This method of surface mounting is unconventional and is not recommended for production projects. If you plan to use the device in a mobile project make sure you seal it.
It's a know fact that the surface mount devices are becoming more and more popular leaving behind the traditional through-hole mounting method. While for the electronics industry surface mounting brings a lot of benefits (one of them being the ability to pack more components per square inch) for the regular hobyst surface mount devices are often times a "show stopper".
This tutorial is for those of you who absolutely need to mount a small surface mount device for prototyping purposes and do not have the means or time to perfom a traditional surface mount. There's obviously a "correct" way to mount a SMT component and I highly recommend reading the Surface Mount Soldering guide at Curious Inventor first before attempting this method. However it's always good to know that there's an alternative method that does not require any investments in expensive SMT equipment or materials.
In this tutorial I'll describe how to mount an analog LIS244AL accelerometer (data sheet), that comes in a tiny leadless LGA-16 package. The size of the component is 4x4x1.5mm. It has 16 solder pads, each is 0.3×0.4mm. This is the tinest component that I ever have to deal with. When I first saw this device my first thought was that there's now way I can mount this little device without creating a custom PCB and using the traditional surface mounting method.
After several trials and errors I came up with an interesting method that uses masking tape in order to create bridges from the device pads to the pcb regular "through-hole" pads. I also made my work easier by bringing the device pads to the same surface as the PCB hole pads.
No custom PCB is required for this tutorial, we'll use a regular proto-board that you can find in any electronics shop. However you can create a custom PCB to minimize the distance between the device and the external pads.
First mark the desired size on the board that you'll use for your breakout board. We'll use a piece that has 4×4 holes (two pieces are marked on the picture below).
Next calculate or measure the diagonal of your component. Our component has a diagonal of about ~5.5mm .
Using a 5.5 drill bit drill a hole in the middle of the breakout board.( it's easier to drill the hole before you cut out the board).
If you're having problems creating a perfectly centered hole and if you have a set of drills of different sizes – first drill a small hole with a small
drill than gradually enlarge it by using larger bits. Or simply use a drill press if you have one.
Make sure the component fits tightly. Use a file to enlarge the hole if necessary.
Note: You can avoid some of the next steps if you find a way to cut a square hole in the PCB. It would be easier to cut a square hole in a thinner PCB using a knife or a small chisel. Also if you can etch your own PCB you can bring the external traces really close to the device's pads so creating bridges would be really simple. However, in this tutorial we'll focus on the beginner hobbyist that has no way to etch a custom PCB.
Since our hole is round and the device is square we'll need to fill in the gaps with some filler.
Before applying the filler, first mask the side of the board that contains the copper pads. (We'll be filling the gaps with putty from the other side)
I used 2 layers of tape one to mask the pads of the device (it's darker in color) and then on top of it I put regular scotch tape over the entire board.
But you can probably do with just one layer just make sure it sticks tightly to the device so the putty doesn't get onto the device pads.
I cut the excess tape with a knife so you can't see any tape hanging around the board on the picture below. (This is not necessary, I guess I did it from inertia since I first cut the other piece of tape around the device).
Turn the board over and fill it with epoxy putty or other filler.
The choice of filler is yours – you might use a non-permanent filler if you'll need to remove the device later.
A screwdriver might be usefully to even out the putty.
Press the putty tightly so it reaches the over side.
Make sure your putty is not conductive! Some metal putty might be conductive.
Before putty dries out turn over the board and inspect the surface that the putty has formed.
It must be on the same level as your board and the device.
Remove any putty that accidentally filled the PCB holes.
Wait till the filler hardens (timing depends on the type of putty you use – read the instruction on the packaging of you putty).
When done, remove the masking tape (on next picture you can see both layers of masking tape). You must get a plain surface – the board , the putty filler and the device must be on the same plane on the copper side.
The board I'm using happens to be same thickness as the device (1.5mm) so it also creates a plain surface on the other side.
If your device is taller it might sticks out on the reverse (non-copper) side.
Now we must create some bridges from the small device pads to the PCB pads. This is done easier by using masking tape.
Notice that using 3 strips of tape we can create a channel that goes from the tiny pad on the device to the larger round pad on the proto-board. Fill the channel with silver conductive polymer ( I used Circuit Writer pen liquid).
