Vejen, Denmark
Offline
Edison Member
Karma: 7
Posts: 1177
RAWR! ^,..,^
|
 |
« on: July 07, 2011, 12:57:05 pm » |
I want to learn how to use eagle. The reason I want to learn how to use it, is because I am planning on ordering some pcb's with a ATmega328 (with crystal and all the other parts needed to make it run), a 433 Mhz receiver, and then some relays/2N2222's/opto's (depending on what application they are going into). My problem is now that I find eagly VERY confusing, and is not sure where to start to be able to "see the light".
|
|
|
|
« Last Edit: July 09, 2011, 03:53:55 pm by bld »
|
Logged
|
|
|
|
|
Leighton Buzzard, UK
Offline
Edison Member
Karma: 11
Posts: 1049
|
 |
« Reply #1 on: July 07, 2011, 03:17:59 pm » |
I have been down that path here's what I did
1. layout the schematic 2. find out how to add libraries so you can choose components
3. when you get to laying out the pcb - don't bother with auto-route at best it's awful! I always route by hand
4. the advanced course is creating your own components so many reasons why it's a good idea and not hard
so baby steps and ask if (ok when) you get stuck!
|
|
|
|
|
Logged
|
there are only 10 types of people them that understands binary and them that doesn't
|
|
|
|
Vejen, Denmark
Offline
Edison Member
Karma: 7
Posts: 1177
RAWR! ^,..,^
|
 |
« Reply #2 on: July 08, 2011, 08:05:21 am » |
Hope I am doing it right... So far I should have the 328 chip, crystal, capacitors, and a 7805.... If I actually did what I thought I did 
|
|
|
|
|
Logged
|
|
|
|
|
Leighton Buzzard, UK
Offline
Edison Member
Karma: 11
Posts: 1049
|
 |
« Reply #3 on: July 08, 2011, 09:09:13 am » |
well it IS a schematic you might want to check the wires though far too many joined together!
also the wire to RESET (pin9) seems to go tright through the resistor you can easily check if it's connected properly drag the resistor and see if the wires go with it CTL-Z will "undo"
|
|
|
|
|
Logged
|
there are only 10 types of people them that understands binary and them that doesn't
|
|
|
|
Global Moderator
Boston area, metrowest
Offline
Brattain Member
Karma: 242
Posts: 16490
Available for Design & Build services
|
 |
« Reply #4 on: July 08, 2011, 09:52:02 am » |
As a beginner, one should also use File:Export:Netlist when done, it will output a list of what is connected to each signal. I find it makes the netlist much easier to read if each net has a name - use the "Label" button (looks like ABC over a line) to give each signal a name - double click to drop the generic N$1 name on a signal line, then right click and select Name to give it a meaningful name. For example, you can see in this schematic that every signal has a name - they will show up as blue lines to be routed when you layout the board; they do not need to actually have lines connecting them on the schematic. Be sure to use +5V and GND for those names, that will ensure they get routed with thicker lines and not just standard thinner lines. I like to start a design by trying the autorouter - it will let you see where you component placement was poor - lots of vias, lots of signals going all over te place. Rip up the signals, move things around, try again. You can eventually get down to minimal vias - this design has just 1 where the regulator output capacitor had to connect to the bottom of the board. If that had been a thru-hole component, even that might have been avoided. 
|
|
|
|
|
Logged
|
|
|
|
|
Vejen, Denmark
Offline
Edison Member
Karma: 7
Posts: 1177
RAWR! ^,..,^
|
 |
« Reply #5 on: July 08, 2011, 09:58:01 am » |
Starting to look like something! I hope...
Got all the basic components on, and still need to add a bunch of 3N3333's and a row of pin headers for connecting to them.
