Desoldering smd: braid or hot air station?

[CharlotteSwiss]

Well how else would you solder a 2 pin component? Tinning one pin and sticking it on while its hot is the only way i know of doing it.

Instead of tinning the pitch, put tin on the tip; I think it's the same thing, the component is already in place and there is no need to fix it .. but the two procedures could be equivalent.
My doubt in the procedure of the video was not the positioning or the welding ... but the fact that after having welded it puts flux on the welds and heats them, it seems bizarre to me, but obviously there will be a reason.

Meanwhile I did other tests this time with smd width 1.6 mm using the 1.6 mm flat tip. So I can easily solder even without flux and they come out pretty well
 

[Berni]

Ah the flux. Well adding some flux and heating it up again is a easy way of fixing ugly looking joints. If you move and reheat a joint too much it gets crusty and ugly looking, but adding a bit of flux and melting it again turns it back into a nice shiny smooth joint because the flux eats away all the crud, letting the solder flow freely into the shape it wants.

[CharlotteSwiss]

Thanks Berni, OK now I got it; I would say that adding flux after soldering is ONLY to be done when soldering looks bad, normally we can do without.
I have already become a better woman with the soldering iron in my hand: up to smd width 0.8 mm I can weld, with the smaller ones I have not yet tried, but perhaps soldering paste and air would be more suitable ...
 

[Howardlong]

Well how else would you solder a 2 pin component? Tinning one pin and sticking it on while its hot is the only way i know of doing it.

Instead of tinning the pitch, put tin on the tip; I think it's the same thing, the component is already in place and there is no need to fix it .. but the two procedures could be equivalent.

If I am understanding correctly, you're advocating applying solder to the iron's tip and then applying it to the pad and part? It so, this is fraught with difficulty, because the flux in the solder will have evaporated by the time you apply it to the pad and/or part.

My doubt in the procedure of the video was not the positioning or the welding ... but the fact that after having welded it puts flux on the welds and heats them, it seems bizarre to me, but obviously there will be a reason.

The reason is that any previous flux, either from the solder itself or explicitly applied, will have already evaporated or deactivated.

The only problem with the video is the amount of time the tip's on the joint: it's far too long. It should be a straight in and out procedure, half a second or so, unless you're soldering thermal pads.

Furthermore, with ceramic caps in particular, you risk cracking them with too much thermal stress. Worse still, that failure mode is often intermittent.

Meanwhile I did other tests this time with smd width 1.6 mm using the 1.6 mm flat tip. So I can easily solder even without flux and they come out pretty well
 

I think you may be over analysing the tip size. I only use two tip sizes for soldering, a 0.2mm conical 40W cartridge bit for most SMD connections, and a 2.4mm x 0.8mm chisel for joints with higher thermal mass on an 80W temperature controlled iron. Both are dialled in at 350 deg C.

For soldering QFPs, it's as in the video.

For QFNs or other devices with power pads underneath, I lightly tin all the pads, removing excess solder with braid, apply flux, hold the device over the pads with tweezers and reflow with hot air. Once the solder has melted on all of the pads, you should be able to slightly nudge the part with the tweezers and see it snap into place due to the surface tension of the solder.

For desoldering small SMDs, I use a tweezer iron with a selection of different sized 2x40W cartridge bit sets which manage everything from 008004 to SOIC8. Sure, you can fight with a normal iron with a larger tip, but if you're going to be doing it quite a bit, tweezers are the way to go. Hot air is also an option, but for small SMDs, I much prefer tweezer irons.

For desoldering larger SMDs like QFPs, QFNs, BGAs and power devices, I use hot air, and then clean up all the pads with braid.

The only time I use paste is with a stencil, and typically that's when I'm assembling a final preproduction batch or doing a respin. Otherwise I tend to manually hand solder each part especially if it's only a small board with a under a couple of hundred parts.

To remove excess flux I use an aerosol flux cleaner with a brush attachment for most use cases. For high impedance and ultra low power stuff, I use an ultrasonic bath with detergent followed by a couple of rinses in deionised water.

A good iron and bits will last many years: the Wellers I use are 15 years old and see daily use. The bits, if treated well, will also last years. The chisel bit is exactly the same one that came with the iron all that time ago, and before I switched to cartridge bits about three years ago, the 0.2mm conical traditional bit I used was also still going strong after well over a decade.

