Clever project, well documented, some improvements suggested
For The CD4017 Decade Logic Clock Soldering Kit
Clever project with surprisingly low current consumption: 9 to 12 mA at 5V, depending on how many LEDs are on. Documentation is excellent, with the back PC board silk screen containing the parts list and a block diagram of how the circuit works! Shipping was very fast - I ordered it at night on cyber Monday, got an e-mail within 10 minutes telling me it would ship the next day, and it arrived in time for a Saturday project. The kit was complete, build went smoothly with only one issue (intermittent short between +5V and GND, discussed in next paragraph). The clock keeps time accurately, but due to lack of capacitors between +5V and GND, it does not handle power supply transients well (more details below).
The clock is meant to be built without IC sockets, but I chose to supply my own so I could perform some resistance and voltage checks before installing the ICs. I initially built the clock with the full-size USB jack, and resistance and voltage checks looked good. I installed the CD4060 chip to check out the clock circuit first, but immediately had a short from +5V to GND. I removed the CD4060 chip and the short was still present. Wiggling the USB cable connected to the jack caused the short to come and go. I traced the short to the +5V pin on the full-size USB jack. The plated-through holes for the four pins on the bottom of the jack are a larger diameter than they need to be, which allowed a fair amount of solder to flow through from the top (solder side for the USB jack) to the bottom (where the jack is installed) and form a blob on the +5V pin, just barely touching the metal case of the jack. Since the jack case is connected to GND, this caused an intermittent short between +5V and GND. The clock only uses USB for power (not data), and since I wanted a little more flexibility for power supply options anyway, I removed the jack and installed wires to supply power to the board.
I noticed that most other plated-through holes in the board are larger diameter than they need to be, particularly those for the ICs, resistors, and diodes. These did not cause any problems, just required a bit more solder than usual to get nice-looking joints, with a little more solder than usual flowing through to the top of the board. Most of the pads around the holes were fairly narrow. This also did not cause any problems, but it would have been a little easier to solder with more pad surface. If the board is ever updated, I recommend making the hole diameters smaller while maintaining the same outer pad size, to give more pad surface for easier soldering and require less solder to fill the holes.
With the clock built, I used a bench supply to power it with 5V, and noticed that it only consumed 9 to 12 mA, depending on how many LEDs were on! This got me thinking about powering the clock from batteries. Since the ICs are all CMOS and rated for a wide voltage range (3 to 18V), I tested the kit up to 6V and down to 3V and found that it worked well between 3.5V and 6V, which would be ideal for powering the kit using four 4 AA batteries, and alkaline or NiCd or NiMH would work fine! I'm currently powering the clock from a 5V regulated AC-DC supply which is shared with another project. The other project has a DC jack for the supply and screw terminals which are connected to the jack (always on), perfect for connecting the wires to power the clock. There is a power switch on the project, and I found that when I switched the project on or off, the minutes and hours on the clock jumped one or two counts. I reviewed the clock schematic and noticed that there are no decoupling capacitors, and no capacitor at the power input. First, I added a 47uF electrolytic capacitor at the power input. The holes for the micro USB jack were perfect for this capacitor. That solved the count skipping from the power transients when turning the switch on and off. However, with this capacitor installed, I noticed that the ten minute count behaved strangely - it went 1, 2, 3, 4, then two LEDs came on, then they both moved up a couple of times with the ten minute count, and then it reset to 0. I found that a 0.1uF capacitor across U2 pins 8 (GND) and 16 (VDD) fixed the problem. I also added 0.1uF capacitors to U5, U7, and U8, and the clock has been keeping great time ever since. (Ideally, there should be a 0.1uF capacitor across VDD and GND on every IC, but since it was time-consuming to solder them to the pins on the back of the board, I decided to stop there.) When there is an opportunity to update the board, I recommend adding a 0.1uF decoupling capacitor as close to each IC as possible between VDD and GND, and a larger electrolytic capacitor (10uF to 47uF) at the DC power input to the board.
A suggestion when checking out the clock after the initial build is to make use of some of the other outputs of U9 to test the clock minutes and hours counting without having to repeatedly push button SW2. Using a loose piece of wire, touch one end to the end of R4 that is connected SW2, and touch the other end to U9 pin 15 or 14 to quickly count minutes.
