Books for pcb design

[pucca]

Hi! I'm new in this field and I want to find some books that would help be to make a pcb design. I have to make a module with sensors and other components. What's the first book I should read? I'm an electrical engineer but I have n't ever made such a design.

Thanks!!

 

[MacGyverS2000]

pucca, is this tied to your antenna design thread? If you intend on making PCB antennas and have never done it before, I suggest hiring outside help (unless you have years to kill or lots of money to burn on hundreds of prototypes).

If it's a simple board for attaching sensors to, I'm not really sure if there are any books to recommend as the work is pretty straightforward. Every design program has its own quirks, and those are best learned by using the program and asking other users (typically on package-specific forums run by the manufacturer).

If you have a more specific question, I'm sure we can answer it here...


Dan - Owner

http://www.Hi-TecDesigns.com

 

[pucca]

I have made printed antennas for RFID appl , using ADS of Agilent. In the fist phase of the project I will not make a printed antenna but I will use an external one. But my problem is the PCB design of the rest of the board,as I have never done smthing like that before.

 

[pucca]

It's an apllication for wireless sensors network

 

[MacGyverS2000]

If you've made PCB-based antennas, you know pretty much all you need to know to make a simplistic sensor board.

Dan - Owner

http://www.Hi-TecDesigns.com

 

[mcgyvr]

I'd recommend going through the IPC specs.
Like IPC-2221A generic standard for printed circuit design.
Or any others that apply to your application/board type (thru hole/SMT,etc..)

 

[pucca]

You mean that I should avoid auto routing and do it manually?
The auto routing will create mess or help me?

 

[itsmoked]

Autorouting will cause some of the most screwed up boards you've ever seen. You should hand route several before carefully using autorouting.



Keith Cress
kcress -

http://www.flaminsystems.com

 

[MacGyverS2000]

You would need to spend weeks/months to properly set up the auto-router rules for a small board, and a large, multi-layer board will take you many months if you've never done it before.

If this is a simple sensor board, it shouldn't take you longer than a day or two to lay it out by hand...

Dan - Owner

http://www.Hi-TecDesigns.com

 

[BobM3]

"Printed Circuits Design" by Ginsberg. It covers items that (I'd guess) aren't considered with most layout software like designing for automatic assembly, board materials, soldering considerations, EMI.

 

[GonzaloEE]

I started doing PCBs in Altium Protel99SE some years ago. As I haven't used any oher s/w, I don't feel like suggesting you one. Drawing PCBs it's fairly easy if you have some electronic background.

Just adding up to the above rules:
- Contact the board manufacturer, to make sure you'll meet their manuf. specs (hole sizes, track width, etc.). Define and draw items with fixed positions first(connectors, screw holes, etc.) Start with subsystems first(supply, core processor, RF stage, etc.), arranging their parts and drawing tracks, then place these subcircuits in the main board area, then drawing the remaining tracks. Defined subcircuits help in debugging the prototype.

If you're working on multilayer PCBs, keep tracks on opposite layers perpendicular from each other, to minimize crosstalk. Avoid drawing long, straight tracks if you're expecting current pulses there, as this may act as antenna or RF stub. Draw thick fan-out tracks from VCC supply to subcircuits, and use a big GND plane for current return. This reduces high current loops and weird prototype behavior.
Try to use different GND planes for power supply, logic and analog/RF stages, to avoid noise from spreading through. Join them together with thick GND tracks far from noise-sensitive parts.
Minimize via usage, as they add inductance to RF tracks, while reducing effective conductive section if high currents are expected to get through. Vias are also common source of problems during board manufacturing, reflow soldering, etc. Add testpoints over tracks and vias whenever possible, marking the signal name, to help during debugging process.

Routing RF boards is a tricky business, more suited for artists. My little help would be: avoid sharp corners for RF tracks, as corners may act like radiating points at high freqs. Keep power tracks for RF stages as thick as possible, while pouring a wide GND plane around/below circuitry. There are loads of Internet resources to learn more about track impedance calculations. This would be valuable on creating printed antennas and sensitive tunable parts like filter coils, microwave capacitors, etc.

Good Luck!