Maker Pro
Maker Pro

type of fault in vlsi circuit ?

Status
Not open for further replies.

vead

Nov 27, 2011
473
Joined
Nov 27, 2011
Messages
473
How many type of fault can be occur in VlSI circuit ?

my effort
physical fault during manufacturing process
circuit fault during interconnection
packing fault

Q1 does anyone know another fault ?
Q2 what is fault model and why does we need in vlsi circuit?
 

Harald Kapp

Moderator
Moderator
Nov 17, 2011
13,728
Joined
Nov 17, 2011
Messages
13,728
Some that come to mind:
  • fpga: single event upset (Google)
  • operatig under wrong conditins (overvoltage, overtemperature...)
  • electromigration (Google)
  • metastable states (Google)
  • faulty design
  • ...
 

vead

Nov 27, 2011
473
Joined
Nov 27, 2011
Messages
473
you did not understand actually what I am asking

when we make chip It may be work or it may be fail due to some fault
chip may be fail due to fabrication process
chip may be fail due to circuit connection
chip may be fail due to packing

how we will detect fault in chip and what is fault model in vlsi system
 

Harald Kapp

Moderator
Moderator
Nov 17, 2011
13,728
Joined
Nov 17, 2011
Messages
13,728
A typical fault model ist the stuck-at fault. One assumes that a net is either permanently high (stuck-at-1) or low (stuck-at-0). One simulates the chip with and without fault and compares the results. If the results are different, the fault can be detected. If the results are the same, the fault is not detetcted by this test pattern and a new test pattern is created. Creating good test patterns is a science in itself.
There are other types of stuck-at faults (stuck open...) but they are increasingly difficult to simulate. Often the simple stuck-at-1/0 model gives satisfying results.
Large VLSI circuits also contain additional test circuits. E.g. a ring oscillator , here the frequency of the oscillator is a measure of the quality of the manufacturing process. Or test multiplexers that are activated to gain acces to internal circuit nodes that are otherwise not available for testing (an interesting note on the side: sometimes special test points that need to be accessed after a chip has been assembled are connected to pins which are labeled NC - not connected. This is typical e.g. for voltage references for trimming purposes. One should really not connect these pins to any external circuit because the operation of the chip can be unpredictable. But I stray...)
 

vead

Nov 27, 2011
473
Joined
Nov 27, 2011
Messages
473
I am little bit confused about fault , fault model , and testing

fault - component fault , circuit fault
fault model - delay fault, stuck at fault
testing - testing for component , testing for circuit connection

can you explain little bit all three things in vlsi system
 

Harald Kapp

Moderator
Moderator
Nov 17, 2011
13,728
Joined
Nov 17, 2011
Messages
13,728
I don't understand waht you don't understand:

fault - the actual fault, whatever went wrong durcing production
fault model - a simplified model how a real fault will influcence the operation of a circuit
testing - find out if and where a fault is in the real product

There is no real difference for these things between a simple logic circuit and a huge VLSI circuit - only the size of the circuit.
 

vead

Nov 27, 2011
473
Joined
Nov 27, 2011
Messages
473
I don't understand waht you don't understand:

fault - the actual fault, whatever went wrong durcing production
fault model - a simplified model how a real fault will influcence the operation of a circuit
testing - find out if and where a fault is in the real product

There is no real difference for these things between a simple logic circuit and a huge VLSI circuit - only the size of the circuit.
VLSI design are automated

customer requirement
specification
functional design (by using hardware description language )
digital design (using gate level model )
circuit level (using gate are implement using transistor )
layout design
fabrication process ( layout is convert into real chip )
chip manufactured

suppose this chip is not working
how we will determine fault in chip
 

Harald Kapp

Moderator
Moderator
Nov 17, 2011
13,728
Joined
Nov 17, 2011
Messages
13,728
VLSI design are automated

No. There are a lot of tools to partly automate routine work and to assist the developer. But in the end, it is still a human developer who has to interpret the requirements and translate them into something the VLSI design tools can "understand".

suppose this chip is not working
how we will determine fault in chip
BY testing, as I explained before (see especially post #4)
 

vead

Nov 27, 2011
473
Joined
Nov 27, 2011
Messages
473
I understood testing is used to determine fault in chip
fault may be occur during fabrication , during connection , during packaging

Q1 If fault is present in component , which testing method we will use to detect fault ?
Q2If fault is present in interconnection , which testing method we will use to detect fault ?
Q3 If fault is know, can we recover fault in chip ?
 

