Solder paste flux residue after reflowing

[jpep]

I have to assembly a relative high number of boards using SMT, so I decided to give reflowing a try and purchased a cheap T962 reflow oven. After some testing the reflow process seems to be working fine, but I'm getting some ugly flux residues.

The residue is solid and looks like solid flux rosin. It is crystallized and when pressure is applied it cracks. I have attached some pictures for you guys to see it. The solder paste I'm using is Sn63/Pb37 XG-Z40 from mechanic (It's probably not the best solder paste in the world...).

The thing is that I'm having trouble to clean the residue. I have tried the usual isopropyl and brush to try to clean it without any luck. I even tried using acetone which didn't work neither.... I could try using some other solder paste brand, but maybe someone knows a proper way to clean it.

I was also wondering if this could be due to not applying the popper amount of temperature during the proper amount of time to allow the flux to evaporate, so I reflowed the paste using a hot air instead, and well... some more of the flux can be evaporated, but it still leaves that nasty solid flux residue.

Do you guys have any suggestions to clean this or to reduce the amount of residue after reflowing?

EDIT: Last two pictures show the cracked residue after applying pressure with tweezers.

[dave_k]

Try a better cleaner like CHEMTRONICS ES835BE Flux Remover.

Or, you might be better off ditching your current solder paste and using a different type.
Look for paste that uses aqueous flux, or is labelled as NC (no clean).

[M4trix]

Some pads are not completely covered with tin. Perhaps the board is tinned with immerse tin ? 

[jpep]

Try a better cleaner like CHEMTRONICS ES835BE Flux Remover.

Or, you might be better off ditching your current solder paste and using a different type.
Look for paste that uses aqueous flux, or is labelled as NC (no clean).

I will try that cleaner thanks for the advice. Do you have any specific recommendation for solder paste? all the ones I have seen are really expensive

Some pads are not completely covered with tin. Perhaps the board is tinned with immerse tin ?

It's a home etched board, and I applied the solder paste randomly just for testing so ignore my poor technique for applying solder paste for now 

[GBowes]

The Mechanics solder paste is not great. But I use it successfully.
I use IPA to clean my boards and have only rarely had resistant patches of baked flux that need scraping and then further cleaning with IPA.
I brush on IPA with a very wet toothbrush... let it sit for 30-60 sec and then wipe it with a lint free cloth.
I repeat this process several times but lay the cloth over the PCB and use a dry toothbrush to soak up the resin and IPA.

I notice a couple of issues on those boards that might help.

1. As someone else mentioned, you are not getting good flow over the whole pad. In addition, to the flux issue, you may also encounter problems with tombstone caps etc.
2. It appears that generally you have way too much paste on the pads. The problems with this are exacerbated by #1.

To improve this process, I think you need to look at your reflow cycle.... You are not allowing enough time for the solder paste flux to do its job.
You also may be too hot and too long in the solder phase. Mechanics paste has a fairly low melting point and that flux tends to bake on if it hasn't evaporated. I usually limit my peak temperature to about 200°C. (measured on the PCB)

For #1, you need to make sure that your PCB is perfectly clean. Then, with home brew boards I usually flux the entire board with a flux pen before applying the solder paste.

For #2, you need to reduce the amount of solder paste. Remember, the more paste, the more flux but of more concern is that some of the extra solder may end up floating the passives high and then shorting underneath them.

Graham

I see you just posted an response.
Some of your problems are aggravated by the things you suggest we ignore, but some of the above suggestions should still help.

[M4trix]

Well, it's not about poor soldering. My concern was about liquid tin. I made a batch of PCBs and tinned the pads with Seno 3211 but after reading some articles where they mentioned a short shelf life ( couple of months ) ... I freaked out ! I should have asked you, is your PCB tinned with such liquid and was it lying around for some time ? I have also posted in this thread regarding the same issue --> https://www.eevblog.com/forum/beginners/pcb-building-help/msg1268056/#msg1268056

Btw, you could try this cleaner for flux --> http://www.tme.eu/es/details/pcb-cleaner-a-1l/productos-de-limpieza-y-mantenimiento/ag-termopasty/zmywacz-pcb-alkoholowy/   

[jpep]

The Mechanics solder paste is not great. But I use it successfully.
I use IPA to clean my boards and have only rarely had resistant patches of baked flux that need scraping and then further cleaning with IPA.
I brush on IPA with a very wet toothbrush... let it sit for 30-60 sec and then wipe it with a lint free cloth.
I repeat this process several times but lay the cloth over the PCB and use a dry toothbrush to soak up the resin and IPA.

I notice a couple of issues on those boards that might help.

1. As someone else mentioned, you are not getting good flow over the whole pad. In addition, to the flux issue, you may also encounter problems with tombstone caps etc.
2. It appears that generally you have way too much paste on the pads. The problems with this are exacerbated by #1.

To improve this process, I think you need to look at your reflow cycle.... You are not allowing enough time for the solder paste flux to do its job.
You also may be too hot and too long in the solder phase. Mechanics paste has a fairly low melting point and that flux tends to bake on if it hasn't evaporated. I usually limit my peak temperature to about 200°C. (measured on the PCB)

For #1, you need to make sure that your PCB is perfectly clean. Then, with home brew boards I usually flux the entire board with a flux pen before applying the solder paste.

For #2, you need to reduce the amount of solder paste. Remember, the more paste, the more flux but of more concern is that some of the extra solder may end up floating the passives high and then shorting underneath them.

