LDE parameters with APS simualtor

[yvk]

Hi,

While doing post layout simulations using APS and merging mfactor devices what are the effective parameters for Layout dependent effects like sa,sb,sc etc.

I guess its not suggested to use this option of merging mfactor devices because those effects are different for every instance.

Thanks,
yvk

 

[Andrew Beckett]

On 02/16/13 18:49, yvk wrote:

Hi,

While doing post layout simulations using APS and merging mfactor devices what are the effective parameters for Layout dependent effects like sa,sb,sc etc.

I guess its not suggested to use this option of merging mfactor devices because those effects are different for every instance.

Thanks,
yvk

I don't think it will merge them if the parameters are not identical.

Andrew.

 

[yvk]

So you think that the parameters are not merged when not identical.
Then theres no meaning for merging mfactor devices??
so would it override the settings?

I only wanted to know if theres any difference in sim results with and without using merge options in APS with post layout.

Thanks,
yvk

On Sunday, February 24, 2013 7:24:41 PM UTC+5:30, Andrew Beckett wrote:

On 02/16/13 18:49, yvk wrote:

Hi,



While doing post layout simulations using APS and merging mfactor devices what are the effective parameters for Layout dependent effects like sa,sb,sc etc.



I guess its not suggested to use this option of merging mfactor devices because those effects are different for every instance.



Thanks,

yvk





I don't think it will merge them if the parameters are not identical.



Andrew.

 

[Andrew Beckett]

On 02/24/13 15:23, yvk wrote:

So you think that the parameters are not merged when not identical.
Then theres no meaning for merging mfactor devices??
so would it override the settings?

I only wanted to know if theres any difference in sim results with and without using merge options in APS with post layout.

Thanks,
yvk

There should be no significant difference between results (there may be
a slight difference within tolerances because the matrices will be
different and so the result could converge slightly differently; if the
circuit has multiple stable operating points it's also possible that you
might not get the same operating point. In general though the results
should be the same, which is why we do it - and even if the results are
different you could have got that answer without merging just by
altering something else minor).

Regards,

Andrew.

 

[yvk]

Thanks Andrew,

Ok I will take it granted that merge option shouldnt alter the sim results.(as its not any bistable OP circuit)

However trying to understand your statement ""it doesnt merge them if parameters not identical"" so the advantage of using APS for speeding up the sims is lost then??

-yvk


On Monday, February 25, 2013 3:05:59 AM UTC+5:30, Andrew Beckett wrote:

On 02/24/13 15:23, yvk wrote:

So you think that the parameters are not merged when not identical.

Then theres no meaning for merging mfactor devices??

so would it override the settings?



I only wanted to know if theres any difference in sim results with and without using merge options in APS with post layout.



Thanks,

yvk





There should be no significant difference between results (there may be

a slight difference within tolerances because the matrices will be

different and so the result could converge slightly differently; if the

circuit has multiple stable operating points it's also possible that you

might not get the same operating point. In general though the results

should be the same, which is why we do it - and even if the results are

different you could have got that answer without merging just by

altering something else minor).



Regards,



Andrew.

 

[Andrew Beckett]

On 02/25/13 06:57, yvk wrote:

Thanks Andrew,

Ok I will take it granted that merge option shouldnt alter the sim results.(as its not any bistable OP circuit)

However trying to understand your statement ""it doesnt merge them
if
parameters not identical"" so the advantage of using APS for speeding up

the sims is lost then??

-yvk

No, because this optimization is only one of many things that makes the
simulations faster, including the core algorithms and support for
multi-threading (although you should see some speed up even without
multi-threading).

Regards,

Andrew.

 

[yvk]

Thats good to know.

Thanks,
yvk

On Tuesday, February 26, 2013 2:02:47 AM UTC+5:30, Andrew Beckett wrote:

On 02/25/13 06:57, yvk wrote:

Thanks Andrew,



Ok I will take it granted that merge option shouldnt alter the sim results.(as its not any bistable OP circuit)



However trying to understand your statement ""it doesnt merge them

if

parameters not identical"" so the advantage of using APS for speeding up

the sims is lost then??



-yvk





No, because this optimization is only one of many things that makes the

simulations faster, including the core algorithms and support for

multi-threading (although you should see some speed up even without

multi-threading).



Regards,



Andrew.