Hello, I am trying to replace a broken chip in a motherboard (MacBook Air), I have the replacement, that looks something like this:


As you can see, not many pins, not the hardest to remove apparently.

I've been soldering with a soldering iron for many years and I am quite good at it, of course I can always learn from better technicians but I've repaired hundreds of things with it. However, with a soldering heat gun I have very few experience with SMD soldering. I have this heat gun Atten ST-8800D 800W:

I know it's not expensive but it's neither entry level, I have 8 nozzles, some angled, and I've used it a lot to heat up things with glue, open in ear headphones and many other devices with stubborn glue that you can't disassemble otherwise, also to remove stickers correctly, and many other things, some precise and delicate, but, as I've said, I've barely used it to SMD soldering.

The first thing I did yesterday and today was practice a little bit with broken boards of similar specs apparently (number of layers, etc...), I've practiced on a broken SSD board, broken HDD board, etc...

I normally set the temp to 380ºC and air speed to 50%, and after a while, I can remove fairly well some capacitors SMD and resistors and other components. Then I upgraded to a harder target like a small square chip, and realized to do it "quick" and better I had to put the heat gun to 420ºC. After 2 minutes or so, I was able to remove it.

Then, I tried to remove a big memory chip, way bigger than the one I am trying to desolder, with dozens of pins under it, and had to put 450ºC and 60% air speed to remove it, and took 3-4 minutes to remove it. I was using flux but realized for the easier parts it's not even needed to remove them. I even tried to remove another big memory chip without flux and I was able to do it.

SO... about 1h ago I finally dared to work on the MacBook Air board to replace the chip. First, in order to avoid kicking out some tiny components real close to the chip, I used aluminum foil (or tin foil) all around the chip, so only the chip gets heated up and I don't melt the surrounding area.


The difficulty was between 380ºC and 420ºC, that was my guess, so I put it in 400ºC, 50% speed. After 3-4 minutes on it, with flux, nothing, like a rock. I noticed the CPU heatsink was getting really hot to touch (maybe 50ºC or so?), so I thought about "OKAY, may be the chip is grounded and I need more heat to get it to heat up high enough I can remove it. Tried 450ºC at 50% speed fan, but nothing, after 5 min, same rock.

I then removed the heatsink (it was in the other side) to avoid heat going there, and tried again. Nothing.

Already getting mad, I put full blast 100% and top 500ºC temperature. And freaking nothing, 5 min, 8, 10 min... it just won't come off... I don't understand how I was able to remove away bigger chip within 4 min at 450ºC and 60% speed, and I am maxing out my heat gun and still can't remove this stupid tiny chip.

I am kind of desperate, I don't know what to do. One thing is to remove all my protections so I can heat up better all the area around the chip, but this is very dangerous as I am pretty sure some component will go away flying.

How is it possible I can't remove that simple chip with full blast 800W on it?

I know some people preheat the boards, but I don't have an oven to do that (I own one for food only), and also I can't believe I need to preheat the board if I am using 500ºC at 100% speed for minutes.

Also, when removing these hidden pins chips... is it the whole chip reaching melting temperature of the solder so you can remove it, or is it the motherboard the one heating up so much it melts the solder of the pins and the chip is released?
I am asking because when you are aiming to the chip with the nozzle, should you stay on top, aim at an angle from the sides so heat goes below the chip, don't heat the chip itself but the sides only (then the center how would heat up?)...

Here's a video of the exact board I own:

What he does is more or less what I was doing while practicing, same times and temps and all. I don't know, maybe the protection is what's messing with my results...

 

I don't have an answer, but I'm almost certain that large metal pad is also soldered to the board, and what you're struggling with.

 

If the chip went really bad, it could have essentially welded itself to the pad and potentially to some additional layers below it. No amount of heat is likely to allow it to be removed in that case. Is there any other damage to the board near there?

 

Have never tried this, but what about using a Dremel tool to grind the chip down getting close to the solder pads, the desoldering what is left. Anyone done this?

 

Perhaps you could heat the chip until the solder underneath it melts, and you can then remove it.

