The assignment for this week is reproducing a programmable PCB and adding input devices and output devices.
To give an overview from the concept of the different devices and to give an explanation about the input
and output devices I made the video below.
For the input devices I choose for 3 switches and for the outputdevices I choose for an RGB-led.
For this moment I think the program wil work like this:
- Input: push botton 1, output: blink red (1 sec), blink green (1 sec), blink blue (10 sec)
- Input: push botton 2, output: blink red (1 sec), blink green (10 sec), blink blue (1 sec)
- Input: push botton 3, output: blink red (10 sec), blink green (1 sec), blink blue (1 sec)
The video above presents three steps and these steps are given in the pictures below. (Left) Connecting the new pcb (Hello World pcb) to the ISP-tiny which is connected to a usb-port of the computer.
Now the Code can be send to the Hello World pcb. Once the code is installed the Hello World pcb can be disconnected from the ISP-tiny (middle). With an other cable the board can be connected
to the computer to control what is happening on the board (right).
In the picture below the three push buttons are drawn with a connection to the ground (GND). In the middle the Chip is drawn (IC) with 14 pins.
On the right the RGB-led is drawn also with connections to GND on the cathode side. On the top of the drawing there are two connectors.
On the top left the connection for programming and on the top right the connection for reading.
On the right picture the wires are drawn between the components and some additional components are drawn like the capacitor and the resistor.
One component is missing for this moment, the resonator, which is connected in a loop to the IC and it is connected to GND. Because the traces from the two connectors cannot be connected
to individual pins on the IC they should share the connection if possible (there are only 14 pins). This is possible because programing is separate from
reading / using the board. In de right drawing there are mistakes in the connections of the wires, but the idea is that wires are shared where possible.
After doing the tutorials about how to use Eagle from Jeremy Blum (below) I started drawing the schematic from the board. I choose to use the connect to ground GND and
connect to power VCC in stead of linking all wires together in the schematic. From the left to the right the components are: A) the connector to the ISP-tiny, B) RGB-led with 3 resistors,
C) Chip with 1 resistor, D) Connector to the PC for reading, E) Resonator, F) Switches and G) Capacitor.
The initial idea was to make a single side copper board. This would me easy for milling and I thought the board would not be difficult to make.
After some ittarations and checks I managed to get it all connected. To attach all the components together many cross overs had to be made with 0 ohm risistors.
In total I need about 18 of the 0 ohm resistors. And in my opinion this is too much. Maybee it can be optimized in the layout, but I was happy I was able to make it thhis far.
In the picture below the single sided schematic is shown.
To check the layout on mistakes I used Sketchbook Express on my Ipad. Below I will show how I debuged the traces. The mistakes were (right column):
- thin yelow lines, because traces were not connected properly
- one 0 ohm resistor too much
- Trace uderneath chip is connected to a wrong pin from the chip
- Two traces were connected that should not be connected
After some changes I checked all lines again.
Electrical Rule Check (ERC)
In this case I did the Electrical rule Check after making the board. Ofcourse this should be the other way around....
The mistakes in the schematic were not a big problem for making the board. Green lines were not stopping at the right place.
The line MOSI did not have a label displayed in the schematic. Also not a problem.
Here might be a problem. I don't know what the warning "Only one pin on net RTS and CTS" means....
It asks for a value for the switch. Maybe it has no vanue?
Design Rule Check (DRC)
The Design Rule Check was also executed after making the board.... Hopefully the errors wil not be a big problem.
Most of them are only warnings for dubble lines. I had to draw them double to let the air wires (yellow lines) disappear.
For milling the board on the LPKF-machine we had to draw two holes in the board. DRC gives a warning on these holes.
I now see a mistake in the 3x2 connector. There is an overlap in the traces.
Double side board
I wanted to improve the board design by using two layers and by using a ground layer on two sides.
It turned out much better and smaller and no 0 ohm resistors had to be used. Below there is a picture
from the two layer board.
In the schematic you see no extra 0 ohm resistors.
Here yoy see all layers on top ao eachother. The geen dots are drilled holes (via).
This is the top layer.
This is the bottom layer.
Milling
To operate the LPKF milling machines Gerber files are needed. From Eagle these gerver files can be generated. The third video tutorial
from Jeremy Blum explains how to do this. Basically every layer for the board ( this can be many ) will be experted to a new file
withe a specific extention name.
The single side board was
Selecting components (BOM)
With the Eagle Software a Bill of materials can be generated semi automatically. I did not do it because .... no time.
Soldering
In the pictures below the soldering of different components is shown.
The final result of the single layer soldered PCB.
Mistake in the layout close to connector
In the board layout design process the connecors wer moved several times. The connecting traces should also move accordingly.
By mistake this was not done properly. See picture below.
The problem can be solved by desoldering the connector. Cutting the trace between the middle pins and solder a wire
from the pin to the right trace, see pictures below.
Problem with delamination of copper
I started soldering without a mangifier. On the two layer board the GND layer of copper is very close to all ather traces.
I soldered the IC wrong and with a hot air gun I wanted to take it of. I applied too much force and damaged the top layer.
Micro Capture Pro
It was cleat to me I had to use a magnifier to make soldering possible without damaging the board. In the Fablab Enschede we have a Micro Capture Pro.
It is a kind of webcam for zooming in. For me this was a very good tool to use.
In the video below you see the soldering process. Don't mind the component is soldered on the wrong place...
After measuring the soldering from the chip there were two pins not soldered properly. The resoldering was not done in a good
way. There was too much heat on the copper traces and they came loose, which is very bad. With two wire I tries to bridge the gap.
This soldering was not easy. At the end I decided to cut off all but one othe the wirebush and solder this single wire to the IC-pin.
Finally I succeeded and tested the connections with the multimeter. Tomorrow we will check if it works properly.
As you can see in the picture the board looks very bad and with testing the board there were problems with the IC. Therefor I had to replace the IC.
In this case this was quite impossible with the wires and the copper traces damaged. So I decided to make the whole board from scratch again.
The result is shown in the picture below.
Files to work with:
Please go ahead and use my files to modify it to something for your need.
