Hot air soldering, what am I doing wrong?

[TheHolyHorse]

Hi, so I'm working on a project that has an LQFP48 package. It needed replacement so I de-soldered it and though I'd use the hot air to get it back as well, because it's a pain in the ass to use the iron with all the other stuff in way.

So now to the issue. I've been watching a bunch of videos of people soldering QFN QFP etc with hot air and the chip just gets suck in place no problem. But that never happens for me. I add solder to the pads and a bunch of flux, but the flux burns up before the solder even melts so I think that's why it doesn't help me and it doesn't help that I'm shaking like a leaf.

But how do other people do it? What am I supposed to do to prevent all the flux from burning to early?

I'm truly lost

[TomS_]

Is the part not sticking or blowing out of the way? Maybe too much airflow, or holding the gun to the side instead of top down? Dont get in too close too quick, the component and the surrounding board need to soak up some heat first, so take your time.

Ive seen people prod the part with tweezers etc once all of the solder has melted, but I couldnt honestly say I would do that if it has self centered.

What sort of flux are you using, is it a "paste" or liquid? How much are you using?

You could take a look at Louis Rossmann on youtube. He would seem to have even BGA reflow down to an art, and does not spare the flux. Even with the amount he uses, it almost all disappears before the soldering process is complete.

So probably it will all dissolve before the part solders, but that should be fine as it is only needed to clean the surfaces to be soldered, not make the solder melt or form joins (except by ensuring the pads/leads etc are clean of contaminants etc).

I havent personally done any hot air reflow, but I have also watched a few videos and other people that do it, and Im picking up bits and pieces I need to give it a go eventually. For now I use drag soldering and that seems to be working just fine for me.

[madires]

Take some e-junk, practice and figure out which settings of your hot air station work best. Flux is meant to remove oxidation and needs heat to do its job. So it will evaporate during the whole process and leave some residue. The solder's surface tension is what drags components into position.

[kripton2035]

may be you're using too much air flow ? try half the max value to see if the chip stays in place.

[xmetal]

Using leaded solder helps as the chips tend to jump in to to place compared to lead-free. Also it melts at a lower temperature than the latter.

[nali]

The board or rather the pads also need to get heat, are you just directly heating the chip or the surrounding board also?

If the board has a lot of copper either as pours or internal power planes then you'll save yourself a lot of hassle by getting a preheater to warm the board, a dead giveaway is a chip with a thermal pad on the underside for example. You can use the air gun to heat the board area but a preheater makes life a lot easier.

As mentioned, leaded solder is much easier if you can use it.

TBH personally I use a gun to remove, then iron to replace with lots of flux & "drag soldering". Sometimes it's a bit of a pain if there's a connector or something close by but often it's worth the extra few minutes to remove the connector or whatever to make room then faffing around for ages in a confined space.

[TheHolyHorse]

Is the part not sticking or blowing out of the way? Maybe too much airflow, or holding the gun to the side instead of top down? Dont get in too close too quick, the component and the surrounding board need to soak up some heat first, so take your time.

Ive seen people prod the part with tweezers etc once all of the solder has melted, but I couldnt honestly say I would do that if it has self centered.

What sort of flux are you using, is it a "paste" or liquid? How much are you using?

You could take a look at Louis Rossmann on youtube. He would seem to have even BGA reflow down to an art, and does not spare the flux. Even with the amount he uses, it almost all disappears before the soldering process is complete.

So probably it will all dissolve before the part solders, but that should be fine as it is only needed to clean the surfaces to be soldered, not make the solder melt or form joins (except by ensuring the pads/leads etc are clean of contaminants etc).

I havent personally done any hot air reflow, but I have also watched a few videos and other people that do it, and Im picking up bits and pieces I need to give it a go eventually. For now I use drag soldering and that seems to be working just fine for me.

No it doesn't blow off it just sits there and I have to fiddle with the tweezers to get it in place. I have some chipquick flux pen and I tried getting as much flux as I possibly could, but if that has no part in helping the solder suck the chip in place it's not the problem. I heated the board so that the solder on the pads was molten when I placed the chip on the board so every thing should be hot and toasty.

The board or rather the pads also need to get heat, are you just directly heating the chip or the surrounding board also?

If the board has a lot of copper either as pours or internal power planes then you'll save yourself a lot of hassle by getting a preheater to warm the board, a dead giveaway is a chip with a thermal pad on the underside for example. You can use the air gun to heat the board area but a preheater makes life a lot easier.

As mentioned, leaded solder is much easier if you can use it.

TBH personally I use a gun to remove, then iron to replace with lots of flux & "drag soldering". Sometimes it's a bit of a pain if there's a connector or something close by but often it's worth the extra few minutes to remove the connector or whatever to make room then faffing around for ages in a confined space.

