Update. It turns out the issue I had with the barrel jack drill holes (discussed below), turns out to be a constraint of OshPark’s fabrication process. While this is not discussed in the design rules (link), it is clearly discussed in this support article. Bottom line – stick to circles for drill holes!
A YIPPEE moment yesterday…delivery of a purple envelope containing 3 Healthy pH Dev-Rev2 PCBs… so….how’d it go? What works what does’t? As a process check – how long was the turn around time from Gerber submission to having the PCBs in hand?
The goal of this post is to start the debug process of the Healthy pH Dev-Rev2. This post will focus on the Voltage Regulator Circuit.
Thanks to Those That Went Before
- Chris Gammell – I took Chris’s unique/fantastic Contextual Electronics course. It is unique because it truly is like going to the deep end of the pool, jumping in, and then having to figure out how to swim. Only, Chris is swimming right along with us showing how to move forward and what pitfalls to avoid. Since taking this class, I have benefited greatly from his additional guidance. Besides all this, Chris is passionate about electronics, how we learn electronics, and has always been a person of excellence. I recommend his courses and podcast. Not because I get any monetary benefit – because I do not. But because they have allowed me to make what I consider incredible (by my past history) progress in learning electronics.
- Laen and his company OshPark – I can’t build a prototype without a PCB. Thank goodness for OshPark and how simple they make it to submit Gerbers. In addition, their service and support is excellent. Not only does the person responding to a question respond within 24 hours, but they take time to explain stuff in more detail. Also, folks answering my questions have been extremely positive – which makes for an overall very positive experience.
Cost and Turn Around Time
The more I can test and faster I can iterate on a rev means the sooner I can have a V1. The long pole in the iteration is fabrication. Typically, from submitting Gerbers to receiving the PCBs took around 15 days. OshPark offers a Super Swift service so I tried it for the rev of the Healthy pH PCB. The turn around time was cut in 1/2 for roughly 4x the cost. Besides the cost of the PCB, shipping needs to be thought through. With free shipping, it took four days to arrive. Paying $5 for a better shipping option skimmed off two days from shipping.
Takes longer, Costs Less
I sent Dev-Rev1 out on Thursday PM. I received them 15 days later at a cost of $28.50.
|3 boards at $28.50 per batch of three.||$28.50|
|United States: Free Shipping||$0.00|
Faster, Costs More
My current experience with OshPark’s SuperSwift service / USPS priority mail is 9 days at a cost of $122.50.
I sent the Gerbers in on a Monday in the afternoon. I wanted to try out the Super Swift Service as well as some for shipping to see what the time difference is.
|3 boards at $28.50 per batch of three.||$28.50|
|Super Swift Service||$89.00|
|USPS – Priority Mail||$5.00|
It took 2 days for shipping. I received the boards 9 days later on Wednesday.
According to the great folks at OshPark, panels go out each day at 11AM. If I had been able to submit the Gerbers prior to 11AM Monday, there is a good chance I would have received the boards two days earlier. Perhaps if I paid for overnight shipping I would receive the PCBs on Saturday – a total of 6 day turnaround.
Circuit to Supply a Regulated 5V
The voltage regulation circuit is in the WallWart.sch kicad schematic (GitHub link). So what worked and what didn’t? Time to solder and test.
(UPDATE: NOT D’OH but a Design constraint of OshPark’s fabrication process…USE CIRCLES NOT OVALS FOR DRILL HOLES).
This is embarrassing. AGAIN? Are you kidding me??? Now the pads are larger but the drill holes are STILL too small. **D’OH!** I double checked the holes on Dev-Rev2. Chris pointed out the darn holes still looked too small…but……I was sure I corrected this. WRONG.
Looking at the layout in PCBnew (kicad files are at this GitHub repository)
compared with the test points, they do look rather small. Um…this was the exact feedback Chris gave me on review of this layout. I clearly didn’t get this right! Here’s an image of the barrel jack part next to the pads and holes on the board:
I was able to solder wires to the barrel jack pads. YIPPEE! I got Power, which I set to 10V. The fun part was using the hot air gun to connect longer wires so that I could attach the ends to my power source. It is so much more fun to use the air gun than a lighter!
I soldered on the components (yes, I need practice!):
Here is a drawing I made of this part of the Healthy pH Shield:
Hmmm…I expected to see 5V on Vout …what’s with this punky .63V (.68V when the green LED was attached). I expected to see closer to 5V. Checking the wiring of the pins on the MC78L05ACHX Voltage regulator (data sheet).
**D’OH!** Pins 1 and 2 are swapped. Excuse me for a bit while I bang my head on a wall…clearly nothing is going to be hurt…
I skipped over using the voltage regulator by soldering a wire between pin 1 of the barrel jack and pin 2 of the Wall Wart extension pins (P2):
YIPPEE! I Saw the Light
The board works, the green LED turns on. This is great because it is a visual indicator that the power supply is on…which is most inconvenient when I’m soldering on parts!
Not very pretty, but at least I can move past the voltage regulator circuit.
There are a few additional changes I will be making to Dev-Rev3. These changes became obvious after playing around with Dev-Rev2.
Get Rid of Multiple Test points
Each additional test point adds complexity to the layout and debugging process. Multiple places to insert a probe for the same reading include
- Remove TP19 and TP10: There are three holes where a DMM or scope probe can go to get to AGND: TP19, TP10, pin 1 of P2 (closest to edge of board). Since there is a through hole (pin 1) available, I will get rid of TP19 and TP10 and use pin 1 of P2 to probe AGND.
- Remove TP8. TP8 is on the same node as the through holes for P2. Bah Bye.
To Add: Text and Logo
I have not added any additional text or art to the PCB. The text is the most important. It will be the way I can tell the differences between revs of the boards. Right now I look for the epic fails between boards. While – sadly – there are many epic fails to help identify the board, it will be easier with identifying text like “Dev-RevX.” I plan to add a sacrificial draft of a logo. Creating artwork that identifies my work is really hard so I thought I’d start the iterations now just to start getting visuals in my mind for behind the brain processing.
Closing Thoughts/What’s Next
I am glad to have the test points scattered throughout the PCB. This has made it easier to map expected values from the schematic to circuit location in the layout. I started studying electronics in January, 2014. I need more time to absorb basic circuit design concepts. I like this way of building, measuring, comparing to ideal/expected behavior. It is an incredible learning opportunity. I’m excited to press forward…although I seem to be losing out on sleep in anticipation of figuring out why something doesn’t work and then learning from it.
I plan to go through the pH schematic/circuit after this….until then…
Thanks for reading this far. Please find many things to smile about.