[CharlotteSwiss]

If I am understanding correctly, you're advocating applying solder to the iron's tip and then applying it to the pad and part?
...

thanks for your excellent contribution 

you got it right, I put a small drop of tin on the tip and then solder the first pin; But first I put some flux on the pitch and pin ..
However, even by stagnating the pitch and then approaching the component at a later time .. and re-melting the pond, the flux will have vanished ...
In my opinion the two procedures are the same, in both cases it should be fluxed before carrying out the welding.

Timing to carry out the soldering on ONE pin of the smd: do you maintain that the tip support must last only 0.5 seconds in total? Isn't that too little?
If we talk about soldering with tip and wire, I heat the area for 1 second .. then we have almost 1 second to bring the wire closer and melt it .. and then important still 1 second more so that the tin is well arranged. Total about 3 seconds (note: I use a low temperature for SMD, 280 °)

I'm still learning about tin soldering:
.THT desoldering with pump I succeed well
.SMD desoldering with hot air is good for me
.THT soldering with iron ok
.two pin SMD soldering with iron ok
Now I am experimenting with soldering of SMD IC, and from some timid tests I have begun to understand that:
.with a fairly wide Pitch I can solder in the traditional way (pin to pin)
.with very narrow pitches I have two possibilities: (solder paste I am not interested in for now)
1) drag welding
2) hot air welding (first, however, I have to tin the pitches with the soldering iron)

For now I have tried only a 14 pin smd IC with 1.27mm pitch: I have tinned the 14 pitches (easy), I have placed the IC on top and held with tweezers (not easy operation, the pond on the pitches does not help ..); I heated with air to 300 °, after a few seconds I saw the pond melt and the 14 feet that slowly drowned .. it came out well! (in trying again, however, I struggled, maybe it's not that easy)

The experiments on IC smd soldering continue... 

[tooki]

Charlotte, you are again using various terms incorrectly, which makes me suspect you've misunderstood what other people mean when you read their posts. So here are the correct meanings (in correct English; nobody needs to reply that their native language uses them differently, nor what the incorrect use by nonnative speakers is):

soldering: joining two metal objects by melting a different metal (the solder), which melts at a lower temperature, between them. Soldering is a comparatively low temperature process. This is what we use in electronic circuits.
(There also exists "hard soldering", which is still soldering, but with higher temperatures. It's used in things like plumbing and sheet metal roofs, but not in electronics.) "Welding" is not an acceptable term for soldering.

welding: joining two objects by melting them directly and/or joining them with the same material. Soldering is not welding. None of this discussion has ANYTHING to do with welding.

solder: the alloy used to solder two things together. "Tin" is not an acceptable term for solder. ("Solder wire" is solder in wire form, essentially a metal tube with a flux filling. "Solder paste" is powdered solder suspended in flux gel or paste.)

tin (verb only):

• applying fresh solder to a soldering iron tip. (Proper tinning procedure is to wipe the iron clean, then apply solder to the tip. If the old solder on the tip was very oxidized, it can be helpful to repeat the process to ensure the old, overcooked solder is really gone.)
  
• applying a small amount solder to a pad, wire (especially stranded wire), or component terminal in preparation for soldering it to another object.

joint: the connection point where objects are soldered together. (This word applies both before and after the soldering process is completed.)

fillet: the solder in a soldered joint. (Like in the sentence "In a THT joint, the solder fillet should be concave.")

pitch: the repeating distance between pin or pad centers. So "0.5mm pitch" means the centers of the pins are 0.5mm apart. The pins, necessarily, will be narrower than that.

pad (or land): the metal area on a PCB to which a component pin/terminal is soldered. "Pitch" is not an acceptable term for a pad.

flux: a chemical that removes oxides. In electronics soldering, flux comes from two sources: 1. a flux core inside the solder wire. 2. separate flux, like liquid, gel, or paste. ("Tacky" flux means a gel or paste that has a sticky texture, which can be great for holding down an IC when hand-soldering it.)

Now that this is clear, some points:

Berni... true what you say, but it scares me enough to have to remove all the bridges that are created with the dragging of the tip. However with the gun I have unsoldered many SMD multi pin, as soon as I can I will play to figure out which method could be easier for me.