I use leaded 60/40 solder.
I also prefer the iron for soldering but it was such a pain and my patience was running low But even when I got it in place I had to touch up some joints that never made proper connection.

[TheHolyHorse]

Take some e-junk, practice and figure out which settings of your hot air station work best. Flux is meant to remove oxidation and needs heat to do its job. So it will evaporate during the whole process and leave some residue. The solder's surface tension is what drags components into position.

Yeah I'm gonna have to do that Got some boards from the previous revision and some dead chips.

[bob91343]

I did a good job when I held the new part in place with a small C clamp.  That way I could adjust its position without heat.  Once in place, a slight touch of heat finished it.  If you don't have a clamp, you can use tweezers with locking tips or even chewing gum.  Or an alligator clip or a bench vise.

Trying to adjust position while solder is flowing is difficult.  In fact, once held in proper position, almost any soldering tool works well.  I have used a tiny iron successfully.  If you short adjacent pins, heat it again and it should clear up.  Use the least amount of solder possible.

[TheHolyHorse]

I did a good job when I held the new part in place with a small C clamp.  That way I could adjust its position without heat.  Once in place, a slight touch of heat finished it.  If you don't have a clamp, you can use tweezers with locking tips or even chewing gum.  Or an alligator clip or a bench vise.

Trying to adjust position while solder is flowing is difficult.  In fact, once held in proper position, almost any soldering tool works well.  I have used a tiny iron successfully.  If you short adjacent pins, heat it again and it should clear up.  Use the least amount of solder possible.

It's kinda tedious to place the chip when I have lumps of solder on the pads. I guess if I remove the solder it might work.
I've gotten somewhat good at soldering it with just the iron as long as there's nothing in the way.

[iMo]

QFN:
1. put the solder on the chip's pins, not on the pcb pads
QFN, BGA:
2. you need a LOT of flux, the chip must swim in the flux, flux must stay there during soldering, so set the airflow such it does not blow the flux off
3. put your pcb on a ceramic support, heat the pcb up before soldering, lot of flux
4. then put the chip on the pcb and still heat, maintain a lot of flux, change the temperature (heating profile) such you change the air gun distance from the pcb, move the gun slowly down and up
5. gently (0.2-0.3mm) push the chip from a side - when the pins are soldered the chip moves back by wetting forces
6. the air flow and the temperature have to be adjusted - needs some training.

LQFP:
1. place the chip on the pads
2. bond the chip with solder on two-three places such it holds firm on the pads, shorts are not a problem
3. put a LOT of flux on the pads
4. solder all the pads with iron tip by moving the iron tip over pins, lot of flux and solder, shorts are no problem
5. use the copper wick and a lot of flux - place wick on the pins and heat it up with the iron from top (the wick must be wet of solder), pull the wick over the pins row from left to right (or right to left) slowly at one side
6. repeat at all other sides.

I did a few LQFP packages with air gun, but had to use the wick afterwards to remove several shorts. So it seems to me it is easier and faster to do it with iron instead.

The key is to maintain a LOT of flux all the time - it always works fine when there is a LOT of flux, chip must always swim in the flux.

[TheHolyHorse]

QFN:
1. put the solder on the chip's pins, not on the pcb pads
QFN, BGA:
2. you need a LOT of flux, the chip must swim in the flux, flux must stay there during soldering, so set the airflow such it does not blow the flux off
3. put your pcb on a ceramic support, heat the pcb up before soldering, lot of flux
4. then put the chip on the pcb and still heat, maintain a lot of flux, change the temperature (heating profile) such you change the air gun distance from the pcb, move the gun slowly down and up
5. gently (0.2-0.3mm) push the chip from a side - when the pins are soldered the chip moves back by wetting forces
6. the air flow and the temperature have to be adjusted - needs some training.

Thanks

But in this case it was a LQFP package but still valuable information.

[iMo]

I added the LQFP too. Btw the easiest soldering is with BGAs

[TomS_]

I heated the board so that the solder on the pads was molten when I placed the chip on the board so every thing should be hot and toasty.

Hmm. So you didnt have the chip on the board at the same time? Have you tried with the chip on the board while heating it up as well, so its leads also get cleaned by the flux, and also so the parts heat up at the same rate? Placing a cold part in to hot solder doesnt sound like a great idea.

I also wonder if maybe you need more solder on your pads? If there isnt enough, perhaps it doesnt have anything to "float" on in order for surface tension to align it.

[wraper]

Liquid flux as in flux pen usually is hopeless for hot air. You need tacky (gel) flux. But don't buy Chipquick SMD291, it's junk.

