10
\$\begingroup\$

This is a bit of a fringe question, SE-wise, but I think EE is the best place to ask it.

Some acquaintances of mine had a problem with their laptop - the graphics card stopped working. They've received (from someone else) instructions as to how to "reflow" it.

Unfortunately I don't have the link to the text (and it's not in English anyway), but the relevant bit instructed to heat the card in a convection oven set at 180 C for 5 minutes.

I was asked about my opinion about this fix and answered that while it's conceivable that it might work (my only idea was that it could perhaps remove solder connection defects, since the board is old enough to be pre-RoHS), I advised strongly against performing the fix in an oven still used for food preparation.

To my dismay I recently learned that the acquaintances in question went ahead and performed the fix, in an oven used for home cooking no less. Interestingly enough, it worked, at least for now.

Therefore, I want to ask the following two-part question:

  1. What are the health hazards associated in performing this sort of rework in an oven used for preparing food, and how to deal with those dangers? I would prefer for the answers to concentrate on the possible contaminants left over in the oven, rather than, e.g. increased risk of catastrophic failure of the PCB's components.
  2. How can this sort of rework be effective in rendering the board operational again?

Preferably, I'd like an answer that addresses both parts, but for obvious reason I'll also be satisfied with a quick response on only the first one.

UPDATE: Thank you for the answers so far. Since the most popular ones present opposing opinions, and the number of the respective upvotes is comparable, I think I should wait a day or two before accepting either. This is in the hope that someone would be able to provide concrete data relating to the subject.

UPDATE 05.01.2013: I'm still going to leave the question without an accepted answer for some time. Seeing as how none of the answers have hard data to support them, I'm a bit apprehensive with going the either way. Sorry for that.

\$\endgroup\$
3
  • \$\begingroup\$ As you guessed, the fix is, in essence, using the convection oven as a reflow oven. I've had luck reviving my old nVidia GTX260 by sticking it in my oven for a couple of minutes. Before, the computer wouldn't POST, after the card worked just fine. People have had luck doing the same for the XBox360, where some of the BGA connections under the GPU/CPU crack due to board flex and high heat, and the board malfunctions due to loss of contact between the GPU/CPU and the mainboard. \$\endgroup\$ – Shamtam Jan 2 '13 at 17:37
  • \$\begingroup\$ I know people that use their reflow oven for baking pizza's ... \$\endgroup\$ – jippie Jan 2 '13 at 19:53
  • \$\begingroup\$ I would avoid using anything you'll later use for anything else. Why take the risk? I've had luck reflowing not with an oven but with a skillet. Results are very nice. I didn't cook anything on it though :D \$\endgroup\$ – Gustavo Litovsky Jan 2 '13 at 19:58
7
\$\begingroup\$

They are almost certainly not in any danger. Not any more than the fact that they handled the board and possibly ate something before washing their hands. Direct transfer probabilities are much higher than a double outgassing transfer, first to oven and from oven to food. In the oven of course you have the potential for outgassing, which then coats out on the inside of the oven. I doubt that there is a lot of this. On top of it, people typically preheat the oven which would spike the temperature up again beyond this "reflow" temp and this would liberate the bad stuff (if it was there) and carry it out of the oven. i.e. it would be baked out.

Boards are relatively clean, lead is not very volatile, there are VOC present but once heated they dissipate etc. you wouldn't want to use an oven that is exclusively used for this, but once in a while is most likely very very safe.

In fact the oven elements when new outgas on first use, after cleaning the oven there are chemicals left on surfaces that outgas. Those would be the best comparables and more concerning than the PCB.

\$\endgroup\$
2
  • 3
    \$\begingroup\$ +1 for going with a realistic response in direct opposition to the increasingly commonplace paranoid and politically correct blather everyone seems to spout. One gets tired of the "Ooooh, it's dangerous!" RoHS conform-speak. \$\endgroup\$ – Anindo Ghosh Jan 3 '13 at 2:38
  • \$\begingroup\$ @AnindoGhosh It's good to point out that this is still somewhat dangerous/risky. Not that it's stopped me from using it to fix game boy games without an iron handy. \$\endgroup\$ – Wyatt Ward Sep 20 '16 at 22:53
3
\$\begingroup\$
  1. Who knows what kind of junk is released from plastics, metals, solder, etc. I wouldn't do it in my oven, but hey...

  2. It's a ghetto way of reflowing the solder balls on BGA chips. Technically, there are "reflow profiles" that should be followed to melt the solder properly and all that jazz. A reflow profile is a description of the way a temperature should ramp up or down over a certain period of time. You can Google around for more info on that, but a regular kitchen oven is not going to be able to follow any sort of reflow profile correctly. It should not be considered a long-term fix. There is still a potential the same exact problem will occur again.