My crystal is a lot bigger... hmmm, wonder if I took a wrong one in the list.
|
|
|
|
|
Logged
|
|
|
|
|
Left Coast, CA (USA)
Offline
Brattain Member
Karma: 279
Posts: 15314
Measurement changes behavior
|
 |
« Reply #6 on: July 08, 2011, 10:20:52 am » |
One word of caution from my past experience. I once bought a picaxe based bare board of similar design and used right angle male header pins so the board would mount vertically when plugged into a solderless breadboard. The force needed to plug the pins into and removal from the breadboard soon broke many of the header solder joints and lifted the trace pads. I think plated through holes would have given more mechanical strength to the joints, or possible smaller pins instead of the .025 square header pins would be better for breadboard insertion and removal ?
Lefty
|
|
|
|
|
Logged
|
|
|
|
|
nr Bundaberg, Australia
Online
Tesla Member
Karma: 71
Posts: 6816
Scattered showers my arse -- Noah, 2348BC.
|
 |
« Reply #7 on: July 08, 2011, 10:47:00 am » |
Your getting there, a few comments
C1 and C2 should go to GND.
C3 and C4 are normally 1/10/22uf tant or electro types and therefore larger than you have shown.
There are no decoupling caps, put some as close as possible to the chip's GND and VCC pins.
Crystals normally have an X designator. Q is for transistors.
What is a 3n3333?
______ Rob
|
|
|
|
|
Logged
|
|
|
|
|
Vejen, Denmark
Offline
Edison Member
Karma: 7
Posts: 1177
RAWR! ^,..,^
|
 |
« Reply #8 on: July 08, 2011, 11:52:43 am » |
C1 and C2 should go to GND.
Fixed  C3 and C4 are normally 1/10/22uf tant or electro types and therefore larger than you have shown.
I think I got the right ones now... There are no decoupling caps, put some as close as possible to the chip's GND and VCC pins.
What would they be? Same as for the 7805? What is a 3n3333?
Brain-fart, ment 2N2222 And now... The headacke... Looks like a MESS now... heh, but most of the components should now be on. Still problems with the trace width on power though...
|
|
|
|
|
Logged
|
|
|
|
|
Leighton Buzzard, UK
Offline
Edison Member
Karma: 11
Posts: 1049
|
 |
« Reply #9 on: July 08, 2011, 12:05:47 pm » |
you need to check C1 and C2 one end of each should connect to the crystal (and chip) the other end should go to ground
one of them seems to wander off to the 2N2222 group also I would try and get the crystal as close to its pins as possible, you seem to have room to nudge it down a little
|
|
|
|
|
Logged
|
there are only 10 types of people them that understands binary and them that doesn't
|
|
|
|
Vejen, Denmark
Offline
Edison Member
Karma: 7
Posts: 1177
RAWR! ^,..,^
|
 |
« Reply #10 on: July 08, 2011, 12:21:07 pm » |
you need to check C1 and C2 one end of each should connect to the crystal (and chip) the other end should go to ground
one of them seems to wander off to the 2N2222 group also I would try and get the crystal as close to its pins as possible, you seem to have room to nudge it down a little
They should both be connected almost directly to GND (top left corner). Tried to move the crystal a bit closer. I also tried the name thing, but still only getting thin signal traces for the power too. Netlist
Exported from 8x 3N3333 board.sch at 08-07-2011 19:19:12