Bridges aren't a big problem when drag soldering, as long as you use a good flux. (And periodically re-tin your soldering iron tip well, to "rinse" it of overcooked solder.) I recommend a gel flux — my favorite is ChipQuik's SMD291NL, which smells like toasted hazelnuts during use instead of like burning acid.

Well how else would you solder a 2 pin component? Tinning one pin and sticking it on while its hot is the only way i know of doing it.

Instead of tinning the pitch, put tin on the tip; I think it's the same thing, the component is already in place and there is no need to fix it .. but the two procedures could be equivalent.

It's similar but not the same: the solder on the iron tip will oxidize the entire time it's on the tip, since it stays hot. The solder on the pad isn't kept hot as long, so it won't oxidize as much.

Oxidation leads to bridging, dags (spikes formed when pulling away the soldering iron tip), dull surfaces, and difficulty in solder flow.

Flux doesn't remain active forever.

Traditional rosin fluxes have two temperature ranges before burning:
   inactive (cold) <-> active (hot) -> burnt (hotter).
Modern no-clean fluxes have three temperature ranges before burning:
   inactive (cold) -> activated (hot) -> deactivated/neutralized (hotter) -> burnt (even hotter)

The soldering iron temperature is hotter than the temperature needed to burn the flux, so it's only active for a brief moment before it gets too hot, deactivates, and burns up as smoke. So when you apply solder wire directly to the tip, all the flux from the flux core gets destroyed a moment after it's applied.

This is why in traditional hand soldering, the rule is to always apply the solder wire to the joint, not the tip: the flux in the core melts a fraction of a second sooner than the solder itself, flowing across the joint just before the molten solder follows. When you try and carry solder to the joint on the iron tip, there is no flux.

The exception, somewhat, is joining solder to solder: when both objects to be joined (like a pad and component terminal, or two wires) are already freshly tinned, you can often get away with reflowing without added flux. But for this to be successful, oxidation must be kept at a minimum: the tinning processes had to be quick and properly fluxed, and the iron tip must be freshly tinned without oxidized solder. It won't work if the iron is dirty with old solder, and it won't work on tinned objects that have sat around for a while, as their surfaces will oxidize.

By adding flux, you give the entire "system" a way to eliminate the oxidation, resulting in a good joint.



Side note about no-clean fluxes in hand soldering: The flux in solder wire always reaches a temperature high enough to deactivate it, so residues do not need to be removed. But when you add additional flux liquid, gel, or paste, parts of the board may get hot enough to activate the flux, but not hot enough to deactivate it (they do not deactivate by cooling off!!), meaning that residues can be corrosive. So it is important to remove these residues in hand soldering. Rosin fluxes, on the other hand, become inactive once they cool down. However, some rosin gels and pastes have ingredients that make them behave a bit like no-clean, so I recommend cleaning those residues, too. Liquid rosin is usually fine, as long as it is able to dry quickly.

Thanks Berni, OK now I got it; I would say that adding flux after soldering is ONLY to be done when soldering looks bad, normally we can do without.

Not in drag soldering. Flux is mandatory for this method, or any other reflow method, because you aren't applying solder wire (and thus its flux core) to the joint. Flux gel/paste is ideal for reflow soldering because it allows the terminals to remain "flooded" in flux, so that as you drag the iron across, there is always flux present.

When doing 2-terminal components (like chip resistors), I normally use liquid flux. The explanation above tells you why.

(I typically tin one pad of each component's footprint, for the entire board at once. Then I brush flux onto all the pads, both the tinned ones and the blank ones, then freshly tin the iron tip, then go in and remelt the solder on a pad and push in the component. I do that for the whole board, then look at the board from the side and make sure the components are laying flat. Any that are not, I brush on flux, then hold the component down with tweezers or a wood stick, and reheat the joint so it lays flat. Once they're all flat, I go back and solder the second side of each component.)

[CharlotteSwiss]

thanks tooki, in fact you had already suggested that I am wrong the terms; I am well aware of what the various terms are, the problem is the translator who sometimes confuses / mistakes the terms! Anyway yes, my fault! 