[iMo]

Even the gel flux liquidize with hot air, therefore you have to find a balance between airflow speed and its temperature such the flux is not blown off the pads/pins. That requires some experimenting. I even had a small table I made with air speed, temperature and distance from the pcb, trying to mimic a profile. The pcb temperature is important as well, you have to avoid cooling it from the bottom side. Tricky.

[wraper]

Even the gel flux liquidize with hot air, therefore you have to find a balance between airflow speed and its temperature such the flux is not blown off the pads/pins. That requires some experimenting. I even had a small table I made with air speed, temperature and distance from the pcb, trying to mimic a profile. The pcb temperature is important as well, you have to avoid cooling it from the bottom side. Tricky.

Liquid flux evaporates, does not stay liquid. Liquid flux is mostly solvent with a little bit of actual flux. Also tacky fluxes have an order of magnitude higher work time.

[TheHolyHorse]

I heated the board so that the solder on the pads was molten when I placed the chip on the board so every thing should be hot and toasty.

Hmm. So you didnt have the chip on the board at the same time? Have you tried with the chip on the board while heating it up as well, so its leads also get cleaned by the flux, and also so the parts heat up at the same rate? Placing a cold part in to hot solder doesnt sound like a great idea.

I also wonder if maybe you need more solder on your pads? If there isnt enough, perhaps it doesnt have anything to "float" on in order for surface tension to align it.

Yeah since it didn't align for me and I lost patience I just gave up and left the chip not aligned properly but stuck. I later added more flux and heated it again hoping it would suck it in place.

[TheHolyHorse]

Even the gel flux liquidize with hot air, therefore you have to find a balance between airflow speed and its temperature such the flux is not blown off the pads/pins. That requires some experimenting. I even had a small table I made with air speed, temperature and distance from the pcb, trying to mimic a profile. The pcb temperature is important as well, you have to avoid cooling it from the bottom side. Tricky.

Liquid flux evaporates, does not stay liquid. Liquid flux is mostly solvent with a little bit of actual flux. Also tacky fluxes have an order of magnitude higher work time.

So for hot air I should get some better flux, preferably some gel stuff? I'll do some research on that.

[wraper]

So for hot air I should get some better flux, preferably some gel stuff? I'll do some research on that.

Anything other than tacky/gel flux will be spent before solder melts.

[iMo]

This gel is pretty popular, there are tons of 223 cheapo fakes, I've done with the fakes too and they worked fine as well.
You need also a bottle of IPA (>=95%) for cleaning the mess up.
https://www.tme.eu/en/details/armalf223tf_35/fluxes/amtech/rma-223-lf-tpf-35g/

[viperidae]

The key is to maintain a LOT of flux all the time - it always works fine when there is a LOT of flux, chip must always swim in the flux.

As Louis Rossman says, the bigger the gob, the better the job.

[jhpadjustable]

So 1 rossmann = how many cc of flux? 

[wraper]

The key is to maintain a LOT of flux all the time - it always works fine when there is a LOT of flux, chip must always swim in the flux.

As Louis Rossman says, the bigger the gob, the better the job.

Louis Rossmann cannot solder properly and he admits it. He uses 50 times more flux than needed. If you have good tacky flux, only a tiny bit of it is completely enough, just enough to flow over the pads. Which it does by itself when heated.

[KL27x]

But how do other people do it? What am I supposed to do to prevent all the flux from burning to early?

I have used MG Chemicals liquid rosin flux to hot air reflow hundreds of QFN, using lead solder to pretin the pads. I find

1. You need enough flux that it bubbles and the chip wanders away on its own, so this is annoying
2. I would pin the chip down with tweezers while heating it, and this solves the problem. So it's not really that annoying.
3. I have not "burned" the flux? I mean, I don't know what kind of flux is in your pen, but I agree with others that a pen doesn't usually lay down much flux. I would put a bead of rosin flux down, with a prominent meniscus, using a syringe.

I think Luis Rossman (slash AvE; why am I not surprised he watches AvE) is mostly right re "the bigger the gob." I have gotten feedback before about an issue where excess undried flux caused failure in production testing. It is a problem that goes away by itself in about 24 more hours. If the PCB is made on Monday and goes into testing on Tuesday, it might still be too wet. Now, if you used too much gel flux and you don't clean the residue, maybe it takes another week to dry out on its own? You can also fix this issue faster by touching an iron to the flux for a bit to dry it out.

Whether you need gel flux or not might depend. I've heard that gel flux is supposedly designed for longer reflow/working time. I don't know if this is true in all (or even any) cases, though. It might be especially desirable to have longer working time if the board has a very high thermal mass/conductivity. This might be something to suspect is your case, but you/OP desoldered the chip in the first place, presumably without a notable problem.