Introduction

Assignments

Final Project

Electronics Production (week 05)

Group assignment: characterize the design rules for your PCB production process. For this section reader is referred to Bomi Doh section.

Individual assignment: make an in-circuit programmer by milling the PCB (AVR ISP), program it, then optionally try other PCB processes.

Several things have to happen before this week assignment minimum requirements are to be met. The first step is of course, designing the PCB circuit before soldering components on it and then program the programmer so it will serve its purpose in future electronic programming to come.

For this part we will be using:

Eagle software to design the PCB traces. (Eagle software)

FlatCam to generate the G-Code needed on the CNC side. (FlatCam software)

Soldering station available in EchoFab

To program the programmer itself, I opt for Arduino IDE software

BOM List:

Part Value Device Package Library Sheet

C1 10pf CAP-US1206FAB C1206FAB fab 1

C2 10pf CAP-US1206FAB C1206FAB fab 1

C3 1uf CAP-US1206FAB C1206FAB fab 1

D1 3.3v ZENER_DIODESOD123 SOD123 fab 1

D2 ZENER_DIODESOD123 SOD123 fab 1

IC1 ATTINY44-SSU ATTINY44-SSU SOIC14 fab 1

R1 1k RES-US1206FAB R1206FAB fab 1

R2 100 RES-US1206FAB R1206FAB fab 1

R3 100 RES-US1206FAB R1206FAB fab 1

R4 499 RES-US1206FAB R1206FAB fab 1

R5 0ohm RES-US1206FAB R1206FAB fab 1

R6 0ohm RES-US1206FAB R1206FAB fab 1

R7 10k RES-US1206FAB R1206FAB fab 1

U$1 AVRISPSMD AVRISPSMD 2X03SMD fab 1

U$2 CRYSTAL CRYSTAL 2-SMD-5X3MM fab 1

U$3 USB_MINIB USB_MINIB USB_MINIB fab 1

Designing PCB:

Normally one needs to select all components needed, declare the links between them and then generate the routes. This time, our facilitator, Francois Auclair provided us with the full eagle project file and therefore we did not need to attend this step.

Find here views of the schematic and board

From this point forward, the necessary steps were done individually as follow (Onto FlatCam!!):

Into Eagle project space a Gerber file was generated and fetch into FlatCam. This step needs to happen to translate, with milling specification, the Gerber file into a (metric here) CNC G-Code file before we can carve the PCB.

Specifically, we used here the copper top Gerber file into FlatCam software.

Here in EchoFab, we retain by default the following settings (preference) into FlatCam to mill our boards:

Critical for us are,

· Tool diameter: 0.1

· Pass number: 7

· Pass overlap: 15%

· Milling in climb mode

· Combine passes

Next step is to set geometry before we can generate the G-code itself.

Critical for us are,

· Cut Z depth: 0.0508

· Travel Z: 10

· Feed rate rapids: 700

· Spindle speed: 24000

Next step is to generate the CNC job object, transfer it to the CNC machine and cut the board.

Note: In reality two files are generated with more or less the same settings, one is the trace file (described here before) and the second one is the profile (perimeter) of the PCB. This is essentially done in the same manner with a bigger tool and single Z pass, only deeper, to cut the piece out.

The CNC machine we are using is made so that the X and Y axes homing is mechanically assumed. On the other hand, Z homing position has to be done manually. Setting Z coordinates take advantage of the conductivity of both the milling bite and the copper surface. Essentially Zero position is done by using conductivity detection on a multimeter. One probe in contact with the copper board, the second with the bite waiting for the conductivity to show up on the multimeter while manually the Z probe system is lowered. Upon contact we know, if steps are small, that we are at the first limit of contact and Z relative positioning can be set.

Thereafter, carving trace is only a matter of executing the G-code, supervising the process in case of a fail and make a quick quality control assessment (visual inspection) before we are to solder the components onto the board.

Miscellaneous images of the milling process:

NB: Annie Ferlatte and I have traced and cut-out our board in a single sheet, this was a first attempt to be cutting two boards in a single run, saving raw material. Essentially this implicates noting coordinate for the first board to be carved, then carving the second before both are extrude of the surface while re-using previously noted home position of the first one.

Satisfied with the board cutting, it’s now time to turn to the welding of the components onto the copper traces.

Surface mount soldering is a technic to acquire, but in my case the challenge was very much it, a true challenge! I have lost my reading glasses and cannot properly see were I’m to solder. I know how but I simply can’t see well. So, after several attempt with the aid of a USB microscope I had to come to turn with my handicap. I managed to solder the parts on the paths, but results were far from being satisfactory. Some welding I’ve done, aiming at the pad before being blinded by the soldering rod, hoping, praying even that I would get it bull eye’s, led some splattering all over the place. Eventually I opt for liquid welding and got better results but not on this journey (see upcoming weeks).

Miscellaneous images of the soldering process:

 But for now, the board is soldered and now we need to program it so it will become an AVR ISP.

The sequence of command is an easy one:

· Lorem ipsum dolor sit amet, consectetuer adipiscing elit. Maecenas porttitor congue massa. Fusce posuere, magna sed pulvinar ultricies, purus lectus malesuada libero, sit amet commodo magna eros quis urna.

· Nunc viverra imperdiet enim. Fusce est. Vivamus a tellus.

· Pellentesque habitant morbi tristique senectus et netus et malesuada fames ac turpis egestas. Proin pharetra nonummy pede. Mauris et orci.

· Aenean nec lorem. In porttitor. Donec laoreet nonummy augue.

· Suspendisse dui purus, scelerisque at, vulputate vitae, pretium mattis, nunc. Mauris eget neque at sem venenatis eleifend. Ut nonummy.

A note: Here are presented the overall process we used answering this week assignment. But in reality, for a fairly long time all team members here at EchoFab had a problem with the programmer we were trying to make. Facilitator from our lab and others joined their neurons trying to understand what was going on. After troubleshooting all possibilities, it was realized that the Zinner diodes we were using were rated to high for our needs. Replacing the faulty parts with the right ones, solved the problem and we were able to move forward.

Onto week 6