Continue to Site

Welcome to our site!

Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

  • Welcome to our site! Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

Summer time = desoldering time (mass removal by heat gun)

Status
Not open for further replies.

Grossel

Well-Known Member
Hi forum.

This is just a resyme of experience about the use of heat gun to remove (all) components from PCB board as opposed to use soldering iron.

Summer is finally knocing on the door, most of last winter snow is gone.
So during the winter months I have collected some PCB boards from defect equipments and I've spend a couple of evenings after work to remove all components.

The pros using heat gun
  1. Once an area on the PCB is properly heated, it's way faster than using a soldering iron.
  2. It almost doesn't matter how many pins a component have - it get's off because a fairly big area is heated. Just be aware where the air flow goes and use smalles possible nozzle. It's even a fairly good method that also allow you to remove ATX power adapters from computer mainboards.

Reality check - now for the cons of using heat gun
  1. As the air gun heats up tha PCB itself, it takes longer time before the solder melts AND the overall heated area is bic compared to using a soldering iron, meaning the components often takes more heat. Not all components is resistent to long exposure of heat. In my experience espechialy film capacitors tend to break where the pins are mounted (those that looks like sleeping pillows) and the other variant the square shaped have the problem that the plastic enclosure start to melt before the component let go. Any component that does have plastic casing may not survive, being damaged due to deformation of plastic. For that kind of components, use soldering iron first and then use heat gun to other components.
  2. DANGEROUS SMOKE/GASS : Using heat gun to a PCB creates a hell of a lot of smoke compared to regular desoldering with soldering iron. This smell does really stick to any clothes, hair and skin, and even after a long shower, I cannot get completely rid of the smell. To somehow compensate for that, I never do this when there is no wind, nor when the wind is shifting. Also I stand over the PCB and try to always point the air gun in same direcion as the wind blowes, and also try to point it downwards.
  3. Some components just won't let go (kind of same as above). When that happens - espechial big components like heavy transformers with several legs and transistors still attached to cooling ribs - it get's into a battle between me (want to get it out) and the component(s) (wan't to stick to PCB). Since I don't have the brain to just give up, it resulting in overheating of PCB, the PCB being more flexible/soft = component just get even harder to breake loose = even more heating. This results in bad smoke as something between PCB layers start to boil/melt. This was one of the first bad experience I had using this method, so I always attemt to remove smaller components first so they don't take damage of overheating.
Some experiences og different types of components - what may survive and what components may not
  1. Electrolyte capacitors. One property that most electrolytes have in common is that it takes time from first leg get loose until the next is loose and it can be removed. It seems to be more likely to happens if the area of PCB was not heated. I did a test (have got a esd meter) on about 50 electrolytes in different sizes, and they all survived the heatgun removal method. That is - I had two old computer mainboards that both was defect - and it turned out they both had the very same electrolyte that was defect (no way the heat gun could have being the reason for that when all the others survived)
  2. Electrolyte capacitors (2 - not actually relevant to use of heat gun, but still need to mention). After having tested the actual capacitance/esr it turns out that for some computer mainboards, it seems that someone have cheated. Found one motherboard where every single capacitor labeled 2200uF, they all measured around 1300uF, esr around 0.2 and otherwise in good condition.
  3. Resistors, diodes etc: For the most part, these components survives. But : when there is an area on PCB crowded by small components like this, the problem is simply that too many of them get loose simoultanely and I can't pick them out fast enough, so there is a risk of overheating.
  4. Resistor bridges : I'll say that 2 of 3 is removed successfully. The last third get broken because one or more legs wouldn't let go, and the material get soft and breake apart. This may also be a problem because I haven't a decent plier that can share the load ofer the surface, so it may have failed even if I was using a soldering iron.
  5. DIP switches. These normally have casing of plastic. By now I have done two PCB's that contains rows of DIP switches, and the tesult varies. One PCB from early 90's that have always being located indoor, was no problem at all - all the dip switches survived (four pins pr switch). The other one had being located in a enclosure outdoor near sea for several years (mid 90's I guess) and I managed to resque 4 of 5 DIP switches. The one that failed had a very poor pin, not a problem with melting plastic. Overall experience is good.
  6. Communication ports (computer mainboards). For the most part, all those survives because they're just basic contact points. The main issue is some PCB boards becomes flexible when heated, results that parts with multiple pins won't let go. The bigger and heavier the component is, the bigger is the issue it seems.
Various lessons (some thing not a good idea)
  1. Using paperboards that have contained milk. Since some of those have a very thin plastic layer inside, some SMD components that I thrown in the box was actually stuck at the inner coating. So after twisting and pooling loose, that resulted in bad contact on the smd components surface. Hint - use metal boxes if possible.
  2. PCB flexing like a spring : Using hot air gun, results in pretty much melted solder on the surface on the PCB. When a component finally get loose, the PCB may flex like a spring, resulting that both soldering and other components may fall to the floor. Removing hardened solder from the floor may involve some wasted time, you may want to put cardboard on the floor to prevent that work afterwards.
  3. Seems that different PCB's have different heat transfer caracteristics. For most PCB's (most of moderns, like in computers) it is less work to use heat gun over soldering iron. But there is certain PCB's that is hard to work with. Had an old AT power supply board that when heated up, it flexed more than other pcb's so I had to fight to get all components out (destroyed some film capacitors on it) and troubled with overheating. It may also be caused by the manufactor had used much more soldering on it compared to other boards.
  4. Don't try to remove components close to vise : Some of the early lessons I made. If there is a component very close to the vise, I always struggle to get it loose and often just have to cancel because PCB overheats before the component get loose. I can only assume this is due to heat loss to the vise itself or the air flow somehow doesn't pass the PCB close to vise. This is one experience that I know, but I don't fully understand why it behaves this way.
  5. Bonus lesson (1) : components that won't get loose. If a component refuse to get loose, most of the times it can be solved by moving the hot air gun so that it aim the air flow at the oposite side of the PCB.