I haven't done the drag test yet, but I'll try it in the next few days; the flux I had taken is exactly what you indicated (perhaps you had recommended it to me when I took it); I'll try with the small concave tip, we'll see ..

So for the pointing of the first pin of the SMD, it is ALWAYS better to tin the pad and then bring the component and tip closer to solder it; so better to avoid the drop of tin on the tip then?
Classic soldering with tip and wire: if, however, the wire I support it only at the joint and not at the tip, it takes enough time for the wire to melt ... for this reason I support it between the tip and the joint, but I understand that it would be better not to rest the thread on the tip ..
For the final cleaning no problem, I always use isopropyl alcohol and a toothbrush, the board is clean

We can therefore state that: every time a solder melts with the tip to fix it (so we do not use the wire), we must first apply some flux

 

[tooki]

If we talk about soldering with tip and wire, I heat the area for 1 second .. then we have almost 1 second to bring the wire closer and melt it .. and then important still 1 second more so that the tin is well arranged. Total about 3 seconds (note: I use a low temperature for SMD, 280 °)

For leaded or lead-free solder? For leaded, that's on the slightly low side, and for lead-free, that's almost certainly too low. I generally set 300C for leaded and 320C for lead-free on my Ersa, but big joints or very small soldering iron tips (which have poorer heat transfer) can require higher temperature settings.

A temperature that is too low can actually increase damage, if it means you have to dwell on the joint too long. This is a classic mistake with, for example, connectors. Intuitively, you think "well, the plastic will melt, so I'd better stay with a lower temperature". But the mass of the contact sucks the heat away and solder won't take, so you have to keep heating. Meanwhile, the plastic is going gooey. With a higher soldering iron temperature, the contact gets hot enough to solder quickly, before the plastic has time to get hot.

Similarly, PCB pads behave the same way: brief high temperatures are better than prolonged lower temperatures. A pad that is heated for 30 seconds (even at, say, just 260C) is much more likely to lift than a pad heated to 340C for three seconds.

[CharlotteSwiss]

I only use lead wire for now.
I used my Ersa like this (before reading your message): THT in general, about 320-340 °
SMD (for now only two pins): 280-300 °
Let's say that THT always balances them at 330 ° (it could be good); here is perhaps better that I raise the temperature for the SMD, no more 280 but better 300-320 ° ....
(I thought the smd could be damaged).
Obviously, by raising the temperature the contact times of the tip decrease ...
For the IC that I have yet to experiment, the soldering iron will use about 300 °, while the air I saw that 300 ° could be fine (with air I have to cover all the extra area, otherwise bye bye plastics .. )
 

[tooki]

thanks tooki, in fact you had already suggested that I am wrong the terms; I am well aware of what the various terms are, the problem is the translator who sometimes confuses / mistakes the terms! Anyway yes, my fault! 

 

I haven't done the drag test yet, but I'll try it in the next few days; the flux I had taken is exactly what you indicated (perhaps you had recommended it to me when I took it); I'll try with the small concave tip, we'll see ..

Could be, I may have suggested it!

So for the pointing of the first pin of the SMD, it is ALWAYS better to tin the pad and then bring the component and tip closer to solder it; so better to avoid the drop of tin on the tip then?

I wouldn't go so far as to say ALWAYS without exceptions, but as a general rule, yes.

Drag soldering is exactly such an exception (with the concave tip being designed specifically for this technique), but the flux becomes absolutely mandatory.

Classic soldering with tip and wire: if, however, the wire I support it only at the joint and not at the tip, it takes enough time for the wire to melt ... for this reason I support it between the tip and the joint, but I understand that it would be better not to rest the thread on the tip ..

You only want to feed the solder wire onto the joint itself, not onto the tip. However, what is usually helpful is to have a small pillow of solder on the tip to act as a thermal bridge. It will shape itself to the component leg and the pad, providing much better, faster heat transfer than a tip that's completely "dry". This way, the joint itself heats up much faster, allowing the solder wire to melt on it quickly. With time, you'll get a feel for this. (And yes, if you begin to heat a joint and then realize the pillow is missing, you can feed a tiny amount of solder wire onto the tip to create the pillow, then proceed with feeding the solder onto the joint itself.)

