Heat gun for GPU reflow

[mbsurfer2018]

Hello!

I'm a rookie repair guy and looking for a good heat gun for a GPU reflow on a Toshiba Satellite C55D-A5380.
Any suggestions will be great and my budget is $40 dollars, yes it has to be cheap. I have been searching on Amazon and ebay but the item description states that they are for removing plastic or paint. Most of them have an operating temperature between 50 - 650 C*. I just saw one on a computer repair site for 20 bucks and a operating temp of 300 to 600 C*.

Thank you and good day.

[stj]

Re: Heat gun for GPU reflow

no chance, you will popcorn the chip and delaminate the board.

spend several hundred on a reflow machine or leave gpu's alone.

[mbsurfer2018]

Re: Heat gun for GPU reflow

Thank you.
I'll keep looking.

[sam_sam_sam]

Re: Heat gun for GPU reflow

Any suggestions will be great and my budget is $40 dollars, yes it has to be cheap.

You are not going to find the right thing to use unless you are willing to up your budget to about $100.00

[momaka]

Re: Heat gun for GPU reflow

no chance, you will popcorn the chip and delaminate the board.

spend several hundred on a reflow machine or leave gpu's alone.

Incorrect.

I've done many reflows now with a heat gun. It's just more tricky and NOT like those YouTube videos show where you simply heat the GPU for several minutes and miracles happen.

Rather, if you do want to do heatgun reflows, you will need also need a type-K thermocouple meter. The "more proper" way to do a heatgun reflow is to have the board on some kind of a grill where you have access to both sides of the board - then you alternate heating from the top and from the bottom. This is one way to get more even temperature on the chip without popcorning it.

The "best" way to reflow with a heatgun: use your kitchen stove top (gas or electric with coiled heating element only) to heat the board from the bottom while you use the heatgun to heat the board from the top. And even with this setup, you will still need a type-K temperature thermometer. For less chance of popcorning and delamination, leave the board on the bottom heat only for about an hour so that the board is sitting around 100-105C. This will remove a good portion of the moisture out of it. Then heat to 140-150C, let the board soak for about half a minute at that temperature, then crank up the bottom heat slightly and turn on the top heat (heatgun) at this point. Start heating from the top and watch your temperature on the thermometer. Depending on the heatgun airflow and the thermocouple placement, you can get very different temperature readings. If the airflow on the heatgun is too high, it's possible for the thermometer to read a really high temperature when the heatgun is ON and much lower temperature when you pull it away. So to avoid that, pick a heatgun that doesn't have a very strong airflow, but still enough power to heat everything up.

All that being said, be prepared to scrap several boards as you learn how to use your setup... or to have several reflows come back as non-working (needing a reflow again) after a few weeks/months of use. Also remember that too little heat is better than too much heat - with not enough heat, the reflow either won't work or won't work for very long. Of course, that's better than a popcorned or completely shorted chip due to too much heat. So just keep that in mind.

My setup consists of a metal grill raised on two pieces of wood so that the board I'm working on sits about 2 inches / 5 cm above the kitchen's gas burner/stove. And my heatgun is a $10 cheapo "special" from Harbor Freight Tools. I used to use a different heat gun, but I found this one works better due to lower airflow (but similar heating temperatures apparently). And last but not least, I'm using a Lutron TM-902C type-K temperature thermometer with very thin Omega thermocouple (the stock thermocouple works okay too, but it's a bit slower with the temperature variations).

[sam_sam_sam]

Re: Heat gun for GPU reflow

What Harbor Freight heat gun did you use model number

My setup consists of a metal grill raised on two pieces of wood so that the board I'm working on sits about 2 inches / 5 cm above the kitchen's gas burner/stove. And my heatgun is a $10 cheapo "special" from Harbor Freight Tools. I used to use a different heat gun, but I found this one works better due to lower airflow (but similar heating temperatures apparently).

Is this the temperature sensor meter you are talking about

https://m.banggood.com/TM902C-Digita...CABEgJ3JfD_BwE

And last but not least, I'm using a Lutron TM-902C type-K temperature thermometer with very thin Omega thermocouple (the stock thermocouple works okay too, but it's a bit slower with the temperature variations).

Omega thermocouple Which one did you use

[flexy123]

Re: Heat gun for GPU reflow

No one is denying that heat gun reflows or "baking" in the oven will very likely give you worse results than a real, professional reflow/reball. That is, IF these DIY methods even work. (TBH, it sorta angers me when I see comments on Youtube videos by pros dissing these methods, because we know that these methods are only last resort, and we also know there is considerable risk. We don't need that being told my the "pros". Problem: Not everyone has reflow equipment, and for some things like old GPUs, it's often more a curiosity thing to try, especially if a card is dead so there is nothing to lose anyway. And...for things like a $25 GPU from 2014 or a soundcard, not everyone wants to get a $xxxx reflow bench or wants to spend hundreds on a professional repair. This is nonsense, when you can get replacements from ebay cheap anyway. For example, my currently flaky GTX 970 where I could simply get one from ebay for €130).

THAT BEING SAID, it is possible to fix GPUs, consoles, laprops etc. with a heat gun or "in the oven", although personally I prefer heat gun. (I mean there are countless people who fixed their Xbox RROD etc. that way). Is this "professional"? OF COURSE NOT. But ultimately, if it can achieve what you want, ie. get your laptop, GPU etc. to work again..and even if just for some months...I think it's an interesting alternative.