(*steve*)

¡sǝpodᴉʇuɐ ǝɥʇ ɹɐǝɥd
Moderator
Jan 21, 2010
25,510
Joined
Jan 21, 2010
Messages
25,510
A1 Your testing of functionality should show up errors in the functionality. Faults which do not change the functionality may not be significant.

A2 The same testing

A3 No, unless you have built something into the design to allow for this.
 

vead

Nov 27, 2011
473
Joined
Nov 27, 2011
Messages
473
A1 Your testing of functionality should show up errors in the functionality. Faults which do not change the functionality may not be significant.

A2 The same testing

A3 No, unless you have built something into the design to allow for this.
Q1 when does we do testing after manufacture or before start up circuit

Q2 chop is not working we don't know what is fault
what is process to determine fault
I mean which method we will use step by step ?

1. path sensitization
2. D algorithm
3. full scan & partial scan
4 built in self test
 

(*steve*)

¡sǝpodᴉʇuɐ ǝɥʇ ɹɐǝɥd
Moderator
Jan 21, 2010
25,510
Joined
Jan 21, 2010
Messages
25,510
I presume you would do testing at all stages.

The chips would be tested on the wafer before they're even split up into individual pieces.

Then there would be testing after they have the bonding wires attached (probably after they're encapsulated)

If they require programming then perhaps you test them after programming.

And you would certainly run some tests after the device is built.

The method you use could be any or all of the methods you propose. There is no magic bullet.

You need to determine if the fault is due to

  1. Design
  2. Manufacture
  3. Programming
  4. Something else

Your simulations should take care of (1), bu note that simulations are not perfect. For errors in manufacture, this will either be systematic (failure of (say) the mask) that might be determined only by examination of the chip under a microscope, or they might be random (in which case only some devices will be affected, and the fauults are likely to differ). If it's a fault in programming then you might have a trivial "hello world" application that you could use to verify that you have the basics working, and beyond that you're into a typical software debugging cycle. If it's (4) (maybe damage due to wrong supply rail, poor anti-static handling, etc) then you need to determine how it's happening and to fix that.
 

vead

Nov 27, 2011
473
Joined
Nov 27, 2011
Messages
473
I presume you would do testing at all stages.

The chips would be tested on the wafer before they're even split up into individual pieces.

Then there would be testing after they have the bonding wires attached (probably after they're encapsulated)

If they require programming then perhaps you test them after programming.

And you would certainly run some tests after the device is built.

The method you use could be any or all of the methods you propose. There is no magic bullet.

You need to determine if the fault is due to

  1. Design
  2. Manufacture
  3. Programming
  4. Something else

Your simulations should take care of (1), bu note that simulations are not perfect. For errors in manufacture, this will either be systematic (failure of (say) the mask) that might be determined only by examination of the chip under a microscope, or they might be random (in which case only some devices will be affected, and the fauults are likely to differ). If it's a fault in programming then you might have a trivial "hello world" application that you could use to verify that you have the basics working, and beyond that you're into a typical software debugging cycle. If it's (4) (maybe damage due to wrong supply rail, poor anti-static handling, etc) then you need to determine how it's happening and to fix that.
VLSI design

customer requirement
specification
functional design at register transfer level using some hardware language (verilog , vhdl )
fault or error detect by functional simulation

next step logic design
functions are decompose into gate level
logic simulation
error is detect by truth table, boolean equation method

next circuit design

logic circuit made with cmos transistor
circuit simulation

layout design geometric pattern

layout verification

fabrication process ( layout is convert into real chip )

chip manufactured

next we will test the chip to ensure that chip is working correctly or incorrectly

In vlsi design we have detected all error by simulation
In manufacturing we are testing wafer

I am confused .I don't understand at which level we will use following method

1. path sensitization
2. D algorithm
3. full scan & partial scan
4 built in self test
 

(*steve*)

¡sǝpodᴉʇuɐ ǝɥʇ ɹɐǝɥd
Moderator
Jan 21, 2010
25,510
Joined
Jan 21, 2010
Messages
25,510
Vead, it's a simple 2 step process.

  1. Find out what's wrong
  2. Fix it.
With your question finally answered we can close this thread.
 
Status
Not open for further replies.
Top