For the final cleaning no problem, I always use isopropyl alcohol and a toothbrush, the board is clean

FYI, not all fluxes (especially no-clean ones) come off with isopropyl. (And even the ones that do, it's usually a bit slower and uses up more solvent.) Commercial flux removers are much better. Here in Switzerland, where isopropyl costs nearly as much as flux remover, it really makes sense to use the flux remover. (The cheapest one I've found here is Distrelec's house brand, RND 605-00128. It's about 12 francs for a 400ml can, which is only slightly more than the 26 francs you pay for a liter of IPA at the pharmacy. I suspect it's actually made by Electrolube.)

We can therefore state that: every time a solder melts with the tip to fix it (so we do not use the wire), we must first apply some flux

Yes, basically.

[tooki]

I only use lead wire for now.
I used my Ersa like this (before reading your message): THT in general, about 320-340 °
SMD (for now only two pins): 280-300 °
Let's say that THT always balances them at 330 ° (it could be good); here is perhaps better that I raise the temperature for the SMD, no more 280 but better 300-320 ° ....
(I thought the smd could be damaged).
Obviously, by raising the temperature the contact times of the tip decrease ...
For the IC that I have yet to experiment, the soldering iron will use about 300 °, while the air I saw that 300 ° could be fine (with air I have to cover all the extra area, otherwise bye bye plastics .. )

For leaded solder, 280-300C will be fine, then.

Hot air temperatures are nowhere near as exact, since even just the distance from the board makes a huge difference in temperature. I usually use a temp of around 270-300C, with a large nozzle, high air flow, and longer distance, to preheat the bottom of the board. Once the board is properly preheated, I'll use 320-340C with a smaller nozzle, low air flow, and short distance from the top to reflow the components. Preheating is especially important for multilayer boards, heavy/large components (like large inductors), and components with thermal pads on the bottom. Also for delicate parts like connectors, which you want to expose to only minimal direct heat.

[CharlotteSwiss]

tooki:

I wouldn't go so far as to say ALWAYS without exceptions, but as a general rule, yes.
Drag soldering is exactly such an exception (with the concave tip being designed specifically for this technique), but the flux becomes absolutely mandatory.

I will use the procedure of tinning the pad then; but perhaps then, before approaching the component, it would be better to use the flux on the tin plating done previously ...

You only want to feed the solder wire onto the joint itself, not onto the tip. However, what is usually helpful is to have a small pillow of solder on the tip to act as a thermal bridge.

ok of course, in reality I always tin the tip before a weld, but only slightly .. I will do some tests without placing the wire on the joint ..

FYI, not all fluxes (especially no-clean ones) come off with isopropyl.

in fact I usually use water flux, it does not dirty very much indeed almost nothing ... The other one I struggle more, but it comes away
(I took isopropyl, because I also use it in MTB maintenance)

For leaded solder, 280-300C will be fine, then.

yes, but for THT I prefer to solder at 320-330 °; with SMD ok then I keep a good range 280-300 (in the tests done at 280 °, I must say that the welds came out well and fast)

Hot air temperatures are nowhere near as exact, since even just the distance from the board makes a huge difference in temperature

With hot air welding I have just started ... so I am at a level below zero; However, I immediately realized that small nozzles heated very little. I will only use this solder exclusively for IC with many pin (and pitch close) or just to do a rework of the component to revive the contacts, so it will happen to me a few times, but it does not matter, I will experiment to learn

I would also add that soldering with hot air only makes sense for a rework, or it would make sense using solder paste; adding solder to each pad (and then heating to solder the component) doesn't seem like a logical job...

thanks 

[CharlotteSwiss]

ok, i tried soldering an smd IC (32 pin, pitch 1.27), this is the second or third time i try soldering an IC. To start I wanted to try soldering one pin at a time in the traditional way. I must say that I thought worse, I welded it without creating any bridge.
The welds of the upper side in the photo, I made them with a 0.8 tip (320 °), but often it didn't weld immediately, or it was cold ball .. maybe with a 0.8 tip I would have had to raise the temperature more I guess ...
While the welds below in the photo, I made them with the 1.2 flat tip, temperature 280 ° (I wanted to try to weld on purpose with this low temperature), but it certainly went better, the tin melted immediately and welded the pin well; maybe I should remove the wire first and maybe pour less tin, but it seems to me that for being one of the first times did not go badly.
Note: I used a soft water-based flux (I didn't bother with the better performing one).
Now I would like to understand if with a lower picht it is still possible to solder pin to pin, and if it will be necessary to do the drag instead ...

