Hand soldering a 6x6mm QFN (28pin)

[treez]

Hi
Do you think its possible to hand solder the 28 pin , 6x6mm QFN version of the PIC16F18856 by hand soldering. Eg…dragging solder along the pins, then mopping the excess up with braid?

Page 654 of the datasheet…
https://ww1.microchip.com/downloads/en/DeviceDoc/PIC16LF18856-76-Data-Sheet-40001824D.pdf

 

[KlausST]

Hi,

Yes, homebrew soldering with a stencil, solder paste and a heating plate.

But if you are asking about a soldering iron with a single tip....it will be far more difficult and unreliable.
Especially "3: The bottom pad of the QFN/UQFN package should be connected to VSS at the circuit board level." ... from page5 is difficult to satisfy.

Klaus

 

[FvM]

Don't necessarily need stencil, applying solder paste with a syringe is possible as well. Reflow by hot plate or hot air.

 

[treez]

Hi,
Thanks, I think we will just have to end up trying to solder it without soldering the centre pad, as some of the small pads of the QFN achieve a ground connection anyway.
The footprint pads emerge 0.3mm outside the QFN body outline all the way round.
We have got a hot plate but I am not convinced a great deal of heat will conduct through the 4 layer FR4 PCB. Some other components on the top side are already placed, as we needed to run the board manually (without the micro) first.
I don’t think we can find a syringe fine enough to deliver solder paste onto each separate pad.
We were wondering about just dolloping solder paste over the general pads area, and adding loads of flux, then put the QFN on, then just hot-airing it and hoping that the solder will un-short-circuit itself from between the pins…..and hope that the hot-airing doesn’t damage the PIC16F18856.

 

[betwixt]

Your board doesn't have a connection to the center pad anyway!

Assuming this is for test/prototyping and not a production run, I would opt for hot air and minimal solder, including the pad. You might get away with the "flux and flood" technique but it is risky and you should thoroughly wash the board afterwards to remove residual flux and any paste still mixed with it.

Brian.

 

[KlausST]

Hi,

It seems on your PCB the center pad is not connected to GND..

I don't think using loads of solder paste is a good idea.
Maybe use the syringe to apply the oaste and a blade (scalpel) under the microscope to move it more precisely.

There are stickers that indicate the temperature. https://uk.farnell.com/search?st=Temperature inducating strip

Preheat the PCB to about 100..150°C for the flux to activate and the solder to dry. Maybe for a minute or so. Then heat up to something over 200°C, depending on solder.

Klaus

 

[FvM]

As already mentioned, the exposed pad should be connected to VSS.

QFN ML package is relative large technology and can be easily reflow soldered with hobby and small business techniques. Fortunately MSL 1 with unlimited floor lifetime.

 

[treez]

Yes thanks to all for pointing out that vss and centre pad should be connected...fortunately, their are ground pins on the edge so we will just use them.
The chip will still work without the centre pad being connected...(?)...(though it was admittedly a mistake not to connect it)

This video seems to encapsulate what the kind contributors here have said....

https://www.youtube.com/watch?v=c_Qt5CtUlqY

i wish he had said what he cleans the burnt up flux off with.....i always use IPA but dont know if its best.
I like the look of that "clear jelly" type flux he uses.....i have no idea where you buy that....(any ideas appreciated)
It seems that "water soluble" and "no-clean" flux's arent very good...it has to be the "rosin" type flux(?)

Also, i reckon we will need to clean up quite a bit, as i am anticipating ruining about 5 or 6 QFN chips before i get this right.

We will put kapton over surrounding components so they dont melt off.
The video says we can just leave off doing the center pad so we will not solder the centre pad.

 

[KlausST]

Hi,

Don't stress.
This is not a "very" diffucult to solder device.

I even don't see the need for extra flux, since a good solder (wire as well as paste) schould already contain it.
I've hand soldered a lot of similar chips ... even handbrew soldering of a more than 250 pin BGA with 0.5mm pitch.

I usually solder it with minimum solder, then clean and review it, then add flux just for the corrections, since the original flux is worn out.

No rocket science.

To protect already soldered devices you may use WELLER spot mask, or similar.

Klaus

 

[treez]

Thanks,
Also, i sometimes wonder if , since i left 0.3mm pad extension outside of the body, then maybe we can use this soldering technique...
https://www.youtube.com/watch?v=BvhE16vBfX4

However, the PIC16F18856 datasheet (in the top post) does not make it very clear if the outer bits of the pads are actually electrically connected to the underside bit of each pad.

 

[betwixt]

.i always use IPA but dont know if its best.

Use "ULS" (UltraSolve) from Electrolube, Farnell sell it.
It is expensive but I find a 1 litre bottle lasts me for about a year of flux cleaning. It is also useful as a degreaser and if you make your own photo resist PCBs, it removes the exposed resist paint cleanly.

Brian.