| Author |
Message |
alam.tauqueer
Joined: 19 Jun 2007
Posts: 128
Helped: 3
|
 01 May 2008 7:56 Latches in th design |
|
|
|
Why we avoid latches in the design, even if they provide only cell delay.
Is there any time related issues??
regards,
Tauqueer
|
|
| Back to top |
|
 |
vikas_lakhanpal27
Joined: 16 Jan 2008
Posts: 39
Helped: 1
|
| 01 May 2008 12:36 Re: Latches in th design |
|
|
|
Whenever latch is enabled it will pass watever is there on its D inputs to Q output. If suppose any glitch is coming on D and latched is enabled it will pass it to q. Glitch always create problem u would be knowing this.
Latches are fast,consumes less power, less area than Flops but Glitches can also come along with this advantages.Thats why we for flops.
|
|
| Back to top |
|
|
lordsathish
Joined: 11 Feb 2006
Posts: 248
Helped: 26
Location: Asia
|
| 01 May 2008 15:01 Re: Latches in th design |
|
|
|
| Also Latches are not DFT friendly... It is very difficult to perform Static timing analysis with latches in your design...
|
|
| Back to top |
|
|
alam.tauqueer
Joined: 19 Jun 2007
Posts: 128
Helped: 3
|
| 02 May 2008 6:27 Latches in th design |
|
|
|
Can u please explain me like
Why it is very difficult to perform STA with latches in the design....
Regards,
Tauqueer
|
|
| Back to top |
|
|
ljxpjpjljx
Joined: 05 May 2008
Posts: 209
Helped: 4
Location: Shang Hai
|
| 06 May 2008 14:32 Re: Latches in th design |
|
|
|
| latch is not good ,since STA is based on posedge of clk to do timing check and latch is level sensitive. also DFT need to do some special step to tackle this latch!
|
|
| Back to top |
|
|
alam.tauqueer
Joined: 19 Jun 2007
Posts: 128
Helped: 3
|
| 05 Aug 2008 11:14 Latches in th design |
|
|
|
can you please explain me what are those special step to tackle the latches in the DFT .
It would be great help for me to understand the problem.
Regards,
Tauqueer
|
|
| Back to top |
|
|
Tan
Joined: 23 Jul 2006
Posts: 123
|
| 06 Aug 2008 6:06 Re: Latches in th design |
|
|
|
| Please tell me how does a flip-flop not allow glitches but a latch does?
|
|
| Back to top |
|
|
natg9
Joined: 17 Jul 2008
Posts: 43
Helped: 7
|
| 06 Aug 2008 6:27 Re: Latches in th design |
|
|
|
HI
Latches will allow the data at the input to reflect at the output till the entire time the Latch is enabled , ie when it is enabled it is called to be transparent ie the output follows the input thus if a glitch appears it will be reflected at the output
but the case with FF is not so , the output follows the input only at the edge of the clock whether positive or negative .
thus any glitch appearing at the input will not be transfered to the output unless the clock edge appears
hope this has clarified your doubt
|
|
| Back to top |
|
|
santhosh007
Joined: 27 Aug 2008
Posts: 53
Helped: 15
Location: Bangalore
|
| 28 Aug 2008 9:34 Re: Latches in th design |
|
|
|
Tauqueer,
If your design is complety latch based design( usuallly IBM designs) then you can use the LSSD scan else normal scan FF based desing, you can make the latches as transparent during testmode. For example clock gating cells has the latch, you ORed with testmode signal.
non transparent latches are modeled as TIEX by the ATPG tool. So the coverage drops.
~C Santhosh Kumar
|
|
| Back to top |
|
|
kambojvikas
Joined: 12 Jul 2007
Posts: 2
|
| 22 Sep 2008 13:41 Re: Latches in th design |
|
|
|
Taqueer,
Before enable of the Latch place a 2X1 mux with sel and one pin tied with test mode signal (OR gate with testmode and actual enable signal). Intent is to make Latch always on while doing DFT coverage..
|
|
| Back to top |
|
|