[tooki]

You talk about drag soldering as if it were this hugely difficult task only to be done if nothing else is possible. It’s actually my favorite type of hand soldering, to me the question is “can I drag solder this?” I don’t understand why you would want to solder pin by pin on anything that can be drag soldered.

And yes, solder paste is definitely the way to go for hot air.

[thinkfat]

If you have a proper paste for hot air and a good method to dispense it, then, yes. I find myself often just tinning the pads, adding flux, placing the component in the bed of flux and then using hot air to reflow. Especially for QFN this is my go-to method. But also for QFP.

[CharlotteSwiss]

You talk about drag soldering as if it were this hugely difficult task only to be done if nothing else is possible. It’s actually my favorite type of hand soldering, to me the question is “can I drag solder this?” I don’t understand why you would want to solder pin by pin on anything that can be drag soldered.

And yes, solder paste is definitely the way to go for hot air.

it was just a pin to pin test, just to understand if I was capable, I thought worse; obvious that my intent will be to solder these IC with drag, as soon as I have time I try (but I think it will be time to use the classic flux ..), for now with two-pin smd and IC pin to pin not I still used it.

With solder paste it would be too easy, my final test with the ICs will be to tin all the pins with wire and then solder with hot air
 

[jmelson]

The chip photo looks quite good!  If in doubt, while under the microscope, poke the leads gently sideways with the point of an X-acto knife.  If the lead moves, it wasn't soldered well.  If it doesn't move, the soldering was fine.  I routinely do down to 0.4mm lead pitch witht this sort of technique.  The 0.4 are difficult, 0.5 is easier, 0.65 mm is quite easy.

Jon

[CharlotteSwiss]

If you have a proper paste for hot air and a good method to dispense it, then, yes. I find myself often just tinning the pads, adding flux, placing the component in the bed of flux and then using hot air to reflow. Especially for QFN this is my go-to method. But also for QFP.

The paste makes the job easier, but I am a traditionalist, I believe that soldering is with tip and wire 

[CharlotteSwiss]

The chip photo looks quite good!  If in doubt, while under the microscope, poke the leads gently sideways with the point of an X-acto knife.  If the lead moves, it wasn't soldered well.  If it doesn't move, the soldering was fine.  I routinely do down to 0.4mm lead pitch witht this sort of technique.  The 0.4 are difficult, 0.5 is easier, 0.65 mm is quite easy.

Jon

Yes, especially with the tip of 1.2 (lower welds in the photo) to be a beginner I was not bad (I want to say, without using classic flux, but a water flux, maybe just water really).
Let's say that I succeeded in the 1.27 pitch measure, lower I don't know; 0.5 seems to me already a close measure, if I have not misunderstood the pin to pin balance?
Thanks for the advice on testing the tightness of the welds, I could use the tip of my pointed tweezers
 

[CharlotteSwiss]

Btw just so you are using the correct terminology, after you solder a component we call that connection a soldered "joint", technically described as a "fillet".

The limit of hand soldering individual pads/legs is your tip size, magnification and patience. As others mentioned though, drag and hot air reflow soldering are also valid techniques and can be more efficient.

Your example looks good, try a higher temp for better wetting with slightly less solder and remember your going for best technique first time, if your correcting joints constantly it's a fail no matter how it looks in the end.

As guide look at some wave soldered packages and aim somewhere in between. When you can still clearly see the whole top outline of each leg your right in the zone for the amount of solder. So it's ok to tin the top of the leg but you don't want a dome of solder on top of the leg.

The reason why you aim for concave solder joint/fillet is you can tell by inspection if you have adequate wetting. Under ideal conditions solder will spread out and form an even intermetallic layer (tinning) and the concave curve the solder forms between the pin/leg and pad shows how well that wetting performance went.

