Update 8/6/20: Grades for the 5V Charger were posted today. I got a 190/200 (95%) on the final project, giving me an A (overall) in Jr Design. If I hadn’t forgotten to show the PCB partially inserted in the enclosure, I would have made a 100!

Update 7/30/20: Revision 2.2 arrived in the nick of time, 8 hours before the final report was due! Luckily, everything went together smoothly and worked on the first try. Below are some pictures of the final assembly and testing.

5V Charger Final Assembly soldering
Final Assembly Soldering

Update 7/24/20: I had a few stupid mistakes on that first revision (that is pretty much expected on Rev 1.0). It was nothing huge, just annoyances. The main problem with the design is the 50mm power plane/trace, is going to the wrong pin on the IC (should be #6, not #7). Could be bypassed with a copper solid-core wire, wouldn’t be nearly as neat though.

Hallman Labs first two PCB revisions compared and discussed
Comparing the two designs revisions

Other updates include: Pad size increased on all components | Spacing adjusted for grounding “block” | SMD Pads added to bottom layer (USB shielding to daughter-board) | Milling on bottom of PCB (for daughter card, instead of doing it by hand) | Labeling/Signature Added to Silk | Polarity of how C2 is shown, fixed | Increased all trace widths, to at least 20mm, versus 10mm (required).

Here is the hand-milled, V1.0 PCB, temporarily attached to the daughter-card/board:

This is a 5V USB Car Charger, for plugging into a cigarette lighter and charging your phone (or any 5V USB device). It’s simple, yet effective, at teaching PCB design. The trick to this design is fitting the (optional) AC ripple filter, which I did. Most did not even attempt to do this, due to the insanely tight space constraints.

Once this assignment is graded and the Summer session is over, I’ll provide more of the technical details. If interest is shown in seeing any of it, such as the custom parts library, or the board + schematic files (for Eagle).

However, I will not show the enclosure or reveal the MPN for the stock part, to prevent any cheating/copying (by any students).

Update 7/21/20: When I went to solder this, I (finally) saw that there is an error with the power plane (which is just a wide trace, in this instance). Pin #7 (where the 30mm plane goes) is IPeak and Pin #6 is Vcc. Pin #6 is the place I intended to send the large trace/plane. (Worth pointing out) It will likely function just fine like this, as it’s only 12V-14V involved. However, this was part of the assignment, so I have to remedy it. It would definitely change the impedance of V_IN to Pin #6 (from the optimal performance seen in MultiSim), if leaving it like this.

There is only one way I know of to “fix” this. Due to the interconnects on Pin #7, I can’t change IC wiring. Even if I could, that would be more involved and requires scraping PCB to expose a new (DIY) “pad”.

All I am going to do, is cut the stock trace and then add in an appropriately gauged, solid-core wire. I will connect at the top of the 0.33 ohm resistor, i.e. on the V_IN side. You could also run it directly, from V_IN, if you have the room.

Stock PCB design on the left and my PCB design on the right

My first custom designed PCB, for Jr Design. ❤❤❤ Every millimeter of the shape, every layer (copper ground plane, for instance) and every line of silk screen, completely made by hand.

I decided to go with all through-hole components for durability, due to chargers being abused, often. Obviously, I could have picked SMD components, for each and every one of these parts. At that point, the durability of the design becomes much less, in the long run.

That is the problem with many electronics these days, too many SMD parts and not enough through-hole parts. Through-hole, has both, a better electrical and mechanical bond between component and PCB. This is known, throughout the ECE world and nobody disputes it. It comes down to the amount of surface area for mech. bonding (which is at least double as much, on through-hole vs. SMD), capacitance, amperage capability, and a few others, small, weird measurements. However, on average, through-hole parts cost about 30-50% more than SMD equivalents.

If this interest you let me know, I can go into more detail on why through-hole is better. I have been soldering for nearly 20 years at this point, so the fact I am just now taking Jr Design is kind of ironic. My journey has been eventful, I’ll say that.

HL Custom PCB #1: 5V USB Car Charger (Dual Port USB-A Female from Daughter Board)
My first custom PCB design. A simple 5V USB car charger, with AC Ripple filter included. Very tight space constraints and part selection (due to max height being 15-16mm).

Files were made in Eagle and exported to Gerber, for production at JCLPCB.

I can’t wait for Sr design in the fall! 🤓

The silk screen has a single error on C2 (polarity flipped on a silk screen bug), but that has no bearing on the function and the final grade.

My overall shape, is accurate to stock enclosure PCB to within 0.01mm. Since we re-use the stock enclosure and USB port daughter boards.

Here is the Bill of Materials, for my design:

Bill of Materials (BOM) for the design.
BOM (Digikey P/N Given)

It’s not required to pick such premium parts for the capacitors, resistors, or inductors, but it definitely helps the Final Report, performance data.

HL Custom PCB #1 parts layout (5V charger)
Laying Parts Out (not soldered)

Join the discussion and let HL know what you think!

This site uses Akismet to reduce spam. Learn how your comment data is processed.