EAGLE Version 5.11.0 Copyright (c) 1988-2010 CadSoft
Net Part Pad Pin Sheet
+5V C4 1 1 1 D1 C C 1 IC2 IN IN 1
GND C1 1 1 1 C2 1 1 1 C3 1 1 1 C4 2 2 1 IC1 11 GND 1 IC1 31 AGND 1 IC2 GND GND 1 JP1 1 1 1 JP2 10 10 1 JP2 12 12 1 JP2 14 14 1 JP2 16 16 1 JP2 2 2 1 JP2 4 4 1 JP2 6 6 1 JP2 8 8 1 T1 E E 1 T2 E E 1 T3 E E 1 T4 E E 1 T5 E E 1 T6 E E 1 T7 E E 1 T8 E E 1
N$1 D1 A A 1 JP1 2 2 1
N$2 R2 2 2 1 T1 B B 1
N$3 IC1 9 RESET 1 R1 2 2 1
N$4 C1 2 2 1 IC1 12 XTAL2 1 Q1 2 2 1
N$5 C2 2 2 1 IC1 13 XTAL1 1 Q1 1 1 1
N$6 C3 2 2 1 IC1 10 VCC 1 IC1 30 AVCC 1 IC1 32 AREF 1 IC2 OUT OUT 1 R1 1 1 1
N$7 R3 2 2 1 T2 B B 1
N$8 R4 2 2 1 T3 B B 1
N$9 R5 2 2 1 T4 B B 1
N$10 R6 2 2 1 T5 B B 1
N$11 R7 2 2 1 T6 B B 1
N$12 R8 2 2 1 T7 B B 1
N$13 R9 2 2 1 T8 B B 1
N$14 IC1 16 (INT0)PD2 1 R9 1 1 1
N$15 IC1 17 (INT1)PD3 1 R8 1 1 1
N$16 IC1 18 (OC1B)PD4 1 R7 1 1 1
N$17 IC1 19 (OC1A)PD5 1 R6 1 1 1
N$18 IC1 20 (ICP)PD6 1 R5 1 1 1
N$19 IC1 21 (OC2)PD7 1 R4 1 1 1
N$20 IC1 1 (T0/XCK)PB0 1 R3 1 1 1
N$21 IC1 2 (T1)PB1 1 R2 1 1 1
N$23 JP2 1 1 1 T1 C C 1
N$24 JP2 3 3 1 T2 C C 1
N$25 JP2 5 5 1 T3 C C 1
N$26 JP2 7 7 1 T4 C C 1
N$34 JP2 9 9 1 T5 C C 1
N$35 JP2 11 11 1 T6 C C 1
N$36 JP2 13 13 1 T7 C C 1
N$37 JP2 15 15 1 T8 C C 1
Isn't this how it should be? Or did I misunderstand something?
|
|
|
|
|
Logged
|
|
|
|
|
Vejen, Denmark
Offline
Edison Member
Karma: 7
Posts: 1177
RAWR! ^,..,^
|
 |
« Reply #11 on: July 08, 2011, 01:50:00 pm » |
And if anyone is interested, this is where it is going to be used in My goal is to make multiple of these devices, and then have different flavors so they can be used in different things. Like one with the 2N2222's to trigger the buttons in this fan, another with 230V relays to handle lights, and one with opto's to do pretty much the same as the first, but with it isolated more. This reminds me... I forgot to add pinheader for the RF receiver.
|
|
|
|
|
Logged
|
|
|
|
|
Global Moderator
Boston area, metrowest
Offline
Brattain Member
Karma: 242
Posts: 16490
Available for Design & Build services
|
 |
« Reply #12 on: July 08, 2011, 01:58:11 pm » |
Did you intend to use a big 40 pin device? Or were you shooting for ATMege328 based board like the Uno uses? H49US is the crystal I think you were after, that's the more normal size.
|
|
|
|
|
Logged
|
|
|
|
|
Vejen, Denmark
Offline
Edison Member
Karma: 7
Posts: 1177
RAWR! ^,..,^
|
 |
« Reply #13 on: July 08, 2011, 02:05:00 pm » |
Did you intend to use a big 40 pin device? Or were you shooting for ATMege328 based board like the Uno uses? H49US is the crystal I think you were after, that's the more normal size.
heh, yeah... I wanted to use the same as the uno and all those boards use... Guess that is named MEGA8-P... Didn't really look that much at it... just saw it had pins and the shape...
|
|
|
|
|
Logged
|
|
|
|
|
Global Moderator
Boston area, metrowest
Offline
Brattain Member
Karma: 242
Posts: 16490
Available for Design & Build services
|
 |
« Reply #14 on: July 08, 2011, 02:12:15 pm » |
If you were after more IO, the bigger part is one way to get there without shift registers/etc. and you get a 2nd hardware serial port also. ATmega644P (or 324, or 1284, different memory sizes).
|
|
|
|
|
Logged
|
|
|
|
|
|