The other benefit of concave fillets is defects in coverage are easy to spot. If too much solder is applied it can mask flux performance and temperature problems.

yes in fact I have problems using the fine tips (in my case 0.8 mm), because it often seems that they do not heat well .. certainly raising the temperature will be better with these fine tips, I could also use 330 °; with the 1.2 flat tip even at 280 ° I saw it stagnate well.
Yes, I spilled too much solder, all pins are solder-free, and it shouldn't be. But they are so small that it is not easy, maybe even my 0,56 thread is too big for these jobs I think ...
 

[tooki]

Btw just so you are using the correct terminology, after you solder a component we call that connection a soldered "joint", technically described as a "fillet".

She has now made clear that she uses uncorrected machine translations and doesn’t care about the correct terminology that has been explained repeatedly to her. Don’t waste your time, I wasted enough for all of us.

[CharlotteSwiss]

Solder wire size, flux and a perfectly clean tip become increasingly more important for control the smaller you go. Small tips drop in temp quicker when they touch the pad. They have also less throughput so temp recovery is longer compared to a larger tip. Go with the largest tip you can but 1.5x pad contact width is about the limit, and obviously if you are bridging pins easily you need to rethink things.

The reason people say use the lowest temp possible is to circumvent poor technique and lifting pads but if you get in and out quickly it's low risk. The other reason why a little hotter helps when you get proficient it preheats the neighboring pad as you go along. With poor technique, a too high temp, inadequate flux, touch ups, you can start to see how damage occurs.

For that exact reason you will see experienced solderers often do alternate pins on through hole packages. As the dwell time needs to be longer for through hole this reduces the components overall exposure to high temp. For smd it mostly depends on board and component density. When you find it difficult preheating the pcb becomes the best way to reduce temp damage.

This video is one of the better ones on smd removal if you have limited access to tools.

Thank you Shock 
using the tip of 1.2 I did not put it flat in front of the pin, but laterally (perpendicular to the IC), by doing so there is no problem with touching other pins. Of course I would have liked to use the 0.5 tip, but I have seen that it hardly melts the tin in a short time. Yes, with the small tips I have (0.8 and 0.5) I have to improve the technique.
To unsolder IC I use hot air, it's very easy (I just have to cover adjacent plastic parts with insulation, if there are any.

[CharlotteSwiss]

Yes, I spilled too much solder, all pins are solder-free, and it shouldn't be. But they are so small that it is not easy, maybe even my 0,56 thread is too big for these jobs I think...

Just in case that got lost in translation, fresh solder doesn't need to travel up the whole leg just to the back of the leg heel where it should be a concave fillet.

A lot of people use 0.5mm diameter solder wire for everything. 0.3mm seems less common to buy due to flux varieties but I think the demand will start to rise steadily, the rule is use what works best for you.

Those Pace videos I recommended earlier have a bit about establishing good joints. There is also some free Nasa training you can find in google, has some good pictorial examples.

So do the pins have a concave part that accommodates the solder? I have to zoom in with the lens and check; one thing that I did not understand when I placed the chip above the pads, is why the legs reached the END of the pad in width, soldering can only take place on the sides and not in front of the pin

---------------------------------------------
Off topic:
English yes / English no / translator: not all people in the world are lucky enough to speak English, so if we want to converse on a worldwide forum we have to use the translator; if the translator is wrong, I can also re-bind 100 times, but I will not be able to understand the error.
I believe that the important thing is not to discriminate against people for the language they speak!
 

[tautech]

So do the pins have a concave part that accommodates the solder?

No.
Imagine L represents the end of a IC pin and the pad junction.
Add solder and it should look a little like a C with the concave part all of solder.
With a wide pad the sides also represent a C.

[thinkfat]

If you have a proper paste for hot air and a good method to dispense it, then, yes. I find myself often just tinning the pads, adding flux, placing the component in the bed of flux and then using hot air to reflow. Especially for QFN this is my go-to method. But also for QFP.

The paste makes the job easier, but I am a traditionalist, I believe that soldering is with tip and wire 

I would not say it is easier. I use paste mostly for passives or larger pad pitch components, because it is faster to solder many parts at the same time. It's also cleaner as it leaves less flux residue on the PCB.

However, for fine pitch parts like TSSOP or anything at or below 0.5mm pad pitch it is quite difficult to apply the right amount of paste if you don't use a stencil. It's almost always too much, leading to bridged pads and requiring a touch up with more flux and braid. It's easier and faster to drag solder those parts.