked with Rossmann (the military dude with nearly bald head who Louis gave the difficult jobs), but we both have the same first name and we both prefer to use as little flux as possible :D
Very good job .my best regards
Appreciate you taking the time to watch!
Enjoyed the video. Could I ask what hot air station are you using?
Atten 862 currently ( see the HotAir Bidet video th-cam.com/video/2FXDj8yconk/w-d-xo.html )
Didn't ow about the part find feature . Definitely handy. Although I don't have many donor boards lol
Usually if you have one or two of the same "generation" it goes a long way towards covering your requirements
Thanks! Side question: Where are all the SMC failures? I feel like back in the day, you would reball every other video. Change of hardware production?
Almost never have jobs like that coming in any more. Even at substantially reduced repair costs the Magsafe1/2 series rarely shows up now, maybe the occasional A1466 and those mostly now seem to either be a dead cap or a failed CPU IMC :(
Dust fest😂…I love it….great video as always thank Paul
It's a dust-filled warzone in there!
Hi Paul, thank very much for this very intersting video i learn more...can u explain how do you create the dark dome for the thermal camera? i don't undertand.. i have a TTI HT-18+ suggest to implente with this.. thanks for sharing with us your great experience bye Francesco Timpano from Flornce Italy
Hello and welcome back. The dark-dome is simply a thick towel or heavy blanket, very simple :D
Hi paul, what wick was that you are using at 15:00, it looks very sturdy , when i try with the hot air it just spreads it around , i think its my wick, also could you tell me which leaded solder you use, thankyou.
Try this stuff - I bought a 100' roll, initially expensive but it'll take a while before I run out.
www.amazon.com/No-Clean-Blue-4-Braid-AS/dp/B001UHQRD6
@pldaniels do you use a dedicated library of functions for pdf search or just classic text search ?
I've written a set of functions for the task within the schematic engine. You can use "classic text search" that can help too but there's a lot of noise in the runes of those schematics.
G,day from Sydney.
Could I ask the name of the company search engine you used to find the electronic components?
🌏🇦🇺
Hello John,
The software is FlexBV5 from pldaniels.com/flexbv5
Yes, genuine Australian software grown right here in this lab.
pldaniels.com/flexbv5/manual/flexbv-manual.html#part-finder
19:22 Could you elaborate as to why Lead-Free solder is 'bad' \ 'worse' then Leaded solder?
is there no Lead-Free solder brand that is of the same quality and strength?
23:30 is that it?
The nice thing about leaded solder is that it has the lower melting point ( but still high enough for practical purposes ) and more importantly it's quite a lot more durable & ductile for this purpose than a lot of lead free solders. Lead free SAC305 is a fairly nice replacement though still has a higher temperature ( ~220'C vs 183'C for 63/37 leaded ).
Has YT fixed the bot problem?
Tend to keep an eye on your recent videos' comments section but there's basically never any these days
Has been fairly quiet lately, thankfully.
@@pldaniels Let's hope I didn't jinx it 😂
👏
1/17W = about the heat 🔥 generated by one gnats ball sac for those playing along at home 🤣
Got to rub them really quickly for that friction to set fire on thing - watch "Primative Technology" dude for an excellent demonstration of setting fire to things with lots of rapid friction. ( www.youtube.com/@primitivetechnology9550 )
That is for sure. The blanket is the only way to have been able to see that tiny bit of heat.
iCloud locking is a big joke. Apple virtually made thousands of absolutely fine working machines inaccessible for "security reasons".
I can appreciate why they do it, though I do wish there was at least *some* sort of pathway to remove the lock; albiet I suspect Apple go along with the concept of "Any valid pathway is a pathway for future abuse".
At least for now it does mean there's a supply of quality donors at lower prices.
🖐👍