Graham

I see you just posted an response.
Some of your problems are aggravated by the things you suggest we ignore, but some of the above suggestions should still help.

Thanks for the detailed response. I will try with a more precise application of the paste. I will also definitely make some more tests to try to improve the reflow cycle profile and will let you guys know about the results.

Well, it's not about poor soldering. My concern was about liquid tin. I made a batch of PCBs and tinned the pads with Seno 3211 but after reading some articles where they mentioned a short shelf life ( couple of months ) ... I freaked out ! I should have asked you, is your PCB tinned with such liquid and was it lying around for some time ? I have also posted in this thread regarding the same issue --> https://www.eevblog.com/forum/beginners/pcb-building-help/msg1268056/#msg1268056

Btw, you could try this cleaner for flux --> http://www.tme.eu/es/details/pcb-cleaner-a-1l/productos-de-limpieza-y-mantenimiento/ag-termopasty/zmywacz-pcb-alkoholowy/

Do you know if that cleaner is just alcohol or does it contain any other components? I have been searching for it, but I don't understand polish and the only english datasheet I found didn't say anything about its composition...

[M4trix]

Do you know if that cleaner is just alcohol or does it contain any other components? I have been searching for it, but I don't understand polish and the only english datasheet I found didn't say anything about its composition...

It's on the page 2 & 3.

Isopropyl alcohol 40-60%
Hexane 8-14 %
Acetone 15-25%
Pentane 5-12%

Edit: oops, I see they don't ship it to Spain --> El producto que desea no está disponible para la venta en su país. ( The product you want is not available for sale in your country. )

 

[boffin]

I have a little spray bottle that I keep filled with 50/50 Isopropyl/Nail Polish Remover (acetone).  Not aggressive enough to harm plastics, but much better removing flux than Isopropyl by itself.

[Ian.M]

From the composition of the flux remover M4trix posted, if you cant get the good stuff locally, it would be worth trying 12% naptha (lighter fluid), 25% Acetone, 77% IPA. 

Caution: nail polish remover often has additives like Lanolin to make it kinder to the skin. You need pure, reasonably dry Acetone, and the best bet for that is a specialist paint or fibreglass materials supplier.

You can also try putting the board in a tightly sealed polypropylene food storage container with a little more IPA or flux remover in the bottom immediately after spraying, so it cant evaporate and has longer to work on the flux.  Leave it a while and the flux should be soft and easily soluable.

[wraper]

Flux residues from mechanic paste are very hard to remove. Even when cleaned with acetone in ultrasonic bath, there will be some leftovers. Not to say nearly impossible to clean under the parts. So best would be to just leave them as is. When you apply correct amount of paste, there will be much less of them compared to your pictures.

[jpep]

Sorry for taking so long on updating, but i had to find some spare time in order to finish modding my reflow oven and make some changes into the firmware code to suit my needs and be able to choose between which combination of probes I want to use during the reflow process.

Now with the results.

About cleaning baked flux, I got some of the products you guys recommended and these are the results I got going from best to worse results:

• Chemtronix ES896BE: I think I got a sightly stronger version than the one recommended by @dave_k and it gave me somewhat successful results, but I had to rub a lot...
• http://www.tme.eu/es/details/pcb-cleaner-a-1l/productos-de-limpieza-y-mantenimiento/ag-termopasty/zmywacz-pcb-alkoholowy/ : recommended by @M4trix. I wasn't able to remove the baked flux with this one
• IPA + Acetone: Nope, It didn't work
• IPA: It didn't work either

@GBowes gave the best advice, and after setting the reflow oven to use only the readings from the temp probes set onto the pcb, and configuring a softer reflow profile with 200°C at peak, the flux didn't crystallize and was cleaned easily removed using IPA  . There is a picture attached of nice and clean solder for you guys to see.

TL;DR I didn't find a proper way to clean crystallized flux, the key was to soften the reflow curve to avoid flux getting baked, that way it can be cleaned easily with IPA.

[wraper]

200^oC is too low for reliable reflow. Will work on tiny boards but you will have problems on larger boards due to hot/cold spots and different thermal mass of the parts. What happened, I would guess that flux did not react completely as intended. As you won't be able to remove it from under SMT parts by just rubbing with solvent, it may have negative consequences such as current leakage and increased probability of corrosion in long term.

the key was to soften the reflow curve to avoid flux getting baked

Do not forget that you intentionally fail to do recommended soldering profile which may result in lower reliability solder joints. Especially considering that all modern parts are lead free and have higher melting point of lead-free solder finish. At such low temperature, lead free finish likely won't dissolve in leaded solder which may result in fragile solder joints.

[jpep]

200^oC is too low for reliable reflow. Will work on tiny boards but you will have problems on larger boards due to hot/cold spots and different thermal mass of the parts. What happened, I would guess that flux did not react completely as intended. As you won't be able to remove it from under SMT parts by just rubbing with solvent, it may have negative consequences such as current leakage and increased probability of corrosion in long term.

the key was to soften the reflow curve to avoid flux getting baked

Do not forget that you intentionally fail to do recommended soldering profile which may result in lower reliability solder joints. Especially considering that all modern parts are lead free and have higher melting point of lead-free solder finish. At such low temperature, lead free finish likely won't dissolve in leaded solder which may result in fragile solder joints.

Thanks for the advice, I will keep trying rising temps and checking the results.