Electronics - Final Academy

Week04. Making Boards. CNC and Welding

Our I-Modela could burn this week!!!

The first process of the week was try send the *.rml from FabModules to the Roland machine, but we had problems with the ports or directly with the code for the machine.
The board choosed in the first try was the ISP that use a Cristal 20MHz. We made some correction or modifications in the size to do more easy the welding process. This week was our first welding time.

Jose Pérez de Lama, aka Osfa, made the best summary of the week of FabLabSevilla. In our team, anybody have experiences with electronic or welding, so we worked together all the week.

Down, the size correction. We name it FLOW ISP

FabModules_vs_I-Modela... I have no many idea about FabModules sender code, Osfa show me something but I dont understaned any code line... ( I´ll study it, I promisse). We can´t send the files directly to the machines, and the *.rml code is writed for other machine, we was probles with the sizes and the movement into the I-modela. While part of the team was trying to connect I_Modela and Fab Modules,  the other worked in manufacturing with vector files.

Quickly, we started working with Inkscape to vectorize the *. png. OSFA had already made changes to the board that incorporates the "crystal", so, with we had two vectorial files to work with the I-Modela (need *.ai) and our old CNC machine Alaris30 (need *.dxf). I dont know why, but all the board have diferents size, almost negligible, may be consequense of the vectorization process. The board use to work was a FR4, we have had problems in the shipping this time, and we had not the FR1 board.

The I-modela got us the best results, we use a 45º bit to work, and making cut with 0.01 mm depth (I know, the depth must be 0.2 or more...but the I-modela controler show us 0.01 mm depth when are working and the copper is deleted totaly)

Pic of vector files used.

Milling Process

Okay, I-Models is not the best machine for making the PCBs, our little toy trembles and resonates with each pass... but the results were amazing!!
To get the cut piece, we used a Proxxon disc-cutter.

Tricks and traps....

The first consequences of use I-Modela is the time consumed in each milling process. The velocity used was 40 or 50 mm/min in X/Y movements and 20 mm/min in Z movement.

To send a work to the machine we used the Roland software, I-Modela Creator. Here, we define cut areas, engraves areas or pocket areas. So, if we want delete all the excess copper we must use the pocked tool, but the time required is so much, accordingly we used the cut tool to make an single cut with 0,01 mm depth, configurating a 0.3 mm diameter to calculate the trace, completed the first process, another trace is sent, but in this case, with a 0,45 mm diameter, and not changing the 45º tool, after, another with a diameter a little bit more big, and other..and other may be...until we get the necessary separation.
Using FabModules, we can choose the number of offset that we want to cut or engrave, but not  in I-modela Creator.

We hope that the FR1 board will be more easy to be mill. For this week we only got three finished boards with some mistakes in the traces. When we was completing the welding process, some copper traces were removed from fiber board... and we had lost the posibility to finish the board totaly. Counting the components, we discovered that we don´t had the diodes required, already been purchased and are sending it.

The pic show the last result of the week

Well, now my summary:

In the up text you can read some happy solutions. But this week was very very long!!!

1. Conect  I-Modela and Fab Modules. Not posible.
No port. No the same language.

2. Try to made the board with the Alaris30 machine. Not posible.
0.016 inch tool diameter, with long flutes, all broken.
The work surface was not horizotal totaly in some tries.

3. Without FR1. We worked with FR4, very hard, and dificult to be cut.

4. I-Modela velocity. Each PCB consumed more than two hours to be completed.

5. First time welding. This not was a problem, I´m happy with the results.

6. The only difficult part to be soldered was the USB conector. I saw the Valentin
board, here the mini-USB conector was removed and replaced with a USB conector directly on the copper board.

My last work this week was draw a new PCB that mix the Valentin´s  idea and the board that we used. I hope it run!! Now we await the arrival of the diodes to continue working.

Week06. Electronics Designs

Following Anna's tutorial Kaziunas.

I downloaded the files of the board "Hello_Echo". I opened the diagram and follow the steps to add the missing components, resistors, LED and button.

After several attempts and corrections self, all components are connected correctly. It's time to route the board.
Now ... the autorouter is a horror ... I tried to make a placement own for components, but most of the connections are crossed. truly, I ran out of week. no time for milling. So, I reset all the components in place, as in the tutorial, and mainly, I modified the "Design Rules", to then, draw the conection by hand with the "route" tool.

In terms of change of the "Design Rules" It took me a while to discover the system of units used for each modification, it is unclear when used when used milimeters or inches. I did not know the unit "mil", 0.001 inch. Knowing the units used, the photograph shows the chosen values. In general, 16 mil for each change, 0.016 inch or 0.4069 mm.

