Common practices of a good analog and mixed Layout Designer.

[ninge]

Hi All!!!

I am new to Analog and Mixed Signal Layout Design...please any one guide me, what are the main considerations and practices to become a good Analog and Mixed Signal layout designer?

Thanks in Advance
Ninge

 

[renwl]

read the book <<the art of the layout>> can give you some ideas.

 

[protonixs]

Re: Common practices of a good analog and mixed Layout Desig

read the book <<the art of the layout>> can give you some ideas.

you may also read "IC CMOS LAYOUT" by Dan Clein

 

[yibinhsieh]

Re: Common practices of a good analog and mixed Layout Desig

1. be familiar with layout tools.
2. be familiar with DRC and LVS checking and debugging.
3. be familiar with different layout techniques such as common-centroidal, interdigital, etc.

4. be familiar with layout floor plan for different analog circuits.

 

[hionyeh]

I suggest "The Art of ANALOG LAYOUT" by Alan Hastings.

 

[dkace]

Re: Common practices of a good analog and mixed Layout Desig

Common guys and girls!

Do you really believe that Hasting's book is the best point to start? Why?
The Art of Analog Layout, is an Encyclopedia, is not a beginners book!

Judging from my start up and from the level I am currently, I believe that "IC Mask Design" of Saint & Saint is far the best book for a zero level layout engineer. It is like having a supervisor next to you on each page. If there are things you don;t understand, then you have Hasting's book on the other side.

But, no matter how many books you read, if you don't implement them right after your study, it is like you read a story.

So a Layout tool with a technology mounted is a must to try your first layout.

And last but not least, you need a designer to guide you through the first steps.

Don't try anything else. It is just a waste of time!

D.

 

[chenjia]

Pratice is very important in becoming a good layout engineer, so using your hand and brain to pratice

 

[xaxtel]

few words that u always need to use : EM requirement, high current branch, matching, parasitic capacitance

 

[dick_freebird]

For me, it's about internalizing the ways-things-go-wrong
and then not repeating them, for the most part.

That's not particularly helpful in getting started, more of
"where you want to end up".

Along the way you want to absorb the "hot button" things like
good (and perhaps more importantly, bad) matching practices;
gain a mental picture of parasitics and how they might
(mis)behave in terms of noise, leakage; develop an inherent
mistrust of provided models and challenge / supplement them
with your own data / fitting / "crutches".

Developing "gut" practices that avoid you trouble without
having to think about every stinkin' detail, every stinkin'
time, is probably your goal. That's learn-by-doing, make your
mistakes in the sandbox rather than on the clock, territory.

You're only as good as your last boat of FOK wafers. Every other
measure of a designer is pretty much vapor. When the going
gets tough, a lot of folks head for Marketing. Bench skills are
often neglected but being able to claw your way out of a pit of
"WTF???" is not only essential to career survival, but in the end
probably your best teacher.

 

[sree_lakshmi]

Re: Common practices of a good analog and mixed Layout Desig

you can get some info from this pdf

 

[love_analog]

Re: Common practices of a good analog and mixed Layout Desig

Other things I would add: shielding, guard rings (substrate shielding), power supply grids.

Other posters have already mentioned other important things - matching, parasitics, EM etc.

 

[sree_lakshmi]

Re: Common practices of a good analog and mixed Layout Desig

Other things I would add: shielding, guard rings (substrate shielding), power supply grids.

shielding is a process where you put the metal line to be shielded in the middle and both sides you draw a path with same metal and cover the three lines with top and bottom metals with all the four drawn layers connected to VSS or GND.

 

[fcfusion]

Re: Common practices of a good analog and mixed Layout Desig

Cmon people, you're telling ninge to read hardcore books? I guess he only wants advice on proper working method. So here it goes.

Advises:

1) Study the circuit's theoretical behaviour and make sure it can really work. If it doesn't work in theory it won't work in practice.

2) Don't rely on parameter super-optimization, such has transistor sizes, resistor values, etc. Instead rely on robust circuits, especially the one who have some sort of feedback. Chip manufacture is far from being an accurate process, some elements can have a 10% error in their values.

3) Work with people who know about the subject. Ask them questions, all the time, even if they start calling you "the annoying guy". Asking a question is easier than reading an entire book searching for the answer.

4) Divide your project in blocks. Design these blocks so that they do not depend of each other, whenever possible.