The problem, as I see it (and I am sure people like momaka know MUCH more about this than I do), is that heatguns vary in what heat they output, and that the window to reach correct temps on your component for reflow is very small. (Saying: you need to reach the exact temperatures, and +10 or +20 more can already mean you destroy components).

This is more or less impossible with your average $15 heat gun, UNLESS of course you have ways to measure temps on your boards with a sensor, and as you try to "emulate" a reflow profile as close as possible with the right duration and distance you hold the heat gun etc. Which you COULD sorta do after you did lots of trials and errors on some old boards so you can get some experience. See momaka's post above. Saying, it can be done as a "last resort" if you have experience doing it and you know what you're doing. (95% of Youtube videos, the people DON'T know what they're doing, when they "fix" things like consoles it's sheer luck)

[flexy123]

Re: Heat gun for GPU reflow

Would you recommend the TM-902C type-K or one of these laser handheld thermometers? Looking around on the usual cheap places, both are equally cheap to have. (about $10 or so).

[vinceroger69]

Re: Heat gun for GPU reflow

i have the TM-902C type-K and would reccomend it and have done a few reflows using momaka s method and all units reflowed are still working fine

[flexy123]

Re: Heat gun for GPU reflow

Cool, yes seems more practical since you need to hold the other type.

[momaka]

Re: Heat gun for GPU reflow

What Harbor Freight heat gun did you use model number

Not sure. But it's the one commonly sold for $14.99 (but really $8-9 with coupons.) Going by the HFT website, I think it's item #96289. It's HFT's own brand: Drill Master. Rated for 1500 Watts with two speeds/temperature settings (572°/1112°F)

Is this the temperature sensor meter you are talking about

https://m.banggood.com/TM902C-Digita...CABEgJ3JfD_BwE

That's the one.
They are usually cheaper on eBay, though. I got mine for $5 with free shipping. One of the best cheap China eBay purchases I made.

Omega thermocouple Which one did you use

Not sure. That one was given to me by a former boss at a repair shop I used to work at. Deal was I fix 3, I get to keep one. Basically, it's the type-K probes with very very thin wires. They are quite fragile, but also respond much faster to temperature changes.

TBH, it sorta angers me when I see comments on Youtube videos by pros dissing these methods, because we know that these methods are only last resort, and we also know there is considerable risk. We don't need that being told my the "pros". Problem: Not everyone has reflow equipment, and for some things like old GPUs, it's often more a curiosity thing to try, especially if a card is dead so there is nothing to lose anyway. And...for things like a $25 GPU from 2014 or a soundcard, not everyone wants to get a $xxxx reflow bench or wants to spend hundreds on a professional repair. This is nonsense, when you can get replacements from ebay cheap anyway.

+1000
Exactly how I feel too.

That said, I do cringe a bit when I see people doing the "towel tricks" on their old Xboxes. I mean, sure those things are dead and useless anyways. But the tower "trick" is so pointless, that you're better off just putting up the Xbox on your local Craigslist, Kijiji, or OLX for $5 or less. Maybe even offer it for free to someone who may be able to fix it a bit more properly (or just offer it as barter material).

So while I do encourage people to try the heatgun/burner and oven reflows, I don't encourage any of the towel tricks or clamp mods - those are a waste of time.

The problem, as I see it (and I am sure people like momaka know MUCH more about this than I do), is that heatguns vary in what heat they output, and that the window to reach correct temps on your component for reflow is very small. (Saying: you need to reach the exact temperatures, and +10 or +20 more can already mean you destroy components).

Yes, it's that, and also a bit how the heatgun dissipates its heat (i.e. airflow and power of the heat gun). With some heatguns, you will reach the goal temperature very quickly, but then maintaining that temperature for a few moments can be the hard part - i.e. you might hit 215-220C, but then the heatgun might raise the temperature further to 225-230C or more rather quickly, and that can damage the GPU. So that's why the temperature meter is a must with a heatgun/burner reflow.

With the oven method, you're relying on the oven's regulator accuracy, which is actually pretty accurate. The only thing to watch out for, especially on electric ovens, is when you first turn on the oven: this can make its heating elements run at very high power and overheat your board/GPU/components via IR heating before the oven even reaches your target temperature. Also, some ovens can overshoot their set temperature a little bit. So in the end, you can end up with a completely hosed reflow. The best way to avoid that is to test your oven with some piece of equipment you care even less about. Also, pre-heating the oven partially or even fully could possibly fix that issue too.

95% of Youtube videos, the people DON'T know what they're doing, when they "fix" things like consoles it's sheer luck)

True.
And also, 95% of the time, it's not the solder joints between the GPU chip and the board that fails - it's the solder joints between the GPU chip and GPU core. Thus, most reflows are a matter of sheer luck anyways. But if done right and if you have a chip that is known to reflow fairly well, you will probably end up with a good working chip again... at least for a while.

Would you recommend the TM-902C type-K or one of these laser handheld thermometers? Looking around on the usual cheap places, both are equally cheap to have. (about $10 or so).

Always TM-902C (or any other physical contact-based measuring device). The laser meters depends too much on the surface of the material you're trying to measure. For example, reflective surfaces always tend to show lower temperature than they really are - in some cases much MUCH lower. So because of this, I suggest avoiding them.

Meanwhile, contact-based meters like type-K don't care about the surface of the material. However, they can be sensitive to air movement, so avoid creating wind or too much moving air near type-K probes.

Cool, yes seems more practical since you need to hold the other type.

Exactly, that too.
With type-K proves, you just place them where they need to be (securely) and turn on your meter. No need to keep your hand(s) occupied holding a temperature meter.