 

Have never tried this, but what about using a Dremel tool to grind the chip down getting close to the solder pads, the desoldering what is left. Anyone done this?

I have done this a couple of times and it can work very good. The problem though is reflowing the replacement device, as the same constraints that complicate removal, complicate replacement. Overheating a 'bad' device to remove is not a problem, but that's not available to reattach a replacement device

 

The "problem" is this chip and the many layered of the board is designed to get heat out fast .
I bet despite the 800 watts , the rest of the board is a lot lot cooler.
In production , we use pre board heating . In it's simplest a hot plate . The complete board is soaked to say 125 degree c, and only when it's all that hot, whilst still heating the board, is the hot air to the chip in question put on.
Then once chip is off, the board is left to cool down slowly ,

 

If the chip went really bad, it could have essentially welded itself to the pad and potentially to some additional layers below it. No amount of heat is likely to allow it to be removed in that case. Is there any other damage to the board near there?

No, the chip is damaged via software, not liquid damage or anything like that.

 

Have never tried this, but what about using a Dremel tool to grind the chip down getting close to the solder pads, the desoldering what is left. Anyone done this?

Hhahahaha, this really looks like a process that fits my profile pic!!

Believe me, when I was full blasting this chip from hell at 500ºC at 100% fan speed to make it feel home, after 10 mins of nothing accomplished I really started to think about radials, hammers, files and many other brute force tools.

But... one, this is a working motherboard except for the chip, so I don't want to perform such a Rambo thing. Two, I've seen videos of techs desoldering this with 400 something temp, so even with my inexpert hands, at 500ºC, I should be able to do it properly. Three, I have to solder the new chip there, so I must be able to do the reverse process, and I don't feel like I will be able to do it properly if I can't even do the opposite process correctly.

 

Perhaps you could heat the chip until the solder underneath it melts, and you can then remove it.

But... that's exactly what I am trying now, and after 10 min full blast 500ºC at 100% fan speed, it's still rock solid. Hence, the creation of this thread.

 

The "problem" is this chip and the many layered of the board is designed to get heat out fast .
I bet despite the 800 watts , the rest of the board is a lot lot cooler.
In production , we use pre board heating . In it's simplest a hot plate . The complete board is soaked to say 125 degree c, and only when it's all that hot, whilst still heating the board, is the hot air to the chip in question put on.
Then once chip is off, the board is left to cool down slowly ,

So what can I do?
I don't have a preheat oven and I don't think I can buy one just for this. Is the aluminum foil making the process a lot harder?
Should I remove it and try it without any protection (I really don't want to risk knocking off one tiny small component, it would be over for good)

 

So what can I do?
I don't have a preheat oven and I don't think I can buy one just for this. Is the aluminum foil making the process a lot harder?
Should I remove it and try it without any protection (I really don't want to risk knocking off one tiny small component, it would be over for good)

The cheats way to pre heat , is to use a second heater.
Most common is one of those hot air guns used for paint stripping , and some creative use of tin foil , kapton tape and stands. The paint stripper guns have a much higher air flow , at a lower temperature than your chip hot air gun
Ensure not to over heat the board , some thermocouples are normal ,taped to bits of the board , but you could wing it !
Leave to soak till all nice and toasty , then use top heat gun to take off chip. The foil on the top is ok , it localises the local heating .

 

The cheats way to pre heat , is to use a second heater.
Most common is one of those hot air guns used for paint stripping , and some creative use of tin foil , kapton tape and stands. The paint stripper guns have a much higher air flow , at a lower temperature than your chip hot air gun
Ensure not to over heat the board , some thermocouples are normal ,taped to bits of the board , but you could wing it !
Leave to soak till all nice and toasty , then use top heat gun to take off chip. The foil on the top is ok , it localises the local heating .

OH WOW!!!

I actually own one of these:


Could I set it aiming upwards at the bottom of the board at 150ºC (safe for all the plastics and stuff there, right?) elevate/rise on top of the heat gun the board with some alligator arms like in my pic, and meanwhile heat up from the top the element with my Atten?
That could do the trick!?

 

OH WOW!!!

