26 September 2012

EE Prototyping 2

Today was the Course Fair, where the professors hold a "trade show", of sorts, and advertise their classes for the Spring term.  I brought along some show-and-tell items to promote the EE Prototyping class.  First off, the following description for the course catalog (written with help from the seminar) was submitted:
Through a series of projects, we will learn to design, build, and debug electronic prototype systems. We will cover multiple aspects of the prototyping process, including circuit and system design, soldering, deadbugging, troubleshooting, component selection, schematic capture, printed-circuit board (PCB) layout, PCB fabrication, PCB assembly, and thermal analysis. We will discuss the tradeoffs among "faster, better, cheaper", and explore examples in the realms of analog, digital, RF, and power. In addition to hands-on reverse engineering and fabrication experience, students will learn technical communication through design documentation.
Secondly, I brought a number of boards to demonstrate the right ways and wrong ways to build circuits (stealing some ideas from Appendix F in App Note 47). First the wrong ways: solderless breadboards and wire wrap (just say no!):

And now the right ways: dead bug on copper clad, as preferred by Jim Williams (described in detail in App Note 47; the board shown is actually a simple oscillator that Jim built for me); and, of course, the good life, a custom PCB (like this Analog Devices eval board):

I also brought along some other examples of interesting PCB boards. Exhibit A: The low-cost circuit board from a floppy disk drive (a single-sided PCB, that includes a "square-wave" trace around the motor for the position encoder).

Exhibit B: The controller board from an inkjet printer, which includes a wide variety of IC packages such as the socketed DIP, several SOICs, and that EPSON ASIC in the middle with a million pins on tiny spacing.

Exhibit C: A ruggedized power supply (you can tell it's rugged from all the epoxy holding the parts in place).

And finally, Exhibit D: The PCB inside a Spectral Synthesis ADDA2218, an 18-bit analog-to-digital audio converter, which uses colored FR4 (the color isn't painted on, it's impregnated in the fiberglass). I thought this was a neat touch, for a circuit board that only one in a thousand customers would ever see (only those willing to void their warranty!).

The reading assignment this week for the seminar was Sections 9.1 and 9.2 (Hardware Design Techniques: Passive Components and PC Board Design Issues) of the Analog Devices Data Conversion Handbook.

Next week, I'll talk about some of the proposed class projects.


Johan said...

Wouldnt the top traces on excibit D have another colour if the sunset was in the FR4? My bet would be a coloured solder mask, but I guess it would be easy to see if one looked at the edge of the board...


Kent Lundberg said...

Good observation. It's both! If you look at the edge of the board, you can see that the fiberglass layers are colored, too.

Unfortunately, my camera doesn't have a good macro lens, so I can't get a good picture of it. I'll try find a better camera and post a photo.

Kixz said...

The third picture is Jim #327 lamp Wien bridge oscillator?!

Kent Lundberg said...

Yes, Kixz, it is Jim's light-bulb Wien-bridge oscillator.

The "Oliver"/"No Oliver" switch adds distortion to the op amp, which affects the settling time of the amplitude. The name is in reference to a paper written by Barney Oliver while he was director of Hewlett-Packard R&D, called "The effect of \mu-circuit nonlinearity on the amplitude stability of RC oscillators".

Felipe Maimon said...

All these posts are rather interesting but I really miss the ones commenting app notes.

Kixz said...

I want to use you photographs "Jim's light-bulb Wien-bridge oscillator" in my documents with Wien-bridge.

I hope to get permission.

I am very grateful. Kixz - lkixz@yahoo.com