PREVIOUSLY ON THE LAST POST…
Rushing back right after work, I made up the list on how to troubleshoot the PCB. Although this might not be a big deal, it does make me worry. This is the board which I playfully flexed, because it is only 0.8mm thick. At this moment, I want to suspect the 4017 IC but I want to make sure first.
And here we are, the troubleshooting List. This is my punishment for skipping a step:
1. Check on the health of the eight green LEDs
2. Check on their orientation
3. De-solder all of them and check the LED solder pads for 4017’s signals.
4. Check all the tracks for hidden shorts or broken links
CHECK ON THE HEALTH OF THE EIGHT GREEN LEDS: OK
Using the multi-meter, I have confirmed that all the eight green LEDs are OK. With the exception of some working LEDs, they had to be soldered since I missed them on the first round. Surprisingly, they still lit up though.
CHECK ON THEIR ORIENTATION: OK
While doing the health check, I discovered another helpful visual clue:
If any of the first four LEDs are soldered wrongly, the left and right pair would not have lighted up. In this case, the first four (on the left) actually did respond and so, I did not think polarity was the issue.
Do note that for the horizontal SMD LEDs, their Negative side are all facing the left.
DE-SOLDER ALL OF THEM AND CHECK THE LED SOLDER PADS FOR 4017’S SIGNALS: OK
With the two green LEDs flashing alternatively, I know that the 4017 is working. But these two LEDs are only getting the signals from two out of the four outputs. Using both meters and the LED method, the 4017 is indeed working. It IS getting the pulse from the 555 and giving out the pulse at its output pins.
So, yeah, it was bad on my part to de-solder and re-solder. By now, the sticky board is full of resin traces and fingerprints. A splash of IPA and kitchen towel helped to clear up the mess because I know, there is more de-soldering and soldering.
And for this step, the 4017 actually survived that.
CHECK ALL THE TRACKS FOR HIDDEN SHORTS OR BROKEN LINKS: OK
I de-soldered the 4017 and the non-working LEDs. Using the multi-meter, I tested for broken links and it all checks out. There is one thing I need to remember, which is that the metal probes on my multi-meter needs to be clean so that I can get reliable results. In the meantime, I have to jab the solder point many times until I remember I have to clean the probe’s contacts. I am lazy, and it is not necessary but hey, I need this PCB to work.
THE MOST PROBABLE SUSPECT
So, why were the LEDs not lighting up when it is receiving the signal from the 4017 IC chip? It’s not even shor…… ting.
Right after I de-soldered and re-soldered one of the LEDs and it worked. This tells me that I need to look at the PCB’s track again, much more closely. When a PCB board is designed to use both sides, copper tracks from one side physically and electrically connects to the other side using what we would call as “vias”. This is a physical wire linking both sides through soldering. Nowadays, it is shaped like a tiny metal tube which is described as though-holes.
Because I am hand soldering and I was not using resin flux, I had to use more solder. The excess solder eventually clumped into the vias which I needed to remove. They create small solder bumps, which can cause problems when I re-solder the SMD LEDs again. Not only that, there is a high possibility that the solder might melt and SHORT the LEDs.
Sadly, this theory is more or less true. The vias can cause complications.
The other LEDs did not encounter this issue as which was a surprise to me.
FINALLY
And so, after much trouble-shooting, I guess I need to be more careful when soldering the LEDs which has vias under neath them. Anyway, you can see the video below. The LEDs are very bright although they were (calculated) to be running on 10mA. A typical maximum current these LEDs can take before affecting its Lifespan is 20mA. So, even at hald the current and at 3.3 volts, they are very bright. But lets just wait until the whole Tricorder is painted and have its sticker put on.
