In case yo plan to solder QFN or similar devices, the easiest method for me is to fix the device by provisional soldering to the PCB on two opposing corners first. Then flux all terminals. Use a „gullwing“ shaped solder tip that is pre-tinned. This will both tin the solder joints and at the same time removes excess solder and bridges. A little practice is needed, but then it is pretty simple. Inaccessible solder joints are a complete different story.
Just in case anyone is interested on this... Recently, I took my chances with LT3042. It can't be much smaller than that! What I found to work nicely was to tin the pcb pads with a soldering iron -just move the tip over the pads and they will hold the necessary amount of solder- then put the chip in place and heat it with hot air. For LT3042 in particular, there is a power pad underneath so I put a drop of solder paste which held it against the air. Very clean and repeatable procedure! I soldered 8 chips in a row like a pro!
I use a 7mm nozzle with air pressure at minimum which I found to work best at least with my soldering station. I set temperature at 400 C. The pcb will see roughly half of it. No preheating. It takes less than 30 sec, I haven't counted but you'll see it happening. No parts loss so far but for reworking -desoldering, resoldering-I would suggest lower temp.
I find SMD soldering to be unreliable a lot of the time.
I carefully clean pads with liquid solder flux.
I then run some paste along the pads and beneath the device to stick it down.
I then tack each opposite corner.
Then put solder on iron and run it along the sides of the pads and pins.
Sometimes it works really well and the joints look very clean and good.
Other times it looks ok but when I power up the device it isnt working and I have re-solder the pins as a joint must have been bad.
Trying to use a heat gun and solder paste just ends up with too much solder shorting out pins and I have to wick the excess off which of course removes most of the solder !
I carefully clean pads with liquid solder flux.
I then run some paste along the pads and beneath the device to stick it down.
I then tack each opposite corner.
Then put solder on iron and run it along the sides of the pads and pins.
Sometimes it works really well and the joints look very clean and good.
Other times it looks ok but when I power up the device it isnt working and I have re-solder the pins as a joint must have been bad.
Trying to use a heat gun and solder paste just ends up with too much solder shorting out pins and I have to wick the excess off which of course removes most of the solder !
I had another go at soldering a SMD SOIC package today.
This time I took a bit more care.
I cleaned the pads first with liquid flux.
I then tacked down the IC on each side.
I then ran some flux paste along the pins.
I the ran soldering iron with a bit solder on it along the pins.
This time it came out very well.
I think previously I ignored the first rule of soldering and didnt use enough flux on the pins before soldering.
Running solder along unfluxed pins is obviously looking for trouble.
This time I took a bit more care.
I cleaned the pads first with liquid flux.
I then tacked down the IC on each side.
I then ran some flux paste along the pins.
I the ran soldering iron with a bit solder on it along the pins.
This time it came out very well.
I think previously I ignored the first rule of soldering and didnt use enough flux on the pins before soldering.
Running solder along unfluxed pins is obviously looking for trouble.
For prototyping doing it the same way as Lee Knatta , adjusting the IC and holding it in position by pressing a finger on it while soldering the two pins for fixation, then solder the remaining pins.
Using a solder paste with higher visocosity helps during positioning and holding it in place (still more safe to press the ic down with a tool of finger during the soldering) or using an adhesive made for this purpose (like Loctite 3609 ) which is convenient if you want to place some more in a row and solder the pcb later.
Last edited:
Thanks, that's how I do it too. But for TSSOPs (0.5MM pitch) the initial soldering always shorts at least two pins, which is what I want to avoid.
My hands are too shaky and I cannot pin down a chip for a while without minor vibrations - which ruins the position I have to adjust under scope or a good magnifying glass.
Thanks for the adhesive type. It is actually quite expensive for DIY. Please does anyone have any suggestion for a well-tested less expensive adhesive which allows fine-tuning the chip position under a microscope yet fixates the chip enough to allow moving the board away from the microscope plate? I very much appreciate any suggestions, thanks.
That's what solder wick is for.
I've seen other techniques where you blob solder across all the pins and then wick off the excess. There's always plenty of solder left to make the connection, assuming your part was flat on the board to begin with.
I do use the wick to remove the shorts, actually I use the method you mention, removing lots of shorts with the wick. But I find the wick method rather unreliable, having to check all joints with a microscope and multimeter with needle probes after the wicking, sometimes a tiny ball of solder burried in the flux still keeps shorting the 0.5mm pins. I would love to avoid the shorting in the first place to improve reliability and reduce the assembly time.
These are techniques my colleagues and I use:
Work under a microscope (e.g. Stereo Zoom, or similar). HUGE help in reducing the shakes!
Pre-tin PC pads with 60/40 rosin-core solder and use solder-wick as needed to get quasi-uniform solder thickness. Don't clean flux from the site--- the semi-cooked flux makes a nice, sticky glue to aid positioning the component.
Choose a corner lead for first solder joint, using tweezers to hold IC in position. Once first joint has cooled, over to diagonally opposite pin. If misalignment is chiefly angular, and less than a pin position in error, I simply twist the chip into correct position while soldering the second joint. Once the second joint is well positioned, I often remelt the original joint to release any captured torque stress. Soldering the remaining pins is comparatively easy.
Re-emphasizing colleagues and I find the microscope essential for success!
Work under a microscope (e.g. Stereo Zoom, or similar). HUGE help in reducing the shakes!
Pre-tin PC pads with 60/40 rosin-core solder and use solder-wick as needed to get quasi-uniform solder thickness. Don't clean flux from the site--- the semi-cooked flux makes a nice, sticky glue to aid positioning the component.
Choose a corner lead for first solder joint, using tweezers to hold IC in position. Once first joint has cooled, over to diagonally opposite pin. If misalignment is chiefly angular, and less than a pin position in error, I simply twist the chip into correct position while soldering the second joint. Once the second joint is well positioned, I often remelt the original joint to release any captured torque stress. Soldering the remaining pins is comparatively easy.
Re-emphasizing colleagues and I find the microscope essential for success!
Last edited:
BSST, thanks. I use exactly the same procedure, including the stereo microscope. I just hoped there would be a faster less laborious method while maintaining the result reliability 
Something like a thick flux or inexpensive adhesive which would allow correcting the position under the scope and avoiding the initial fixture soldering which basically always requires wicking the shorts afterwards.
Something like a thick flux or inexpensive adhesive which would allow correcting the position under the scope and avoiding the initial fixture soldering which basically always requires wicking the shorts afterwards.
BSST, thanks. I use exactly the same procedure, including the stereo microscope. I just hoped there would be a faster less laborious method while maintaining the result reliability
I use a thick flux paste under the pins but also some under the body to hold the IC down while gently solder first pin and opposite pin.
I also found the flux paste better while drag soldering than the thin stuff which just runs off.
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.