Podcast: Play in new window | Download
Subscribe: Apple Podcasts | RSS
- Chris has been troubleshooting a PCB with a dead short on inner layers (put in by board house by mistake)
- Don’t Touch My Gerbers shirt
- “Is there an AI tool that will fix this for me?” … No
- Chris dumped a bunch of current in the board and looked at it with this thermal camera
- 6.5 digit DMM to track down shorts
- Etching problems in the old days
- 100% etest
- Adding rails to PCBs for production
- Reddit discussion thread: why not work on a product? That is, Dave, the wise one.
- Videos
- Live stream issues
- Post from Twitter: Is 2 layers all you really need? This person thinks so, or is trying to convince themselves as much.
- Armchair quarterbacking
- Ian Johnston replacing the display on an 8.5 digit DMM
- Jack Ganssle has posted his final newsletter (The Embedded Muse)…happy retirement! Jack has been on the show twice:
- Ward Christensen, Inventor of BBS and XModem, (and former listener of the show!) has passed away
- Dave is interviewing Lee Felstenstein for our next episode
Thanks guys for mentioning my R6581T OLED upgrade project video!…….and much thanks to MickleT over on the EEVBlog forums for the reverse engineering.
Interestingly, there’s no direct handling of ASCII characters on the STM32 code!
So, it reads a serial data stream intended for a VFD controller, intercepts it, and processes the data to display on an SSD1322 OLED. It takes each packet of incoming data, unpacks the character bitmaps and annunciator flags, and stores them in a frame buffer. The code then scales and converts these bitmaps for the SSD1322 OLED’s pixel format, handling both the main and auxiliary display lines. Finally, it sends the formatted frame buffer data to the OLED, updating the display in real-time based on the intercepted VFD data.
Hi Guys,
Love your podcast, always something interesting. However, I must take a minute to correct some of the incorrect assumptions that I heard on this one. For context as a young man, I produced PCBs for several years in the 70’s and 80’s. So, when I heard that they rolled solder on the boards and that’s why it was all wrinkly under the solder mask. I thought I should explain how it’s done. The solder was electroplated on as tin/lead (From a tank that contained Solder anodes. But it was plated as elemental tin and lead in the proper ratio). This was then used as an etchant resist and protected the copper from the etchant. After etching the boards are fluxed between flux soared rollers on the way into a reflow oven where the tin and lead are melted into the solder alloy. Goes in dull and gray and comes out bright and shiny. I skipped past the resist application and copper plating before the tin/lead plating and the resist removal prior to etching, but, my point wasn’t to explain the entire process, but if you would like me to follow up, I can bring it from the beginning to the end. I have done every process.
Now as far as electrical test we did it very differently back in the day. None of this flying probe nonsense. Which by the way I believe is a netlist generated process. We used to use “Bed of Nails” test machines that had a .1″ x .1″ grid of spring-loaded test pins. A translating fixture would map these pins to another set of pins that lined up to all of the component holes. (None of this Surface mount rubbish.) Then a thoroughly inspected board would be used to program the the matrix with everything that should show as a short. The tester would then compare every point to every other point and if something was shorted that shouldn’t have been or something that should have been shorted wasn’t the machine would stop testing and display the coordinates in question with an error code (Short or Open).
Anyway, there’s my $0.02,
Glenn