Parametric analysis in ADE-XL

Hi.

I'm new in ADE-XL and I'm looking for a way to do the following:

I run a parametric transient simulation. In each simulation I calculate a scalar value. Then I'd like to treat the multiple scalars as a vector and calculate a value based on this vector.

In ADE-L parametric sweep it was done automatically, how do I do it in ADE-XL?

Thanks

Gennady

Hi Gennady,

I think I understand what  you mean, but probably best if you can give an example? In fact better still would be to log a service request and then we can discuss in more detail (you can always put me on cc if you log it via Cadence Online Support, since I think you have my email address?)

Regards,

Andrew.

Hi Andrew,

I'm doing a very simple simulation.

I have a D-FF block which I want to characterize – find its Tsetup and Thold.

I'm running a parametric simulation while moving my clock in each simulation compared to data.

I measure delay of clock to Q of this flip flop – which is a scalar in each transient run.

When the parametric analysis is finished I have a wave of delay vs. my sweep parameter which is the delay between clock and input data.

With some simple math I can calculate Tsetup and Thold from the wave I got.

Now I want to characterize this D-FF in process corners, temp and power supply voltages. I'd like to use ADE-XL for that but I don't know how to get the waveform I described above.

Thanks

I'll open a service request for this.
____________________

Gennady Burdo

B-DeltaCom Ltd.

Tel:         +972-4-6328857

Mobile: +972-54-5252731

e-mail:   gennady@bdc-ic.com

Hi Gennady,

I thought it might be something like that. This is a feature that ADE XL doesn't yet support (the ability to do calculations based on the combined set of results - you can get the wave of delay versus sweep parameter, but not a post-processing of that waveform). It's planned for IC616 (I believe) - but please do log a service request (with me on cc) and then I can file a duplicate CCR so that you can track it, and also ensure that your needs are met by the solution.

Regards,

Andrew.

I think this is the same as http://www.cadence.com/Community/forums/p/17712/1248904.aspx#1248904 (presumably something was blocked in the system, given the flood of old posts which have appeared overnight). I'll merge it with the other post.

Regards,

Andrew.

 I find this absolutely unbelievable!! You have a great sweep tool but you can't evaluate the sweep results?

For goodness sake this needs fixing fast.

This is available in IC615 ISR6, which is ready right now. 

And note that you always could evaluate the sweep results - it was just if you wanted a calculation which worked on the complete set, rather than per point (e.g. you wanted to get the average or max of all the results, say), that you couldn't set this up in the ADE XL outputs. You always could do the calculation in the calculator after the simulation had finished.

Andrew.

Hi Andrew and Firas

OK I will see if I can use ISR6.

I take your point about doing the evaluation in the calculator – but that’s rather inelegant.

Using the optimizer is a good way to go I think. I’m trying that now.

Rob

Hi Rob,

I've asked Firas (who's handling your service request) to check that what we've implemented in ISR6 actually solves your problem.

I agree, doing it in the calculator is not so elegant - I was just trying to express that there was a way, just not as neatly integrated as it should have been.

Regards,

Andrew

So there is a solution to produce a wave from a sweep and do a measurement on that wave?

Is that available right now in 6.1.5.100 or .500?

 Sorry if the question is redundant, it seems that a solution has been found but what is it?