question about reflow- it it ever a long term solution?

[kq702]

Hi I did a reflow on an HP laptop a while back- it was a DV6000 series. The original problem was the internet stopped working- turns out the south bridge chipset was causing the issue (its a very common issue I found out after searching online). Some people said they did a reflow on the south bridge chipset and that fixed it, but lots of people warned that it will not last. They said the only true solution was to reball with lead solder. So I gave the reflow a try- it worked but they were right it only lasted about a month.


So I recently had a customer bring me a mac book pro (2008 model) and the problem was soon as you hit the power button it would turn on and then go right off after 2 seconds. I took the machine apart and saw that the vents were fully clogged with dust. It obviously overheated. I purchased a replacment board on ebay and they said it was a factory refurb board. So I got it and I could tell it was not a factory refurb, I called the guy and he told me it was pulled from a working unit and he reflowed it himself, he said he does it with all the boards he sells, and then sends them to Coremac (a mac repair center) for inspection and they give him a pass certificut. So that gave me an idea- I tried to reflow the chips on the broken board myself. I reflowed the following three chips: 1. GPU,2. CPU, and 3. Northbridge chipsets. I put the board back in after the reflow, and the mac book pro laptop powers up!

My question is will this problem come up again on the mac book pro in a month or so, like the wifi problem came back on the HP laptop? If the problem is likley to return, wouldnt the guy who reflows them all be doing more harm then good? Or was that issue of it only being a short term fix apply to just the south bridge chipset on the HP laptop?

[tron]

Re: question about reflow- it it ever a long term solution?

short term fix if no copper shim mod was used

[momaka]

Re: question about reflow- it it ever a long term solution?

A proper reflow can last as much as the laptop did from the factory. If you're reflows are lasting only a month or so, you're probably not doing it right.
What reflow method are you using? Oven? Heat gun? Rework machine? Other??
Are you using flux?

[Th3_uN1Qu3]

Re: question about reflow- it it ever a long term solution?

A proper reflow can last as much as the laptop did from the factory.

Correct. Just as a bad reball can last just a couple months. It isn't really the method that counts, more than what you do afterwards. Most laptops with the bad nvidia chips also suffer from poor thermal design, and require cooling mods to reach reasonable operating temperatures. If you manage to keep them under 60-65C load, they are unlikely to fail again unless they receive physical shocks (as in, the laptop is dropped).

If you don't use flux then that could be the issue. The way i do is as follows: Put a thin layer of flux on one side of the chip, and using 200C and the thinnest nozzle on the hot air gun, i melt the flux and direct it under the chip. Once most of it has gone underneath, i repeat the process for the other sides. Then i take the nozzle off and reflow. I use 350C for 4 minutes, moving the hot air to distribute it uniformly, and never have the hot air gun closer than 5mm to the chip. If you see a thin layer of smoke coming from the flux then you're doing it right.

From the outside, it seems that flux is more likely to just slide around the sides than go under the chip, but i assure you that it does happen. Out of curiosity, i've pulled chipsets from dead motherboards, one with flux, one without. On the one i pulled with flux applied and melted beforehand, the board had traces of flux down to the very center of the BGA.

[marcellodj]

Re: question about reflow- it it ever a long term solution?

Whatever tecnique you use to blow the bga,hot air,irda,torch or vodoo, i think the most important aspects are:
temperature monitoring
time profiling
flux

i have machines reflowed 2 years ago working fine till today.my 2 cents,regards.

[momaka]

Re: question about reflow- it it ever a long term solution?

The way i do is as follows: Put a thin layer of flux on one side of the chip, and using 200C and the thinnest nozzle on the hot air gun, i melt the flux and direct it under the chip. Once most of it has gone underneath, i repeat the process for the other sides. Then i take the nozzle off and reflow. I use 350C for 4 minutes, moving the hot air to distribute it uniformly, and never have the hot air gun closer than 5mm to the chip. If you see a thin layer of smoke coming from the flux then you're doing it right.

You heat the chip from the top?!! Well, there's nothing wrong with that actually, but in my experience you should to heat the board from the bottom first since it heats up much slower than the BGA chip. Heating the BGA chip from the top can kill the chip in some cases (been there, done that and I've seen it happen to some people at work too) as well as wear the board over time and cause delamination. Bridges are more likely to form too.

Heating the board from the bottom, on the other hand, can never kill the chip unless you use some extreme temperature (but you will likely burn the board as well).