I actually own one of these:
View attachment 339525

Could I set it aiming at the bottom of the board at 150ºC (safe for all the plastics and stuff there, right?) elevate/rise on top of the heat gun aiming upwards the board with some alligator arms like in my pic, and meanwhile heat up from the top the element with my Atten?
That could do the trick!?

That's the idea,
Don't under do the pre heat, you want as much of the board to be in the 100 plus area , slow and gentle ore heat and post cool down is the key,
At least 30 mins is normal ! Set it going , low air flow , cool , then increase pre heat gently over time.
Ensure you can get the top heat on and lift the chip whilst still heating from below,
Practise the maneuver before you turn on the heat..

 

what do you mean 30 min?
I though about heating the board with the gun from below set at 150ºC from 5cm away, wait like 3-4 min, and then start working from the top with 450ºC or so with the Atten, while the gun below is still going. Then clean up a bit the chip area, put some leaded solder and solder the new chip.

Is this process wrong?

 

what do you mean 30 min?
I though about heating the board with the gun from below set at 150ºC from 5cm away, wait like 3-4 min, and then start working from the top with 450ºC or so with the Atten, while the gun below is still going. Then clean up a bit the chip area, put some leaded solder and solder the new chip.

Is this process wrong?

Slow and carefully pre heat and after chip removal cool down.
30 mons pre heat to around 100 / 125 is the way, anything at 800 degrees is the wro g way and will burn board or pop off a track / ball.
Treat the board like a best friend,

 

Slow and carefully pre heat and after chip removal cool down.
30 mons pre heat to around 100 / 125 is the way, anything at 800 degrees is the wro g way and will burn board or pop off a track / ball.
Treat the board like a best friend,

30 min sounds like I'm cooking dinner for Christmas, hahahaha. Isn't that way too much time?
I'm not sure about having the heat gun 30 min straight heating the board, it's really a lot of time to be holding it.

I know, I am picturing, that if a board is all at the max temp possible that respects the plastics, like 100ºC or so, then if I apply heat, it will spread way slower and concentrate more on the area I am working, which is what I want, but again, I am thinking about me holding the gun and heating the board for 30 min and sounds crazy.

I will let you know how I manage. And yes, I know I must do it with precision and care.

 

30 min sounds like I'm cooking dinner for Christmas, hahahaha. Isn't that way too much time?
I'm not sure about having the heat gun 30 min straight heating the board, it's really a lot of time to be holding it.

I know, I am picturing, that if a board is all at the max temp possible that respects the plastics, like 100ºC or so, then if I apply heat, it will spread way slower and concentrate more on the area I am working, which is what I want, but again, I am thinking about me holding the gun and heating the board for 30 min and sounds crazy.

I will let you know how I manage. And yes, I know I must do it with precision and care.

Yep, a good analogy,
That's why you need clamps, not holding ..
Slow and easy ..

 

Wow. W O W.

This is easily the most frustrated I've absolutely ever been about soldering. I want to smash this board into pieces so bad. Freaking 40 min with a big heat gun below at 150ºC heating the whole board, after 10 minutes, started to put full blast 500ºC at 100% fan speed from time to time to check. The whole board was hot to the touch, last 15 min was constant full blast 500ºC at 100% speed plus the below heat gun running all the time. I don't see technicians apply anywhere near the amount of heat and watts I'm using to remove this.

I really can absolutely NOT believe how in the freaking heavens that stupid chip is still rock solid to the motherboard, I don't understand anything, I wanna hammer the c*** out of this board. How is it handling freaking 500ºC right on top + 150º below for 40 min.

Did Apple use solder that melts at 700ºC or what???

I was 99% sure this trick would absolutely do it, yet I am still unable to remove it... next step is instead of using flux, use some liquid low melt solder, which sounds very dumb and dangerous.

 

Start back at square 1. How did you determine that this chip was bad? Did you use a heat imager (TIC) to see a short on this chip? At I stated before, it is possible that a solid short may have welded this IC to the board and NO amount of heat will release it. You can try using a soldering iron (Weller gun type) with the tip directly on the chip. This will apply heat only to the chip. If this also fails, this board can be marked as unrepairable.

-Dan