To make the board, export it in *. Png (monochrome).

VCarve Pro I used to create the file sent to the router, our Alaris30.
In this software, we can create the vectors from images. I used a milling cutter with a diameter of 0.016 inches.

Week08.Embedded programing/ Program Board

This week, no doubt, has been the most difficult to tackle. A field too abstract and unknown to me. Without venturing far, I be sure to get go further over the coming weeks, but for now, I just follow the steps taken by Jose Perez de Lama to perform part of the homework.
Jose Perez de Lama has been helping us during the week to understand concepts that we could not decipher. Meanwhile, Alejandro Campos, Masters in Industrial Engineering in the field of electronics, we explained the most complex part about programming languages ​​and the inner workings of the microprocessor, in this aspect, Juan Carlos Venegas explains on its Web everything relevant.

Course of the week:

Reading DataSheet. I remained stuck in a few days reading the summary. I could not explain what I read, but the notes made ​​by Alejandro and Jose helped me to understand a few lines of the program that we choose.

Down, the three points that I understand:

In ATTINY44 there are three types of memory.
FLASH.4k. Memory hard. Here are stored the necessary programs.
EEPROM.256 b. Memory hard. Here are stored the data needed for the program.
SRAM. 256 b. Volatile memory. Here are stored the data produced or received. Clears the reset.

Clock.

The internal clock runs at a speed ATINY44, we placed a crystal to greatly improve the speed.



Programming. Mournfully, I have not learned to program in C or Python or anything. Although, I have learned something about the different programming languages​​, or rather, about language levels.

High level >> Python >> needs to be interpreted

Middle level >> C >> needs to be compiled

Assembler >> Low >> machine code (for us *. Hex)

Objective. Find a circuit similar to ours. Microprocessor> LED> button
               Find a program (C) to do something with them.
               Understand how it works.

Jose discovered the work of Travis Rich, HTMAA student.

We used their promagra, written in C, to turn on the LED when you press the button, counting certain delay.

We used to compile hello.ftdi.44.echo.c, written by Neil G. an we followed the steps from one of the "class files"

AVRISP mkII was used. First installed AVR dude.

The week has been short. But we were able to load the program to one of the boards that we prepare for Week 06. I hope to learn a lot more about programming.

Happy Embedded !!!

Week10. Inputs

hello.lignt.45                     hello.temp.45                      hello.txrx.45

The production is amazing, in a matter of an hour and a half we had three plates welded and in another half hour, all programs loaded, OLE!

It was the first week dedicated to electronic everything we wanted to do worked, remains to me to learn to do the programs in both C and Python, but I'm on my way.

Board to Board, we use the file. C and "makefile" specific to each board, using our AVRISP2 as a programmer. Hello.temp.45, the first board works perfectly. In hello.txrx.45 we detected welding failure in one leg of the micro, once resolved, compile and program, and after reading in python, perfect. The third board seemed to work, green LED on!! but we could not program the micro, we detected some faults in welds, and again, once resolved, everything flowed.

I repeat, this week I would like to gut the C code to learn how to make some changes, it would be perfect!

Week13. Outputs

Week14. Networking

For this week, the winer word is DEBUG!! and second RECYCLE!!!

I made the hello.bus.45 set. with three nodes.

Only one of the boards was programed without troubles, the others was debuged many times before  to work.

In this electronic week, my idea was instal all the software needed to work in my laptop, running with Windows 7. Until now, I had not installed Python for example, and now work perfectly.

I followed the tutorials from Anna Kaziunas page, thank you Anna.

Like last week, I made the boards without using milling machine.

The chosen system was new, direct transfer from paper. I used sheets Press-n-Peel Blue Transfer Film.

Not all the team worked with the transfer. Here you can see some Hello.nodes made by Jose Perez with our I-Model, and the Hello.Bridge made by Juan Carlos  Juidias with the same machine.

I had prepared three boards, two Nodes and a Bridge, but my Bridge died in a hard debug session.

In this point of the Academy classes our electronic´s part stock is alarming, so we exploit all components of defective parts.

Following the Anna Tutorial for this week, the node was named like '0','1','2'. Here the Viemo link.

My favorite work for this week... the USB-FTDI-NODE wire!!

I was no sure if conecting the Bridge to the FTDI wire, and the Node with AVR , I could write something in the Bridge´s Micro, so I prepared a special conection with prototype wire from Arduino box.