What I do is (at least if I'm reflowing with a heat gun and don't have any fancy machines or equipment that we do at work - this is applicable to hot air as well):
1) slowly heat board from the bottom under the BGA chip to be reflowed. If you have a thermometer, when the board reaches 140-150C on the top side you can stop heating.
2) put flux around the chip. Note: if you heat the board from the bottom like I do in step one above, the flux will flow under the chip by itself - no need to force in under there.
3) continue heating the board for a minute or two until flux has activated well and the top of the board is at 140-150C again
4) move hot air/heat gun to the top of the board and heat the BGA chip. Heat until the top of the board reaches about 195-200C (if leaded solder is used) or 220-230C (if lead-free solder is used).

If you don't have a thermometer, then observe the flux as Th3_uN1Qu3 mentioned - when you see thin smoke coming from it, continue heating/reflowing for a few more minutes.
For step 1), I usually put my finger on the BGA chip while heating from the bottom. When the chip begins to burn my finger, I know the board has reached at least 70C. After that I heat for 1 to 3 more minutes, depending on the board's size. Then I move to step 2.

[retiredcaps]

Re: question about reflow- it it ever a long term solution?

For step 1), I usually put my finger on the BGA chip while heating from the bottom. When the chip begins to burn my finger, I know the board has reached at least 70C.

I'm thinking about getting a Fluke 62 or equivalent sometime in the future. I have seen the cheap knockoffs, but unsure of their quality and safety.

It looks like the lowest price used Fluke 62 on ebay is just under $50. I know some multimeters have a built IR thermometer too.

[Th3_uN1Qu3]

Re: question about reflow- it it ever a long term solution?

You heat the chip from the top?!! Well, there's nothing wrong with that actually, but in my experience you should to heat the board from the bottom first since it heats up much slower than the BGA chip.

I'm thinking of making me a bottom plate heater as well, but i've yet to kill a chip completely by just using the heat gun on the top. If you bring the hot air in slowly, the board will heat up as well and there's no risk of delamination or such. The ones that could not be restored by reflow, i sent them to be reballed and they worked okay. I sometimes monitor the temperature on the bottom of the board, sometimes not.

[lucco85]

Re: question about reflow- it it ever a long term solution?

A proper reflow can last as much as the laptop did from the factory. If you're reflows are lasting only a month or so, you're probably not doing it right.

I would not agree with this.
If the problem is bad solder reflow will help.
If, however, comes a faulty chip, just replacement of the chip can result in a reliable repair.
Unfortunately the defective chips are the most common problem.

[kq702]

Re: question about reflow- it it ever a long term solution?

Thanks for everyone for the feedback and suggestions. What type of flux do you suggest using? I have some Kester 951 no clean flux, would that type be good? Or would a more tacky flux be suggested?

[fddizon]

Re: question about reflow- it it ever a long term solution?

Thanks for everyone for the feedback and suggestions. What type of flux do you suggest using? I have some Kester 951 no clean flux, would that type be good? Or would a more tacky flux be suggested?

that's a good flux to use.

i can't find that type of flux here

[Lumberjack777]

Re: question about reflow- it it ever a long term solution?

The design problem with the HP DV series of laptops is that they used a thin piece of foam (or similar) between the GPU chip and the heatsink. This acted as an insulator, and not as a conductor. So when the GPU chip got very hot; it desoldered itself and pressed up against the foam cushion. Losing contact with the motherboard.

I was very interested in all of the suggestions of reflowing and the use of flux. I might try those in the future. Wow, some of you really superheat that GPU!

My method is to make a heat shield from several layers of aluminum foil. Then press it onto the GPU chip; and cut out the square to exactly expose the GPU. Then I set the heat gun on low and heat the chip for about 2 minutes, from a few inches away. Then I turn the heat gun to high and move it a little closer and move it in a circle around the GPU chip. Since I don't have any temperature monitoring equipment; I wait for what I call "the smell". The smell is the original flux inside the solder balls telling me that the solder has melted. I immediately turn off the heat gun. Press down on the GPU chip with an oven mitt. (one finger in the middle, and 4 fingers on the 4 corners). I hold that for a minute or two.

Then when I reassemble the laptop I cover the GPU with electrical tape. (not the dark middle, just the outsides with the little leads sticking up) Then I use a copper penny or 2 copper shims. (and a thin layer of thermal paste) on the GPU. Now what this does is actually apply slight pressure to the GPU. As it gets GPU gets hotter and hotter during normal use; the heatsink now pushes down slightly on the GPU. Instead of desoldering, we now have resoldering.