Well that is it. This is the experiences that I have made spending two evenings of mass stripping of PCB components.

Any comments is welcome - other things I havent mentioned? I would like some ideas on how to deal with sticky smell, is there a better way to prevent it to stick to air and clothes?
 
I sometimes use a propane torch to melt solder in a spot about 4" diameter then smack the PC board against the work bench the liquid solder fly off. Then parts are easy to pick off the PC board. These days I only remove parts that I really want.
 
... Any comments is welcome - other things I havent mentioned? I would like some ideas on how to deal with sticky smell, is there a better way to prevent it to stick to air and clothes? ...
Proper ventilation: ideally, a ventilation hood.

Even though they may be invisible, the stuff that off-gases from PCBs is very bad for your lungs...

And did I mention a ventilation hood?
 
My last experience which I posted some comments about, maybe some four or five years ago: used a heat gun, with the PCB upside down on a big saucepan. Heating was done progressively on a big portion of the board, shaking it from time to time. You can hear the components when they fall inside.

Good thing you can do that almost everywhere, that is, with ventilation.

Not something I would repeat. As long as I have space to store the boards, I will retrieve whatever I need specifically just when I need it. No more massive desoldering. Not worth my time.

AND THERE IS AN INSIDIOUS DOUBT STILL NOT SOLVED: if the equipment was discarded because of malfunction, how could I be sure that what I save is not the failed component?
 
Instead of buying cheap defective components on ebay, you are wasting valuable time harvesting defective components with a heat gun. Why?
Will you sell your harvested defective and and cooked components on ebay?
 
AND THERE IS AN INSIDIOUS DOUBT STILL NOT SOLVED: if the equipment was discarded because of malfunction, how could I be sure that what I save is not the failed component?
Well - I tend to think of sorting out different components afterwards as kind of relaxing therapy session and have time for it. Actually, the two last pcb I did take the time to test the electrolytes, and it turned out that there was faulty caps on all of the defective devices :)

Instead of buying cheap defective components on ebay, you are wasting valuable time harvesting defective components with a heat gun. Why?
Just so I can prove they're still usable :cool:

Will you sell your harvested defective and and cooked components on ebay?
No, the shipping cost is too high. But if you happens to come across a marked for cooked components let me know ;)
 
Here is another thing.

I did mentioned I dismounted everything on an old AT power supply for use with 110V supply (same as mentioned here, but otherwise totally unrelated). Since I knew that the PSU was broken (any attempt to connect resulted in no output voltage and a very distinctive screamy sound not from the fan).
First I suspected some electrolytes, so I did tested all of them with my DE-5000 meter and they all seems in good condition.

Then I figured out I should test all film anc ceramic capacitors too. I didn't expect to find any of those in bad condition. However it did turns out that two of them I figured was faulty - kind of. This is rare and I have no proper explanation - exept it could be the heat from the hot air gun that have caused them to malfunction.
Whenever I put the test clips (the crocodile variant that ships along the meter) over the legs, it measured a capacitance close to zeero (bending the legs slightly inwards). But performing the same test when test clips are pressing the capacitors legs outwards it measured the expected (rated) capacitance with esr close to zeero.

So of cours I suspected bad connection from test clips, so I did the same test multiple times (three I think) with same result each time. I also tested other caps using same method but got acceptable readings every time for all the other caps. Only those two had this issue.

I'm not sure if those small caps was the original reason why the power supply malfunctioned in first place. If those damages is solely related to the use of heat gun, then I start to suspect the transformer itself, even if it is an unlikely part to go broken.

Ah - very important thing I forgot to mention. The original computer casing have being very ruff threatment during air shipping and when case was opened the impact was strong enough to rip off the CPU cooler from the socket. Therefore in this case I actually beleive that the transformer (heaviest part) may being damaged.
Before I removed the components from PCB, I could not spot any single soldering point.

[edit]
Tested the output transformer insulation with a megger set voltage 500V, all passed. Wonder if I should crank it all the way up to 1kV. I cannot find any specifications on the transformer. Probbly in-house information only, but it says:
Label 1: "360--425D ; SK 1 0038"
Label 2: "DASH 2 B-5" (just a removable transparent sticker, maybe an OK/tested OK mark from manufactor ?)
 
Last edited:
It's funny you should mention faulty film caps - I found my first dead one a few months ago, in a failed machine tool CRT monitor:

View media item 35
It's supposed to look like this; I happened to have an identical monitor with a dead line output transformer, so it was an easy replacement:

View media item 36
 
Now tested all transformers to 1kV megger, and they all tested good. That is one main transformer, 5 windings ring core transformer and then a two-winding smaller ring core transformer. Also, measurement of inductances didn't reveal any potential failures.

Well - one thing to put on notice : when measuring inductance on a multi-winding transformer with the DE-500 meter, and if the probes connects to two pins from two independents windings, it reports readings high values (hundreds of Henry's) probably due to stray capacitance.
I let myself fool a bit thinking I just found the faulty component, until I noticed the <micro> symbol was missing :oops:
 
Just figured out a little trick in order to get loose parts that seemingly are stuck - espechially DIL IC's and their plastic socket (pcb from an industry batteri charger from around '95) - To move the plier to the other side (this socket was hollow in the middle, so I was able to make the move the attach angle of plier from ine side to another) - and it instantly get loose.

But for many larger components, it isn't practical to shift grip when it is almost loose simply because it may fall off at any time.
 
Status
Not open for further replies.

Latest threads

New Articles From Microcontroller Tips

Back
Top