Alternatives: Circuit Works Pen, Silver Microtip Conductive Pen
Wipe off any excess liquid from the tape before it hardens, otherwise when you remove the tape you might damage the bridge.
Also try to remove the tape as quickly as you can before the polymer hardens completely and make sure it's not stuck to the tape before you remove the masking tape strips.
Let the bridge harden, and repeat this technique to create the other bridges.
Prior to creating bridges I soldered some pins to the corners of the board so it can be easily mounted on a breadboard or inserted into a 0.3" wide IC socket.
Consider doing all the soldering prior to creating the bridges if you're using the conductive polymer. Since solder might melt the polymer and destroy the bridges.
Here is the final result mounted on a breadboard, inspect your board under magnification:
In the picture above you can see 4 bridges (used for accelerometer's XOUT,YOUT, GND , VDD).
I did not connect all pads as per device data sheet (Reset , Sleep and some other ground pads).
It seems like the device works like this at this testing stage (but I plan to connect the other pins later according to the data sheet).
Notice that some bridges, like the ground bridge (the thickest one at the left) can go over multiple adjacent pads if necessary.
If you're going to use solder to create the bridges you should use temperature resistant tape (like Kapton Tape).
Also put a thin wire inside the channel to serve as a skeleton for your bridge. Otherwise the solder might not stick to the PCB if the distance between pads is too big.
I attempted to do one bridge with solder (the masking tape you saw above was a temperature resistant Kapton tape), but got lazy – the conductive polymer seemed to be easier to work with and only cost me around $15 (the shelf life of a conductive pen is about 1 year).
Finally test your component. In the picture below I'm testing the X output of the accelerometer that delivers 1.5V in plain position – as it's supposed to. If you tilt the board you should see this value deviating up or down depending on what side you tilt the board.
Finally after testing the board you can seal it with a sealer (epoxy glue or similar or liquid electrical tape) or simply with a piece of tape for added protection.
The conductive silver polymer is not very strong so it might get scratched during handling.
Make sure your sealer will not dissolve the conductive polymer, that's why it's a good idea to first apply a thin layer of sealer and let it dry out before applying the bulk of it. The company that sells CircuitWriter also sells CircuitSealer. However I used liquid electrical tape, similar products can be found at your local hardware store.
Here is the final sealed device inserted in a DIP socket on a proto-board project:
19 thoughts on “Reverse Surface Mounting of Small Leadless SMT Components”
That is the craziest thing … that I might try soon.
You are a super-power-ninja!
Pingback: Prototyping with very, very small ICs - Hack a Day
You are my hero. I usually do the dead bug trick but I use 32 awg wire to bridge the gap between the berg and pins. This is definitely less thermal shock to the IC.
Well done! I just finished(successfully) soldering up an LGA 28. Man, was that a chore. next time I’ll be using your method. Thanks for sharing.
God what a mess why would you bother!
Pingback: Doodle Circuits Blog » Prototyping with very, very small ICs
Pingback: Belgaum news | About Belgaum | Belgaum information | Belgaum district | Belgaum city | Belgaum Hotels | Belgaum People | Belgaum tourism | Belgaum entertainment | Belgaum students | Inside facebook | Hack | make use of | technical news | | Prototyping wit
10/10 for sheer 1337ness.
Seriously cool stuff, I used superglue+baking soda a while back to do something similar but this stresses the chips a lot.
On the other hand it worked, and so far it has never failed me with all chips mummified in this fashion working to this day unless electrically fried by reverse power etc.
Nice. I have soldered an ARM in HQVN33 using enameled wire, but using double sided tape to glue the chip.
I’ve just bought an ic thats lga16 and laughed when i saw the small the pads. Thanks so much for sharing an excellent idea and saving me the frustration of trying to solder it.
Where do you get transparent PCBs?
Can you help me finding them?
Oh, I actually had no idea it is transparent until I etched the copper out. I think any thin PCB would have some transparency to it, so just getting very thin copper clad might do the trick.
Great tutorial! I was wondering what kind of epoxy putty did you use?
I used regular 2-part epoxy for plastic , you can use any type of glue / putty, just stay away from water-based ones or the ones that contain metal. It would be common sense to check conductivity of the material you use with a multimeter , you never know what’s inside !
Brillant! Lots of new cool chips are not maker-friendly :-). I've got lightning detector chip samples I want to experiment with and they came in a package similar to yours. I should try this technique.