I once had a reflow that brought the HP laptop back to life. But the graphics had some slight problems. (I think I was chicken to heat the GPU too long) hahaha

So what I did was leave the laptop running (no power saving, full graphics) for 24 hours. And sure enough; when that GPU really heated up; it resoldered itself back into place due to the pressure of the copper shims.

Just my 2 cents worth. Actually, it was just one copper cent. Literally

[mattbrad2]

Re: question about reflow- it it ever a long term solution?

If I could go around and slap every youtube poster who posted a "penny" fix I would. I don't go a week without seeing one of these "fixes" and after about a week they die, only this time the GPU is now crushed and needs to be completely replaced. The thermal pad has little to do with the problem. It doesn't have the best thermal coefficient for sure but its better than air and in every possible way imaginable better than a penny. I've been doing this for longer than I can remember and have invested a ton of money into the proper equipment to re-work these boards the correct way. To me, there is only one true fix for these HP boards and its been proven time and time again.. reball it with leaded solder. Anything else and your merely buying yourself a little time until it quits booting again.

[momaka]

Re: question about reflow- it it ever a long term solution?

If I could go around and slap every youtube poster who posted a "penny" fix I would.

+1

Add to that all of those people who post "towel" fixes, or just tell you to turn your computer on and let it overheat until the chip "re-solders" itself.
- THE CHIP CAN NEVER REACH A HIGH ENOUGH TEMPERATURE TO RE-SOLDER ITSELF. EVER! (If it did, it will get destroyed - semiconductors don't do well with heat when they are powered on).

Those overheat-type "reflows" do nothing more than make the solder balls on the chip oxidize more and possibly form tiny solder wiskers - which in the right place may temporarily fix the problem. But the chip never actually reaches anywhere near a proper reflow temperature. So in the end, overheating your chip to "reflow" it will only do more damage - both to the chip and (if done enough times) the board. AVOID IT!

Unfortunately the defective chips are the most common problem.

In my experience, that's not what I have seen.
I guess it all depends on the device you're repairing. We've done over 100's of Xbox 360s at work. Most of them work fine again after a reflow (and for a long time too). Only a few don't - those are usually the ones with a dead GPU and a reball very rarely fixes them.

But if you've had many laptops with nVidia GeForce 6xxx, 7xxx, and 8xxx series GPUs, then I understand why you would blame the chips.

[lucco85]

Re: question about reflow- it it ever a long term solution?

In my experience, that's not what I have seen.
I guess it all depends on the device you're repairing. We've done over 100's of Xbox 360s at work. Most of them work fine again after a reflow (and for a long time too). Only a few don't - those are usually the ones with a dead GPU and a reball very rarely fixes them.

But if you've had many laptops with nVidia GeForce 6xxx, 7xxx, and 8xxx series GPUs, then I understand why you would blame the chips.

We are talking about the dv6000 which are generally defective chips, nvidia d / c 06-08 with no exceptions.
I think that we agree.
Reflow method is that I do not use because it is not reliable, but everyone has their own choice
For xbox, I do not know, so far I have not repaired them because rework station that I had been too weak (Achi ir6000).
I'll try it now.

[kq702]

Re: question about reflow- it it ever a long term solution?

that's a good flux to use.

i can't find that type of flux here

I couldn't find the Kester 951 no clean flux here locally so I ordered it from ebay. I got a needle dispenser bottle to dispense it too- its much cleaner then using a dropper. I just feel like it evaporates way too fast though, but I am new to this. I feel like by time the temp gets up all the flux has evaporated off. Is this normal?

[mattbrad2]

Re: question about reflow- it it ever a long term solution?

kq702 - don't worry, that Kester works like magic. You wont be disappointed.

[momaka]

Re: question about reflow- it it ever a long term solution?

Reflow method is that I do not use because it is not reliable, but everyone has their own choice

They can be quite okay if the temperature on the reflowed chip is kept to reasonable levels. I agree, though - lead-free solder will eventually crack again after many heat cycles since it's not as flexible as leaded solder.

For xbox, I do not know, so far I have not repaired them because rework station that I had been too weak (Achi ir6000).
I'll try it now.

Let me know how it goes.
For the old 360's (not the slims), it's the GPU that is at fault most often (like 95% of the time). The other 5% are the HANA (HDMI models only), then Southbridge and RAM. Almost never the CPU, though - I've seen maybe only 1 or 2. This is good, because there are a lot of caps around the CPU that you don't want to overheat.