\$\endgroup\$
3
  • 1
    \$\begingroup\$ The temperature profile is there to bake out the last bit of moisture from the parts before heating to reflow temperatures. If this isn't done, there is a risk of plastic package damage due to trapped moisture boiling off. I doubt many home ovens will do a good job of heating to 180-200 F without going over 212. I also doubt most home ovens can control temperature that evenly across the oven. If you're going to try it, calibrate your oven first by placing thermometers around where you'll put the board, and measure the real temperature vs. the set point vs. time. \$\endgroup\$ – Mike DeSimone Jan 2 '13 at 17:54
  • 1
    \$\begingroup\$ @MikeDeSimone Some components heat up slower than others, and the temp profile also allows time for those slow components. Also, the temp profile allows for a more controlled and even cool-down to allow for less thermal stress. What you say is true, but there are more factors than just that. \$\endgroup\$ – user3624 Jan 2 '13 at 19:37
  • \$\begingroup\$ @DavidKessner: Yeah, but they only give me 600 characters. ^_- \$\endgroup\$ – Mike DeSimone Jan 3 '13 at 0:23
2
\$\begingroup\$

I would not do it. Maybe I'm paranoid, but the risk of exposing my dinner to lead and other nasty stuff is too high. Doing it once might be OK, if you are careful, but I deal with electronics all day, every day, so I'm more careful about such things.

As for the effectiveness... The process of reflowing a PCB is normally tightly controlled. You don't want it too cold, or to hot, or do it too fast or too slow. Too fast or too cold will result in the PCB not reflowing properly. Doing it too slow or too hot can damage the PCB and chips. And ideally you want some flux on the pins/balls that need reflowing.

Reflowing in your home oven does not control for either time or temp. It might work. Or it might not. Or you could make things worse. If the choices were to attempt a reflow or throw the PCB away then maybe it is worth it to try to reflow it. Worst case you'd throw it away anyway-- which you would have done anyway.

Again, I would not do it in my oven. But if I did, here is how I would do it:

  1. Get a cookie sheet and cover it in aluminum foil. Cover the insides of your oven too, as much as possible.
  2. Preheat the oven, with the cookie sheet in the oven too.
  3. Attach metal stand-offs on the PCB.
  4. Remove anything from the PCB that you can. CPU, heat sinks, etc.
  5. Place the PCB on the cookie sheet. The standoffs will keep the PCB itself from touching anything.
  6. When the time is up, turn off the oven and CAREFULLY open the door. Do not shake, bump, or move the PCB. Just let the oven, PCB, and cookie sheet cool with the oven door open.
  7. Thrown away the aluminum foil.

This method should keep your oven as uncontaminated as possible, while keeping the odds of a successful reflow as high as possible. With practice, this might work well. Without practice, your odds of success are probably less than 50%.

\$\endgroup\$
4
  • \$\begingroup\$ What is a 'cookie sheet'? Also be careful not to cover any openings in the oven that are used for airflow / heating / ... \$\endgroup\$ – jippie Jan 2 '13 at 19:56
  • 1
    \$\begingroup\$ @jippie A cookie sheet is a flat thing that you bake cookies on. :) \$\endgroup\$ – user3624 Jan 2 '13 at 22:01
  • \$\begingroup\$ A cookie sheet is a large rectangular metal pan (usually aluminum) with very little (if any) depth (1 cm or less). Heat can sometimes cause a cookie sheet to warp in the oven, which would be really bad for reflow, since it could send some parts flying. \$\endgroup\$ – Mike DeSimone Jan 3 '13 at 0:25
  • \$\begingroup\$ A problem with step 1 in the case of gas ovens is that you could cover the vents used to get the heated air in and out of the oven. \$\endgroup\$ – Mike DeSimone Jan 3 '13 at 0:27
0
\$\begingroup\$

Using the home oven is a terrible idea. Health concerns aside, please see documents from IC manufacturers, e.g. https://aerospace.honeywell.com/~/media/Images/Plymouth%20Website%20PDFs/Magnetic%20Sensors/Application%20Notes/AN216_Mounting_Tips_for_LCC_Magnetic_Sensors.ashx

Excerpt:

Most LCC packages have no special requirements beyond normal procedures for attaching SMT components to printed circuit boards. The exception to this process is Honeywell HMC products that have ceramic or FR4 substrate packages with epoxy top encapsulation. These package designs use two solder types with differing reflow temperatures. Inside these packages, a high-temp reflow solder is used that reflows at 225°C and above to make internal circuit connections. On the package outside, low-temp solder is recommended with a reflow temp range from 180 to 210°C. Three heating zones are defined in SMT reflow soldering process; the preheating zone, the soaking zone, and the reflow zone. The preheating zone includes the soaking zone, and nominally ranges from 2 to 4 minutes depending on temperature rise to arrive in the 160°C to 180°C soaking plateau to active the flux and remove any remaining moisture in the assembly. Preheat rise times must not exceed 3°C per second to avoid moisture and mechanical stresses that result in “popcorning” the package encapsulation.

So, unless you're damn sure it's staying uniformly between 200C-220C and not heating faster than the specified rate, you risk damaging the type of IC being referred to. If you look at a reflow oven, they provide a variety of different profiles that are geared towards different IC types + solder (leaded vs lead-free, etc)... do you really want to disregard all of this and throw it in the oven with your pizza, risking health of both you and the silicon?

\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.