20231113 Re-Flow Soldering on the Razor Crest PCB

20231113 Re-Flow Soldering on the Razor Crest PCB

Finally, with some luck on my side, I am left alone for the afternoon. This also means I have access to the important large table. You see, I have not done anything to reduce the size of the solder paste stencil but rather, left it as it is.


When you hand solder components, the solder you use would be in the form of a thin malleable wire of various thickness. But for re-flow soldering, the solder comes in the form of a grey paste. If you look at the paste with a microscope, you would see thousands of small little solder balls suspended in flux.

Solder paste on the tip of a toothpick courtesy of imgur

The Solder paste I am using is Type 4 from RELIFE. It claims to have good wicking and melts at 183ºC. I am not sure why I used Type 4 but I do remember it is great for SMD components.

This is a RL402 Solder paste from RELIFE and has a melting point of 183ºC.
The small bottle contains 40g of the paste.


Just like silkscreening, you use a squeegee and drag the ink/ paste across the stencil. The process is very easy and you can use any object as long as it has a straight edge. For me, I bought some yellow plastic ones which is meant for car window tinting.

Very quickly, I learnt that there is a technique to apply the paste as it was not a simple as I thought.

Before I start, I cleaned the PCB with IPA (Isopropyl Alcohol)
Using a wooden skewer, I deposited some paste onto the stencil which I have prepared much earlier.
Then I dragged the yellow squeegee across the stencil.
The result was quite bad because I dragged the squeegee on a few directions and with an angle
This made the already deposited paste stuck lift itself off to the moving squeegee.
Wiping the PCB clean again, this time I swept across with the squeegee held upright while pressing down on the stencil.
This results in most of the solder pads filled with the paste.
When I swept the squeegee from other directions, the paste seem to leaked out.
The excess solder paste, which I found out later on, is not an issue because it does not seem to be enough.
This is how the PCB looked like after I populated it with SMD components. I did not notice this until when I did the video as LEDs #12 & #16 have toppled while I forgot about #21.


To prevent thermal shock to the PCB, I placed it on the hot plate as soon as it starts to warm up. It took quite some time for the hot plate to reach required temperature. There were quite a lot of fumes coming off and once the solder paste starts to melt, it means the peak temperature has been reached. The PCB is then let to ‘cook’ for about 18 seconds before the hot plate starts the cooling down process.

20231113 Re-Flow Soldering on the Razor Crest PCB
After the reflow.
The main reason I used this process is to make sure the 6 network resistors have been soldered correctly. In the past, I had to rework them many times and because I bought the wrong type (instead of concave leads), hand soldering is a chore. Oh, and the 0603 LEDs are a pain as well.
One thing I noticed was that the melted solder paste was not shiny in certain point and there was a lot of un-melted solder balls.
Maybe I should have left the PCB on the plate longer, like 40 to 60 seconds.
The Clean up
The PCB was cleaned up with few passes of IPA (Isopropyl Alcohol) to remove excess solder balls.
You can see some of the pads do not have enough solder. Anyway, it still needs more scrubbing with IPA to remove stubborn residue.
A quick test and the 0603 LEDs are working. All except #08 (wrong orientation) and #21 (no LED)
Testing with the Main Board
After soldering the remaining components, the PCB worked nicely with the Main Board


If I used hand soldering, this board would have taken me at least a few hours. With the re-flow, it would be significantly faster. This method gives me a much higher reliability although it has some issues:

Too much/ little solder
I cannot solely blame on the stencil’s 0.13mm thickness because there are other factors too. One possibility is the way I dragged the squeegee and the speed which leads to the paste not settling down. Plus, maybe the stencil was not sitting firmly on top of the PCB which allowed for leaks.

Solder joints
During the testing, there were two LEDs which did not light up reliably. The testing led back to the network resistors which, although has sufficient solder, did connect. There is another possibility that I allowed the liquidous state period to be too brief. This is the peak temperature where the solder has started to melt and (through capillary action) aligns the SMD components. Gravity also made sure the SMD components are sitting on their respective solder pads correctly. All these needs time. Perhaps on the next board, I should extend the period to at least 40 seconds or more.


Here is the video of the re-flow process on the PCB. It was very exciting for me because this is now an actual re-flow process and not some PCB testing anymore. Thankfully, the first time results was better than I anticipated. Enjoy!


With this issue more or less settled, it’s time to make sure the whole system works. Then I would have to start on the BOM aka Bill of Materials to see how much I would need to get the whole system manufactured. Plus there is the Assembly Manual as well.

What started off ans a simple three PCB board design has now evolved to eight…

Posted in A Piscean Works Blog, AMT, Design, Electronics, JLCPCB, Lighting, Model Kits, Printed Circuit Board, Razor Crest, Revell Germany, Scale Lighting, Scale Models, Sci-Fi, Spaceships, Star Wars, Techniques, Technology, The Mandalorian, Video.

Leave a Reply