compare Spectre RF and ADS

[zmliu]

ADS for RF system design,Cadence SpectreRF for RFIC design and layout.

 

[filtershaoyong]

compare Spectre RF and @DS

for my opinion,(in IC design) you can use @DS to get pre-simulation result. but for post-simulation, the spectre is first choice, because the spectre is more accurate.

 

[Mazz]

After a lot of replies, now maybe a brief history can help.
Cadence (spectre) comes from analog integrated world, with all the costraints and benefits: complex circuits, powerful GUI, powerful layout, and parasitic extractor; it has been brought to higher frequency world in time, correcting all the problems that comes with RF. It's now the STANDARD for a lot of IC companies, with all the consequences. The development is now at this point: the limit is the parasitic extractor that does not take into account parasitic inductors (my opinion) and the PSS is not efficient.
Agilent (HP): comes from discrete MW design, with all the costraints and benefits.
Powerful Harmonic analisys, system design, "natural" for RF. Is now going toward RF IC world, coming from higher frequency. The limit is the GUI, the transient, ac analisys.
Maybe there are a lot of other things to say, but I want to summarize.

Now the two companies are coming to commercial agreement (RFDE), having the strength of both products (who has spectre can have today also RFDE), but the product is really new (bugs).

The ideal is to have both, and to practice a lot!!!!!
Yes, because to use properly the tools you need to know a lot of tricks, it's not only "netlist and run".
And, above all, a good designer is not (only) a person who knows well the tool!

I hope it can help.
Mazz

 

[v_naren]

SpectreRF's interface is lousy!!!
but I wonder if ADS can simulate LTV systems such as oscilators like SpectreRF....as if ADS cannot simulate LTV systems then it also cannot predict the phase noise effects such as up-conversion of the 1/f noise in oscilators and all!!

 

[Mazz]

ADS works perfectly on oscillators, even better than Spectre.
I have simulated the same integrated VCO (freq 4.5 GHz) with Spectre and ADS and compared it with measurements. The better fit is with ADS (phase noise).
Furthermore, using HB, you can plot the K_VCO vs freq very quickly, while in spectre it takes longer time.
I hope it can help.
Mazz