Do you Believe Cadence virtuoso?

[020170]

when I desinged circuit using Hspice, I thought Hspice is not convenient to design more complex circuits, for example DAC or ADC.

so nowadays I have used Cadence Virtuoso Schematic composer.

but I have some question about using cadence.

when I designed some amplifier using Hspice, I would like to transfer into Cadence. because cadence is so easy to use.

when I completed transfering my amplifier into Cadence, I realize that my amplifier is not saturation region ! is it possible?

some TRs in Amp is not saturation region, so I'm confused.

do you believe cadence? or Hspice?

I don't know whether accurate simulator is.

thanks.

 

[VSMVDD]

i think by definition a simulator is only as good as what it simulates
very often countrys release inferiour transitors to other countrys and 'tamper' with data

this is obvious

what isnt obvious is that simulators are only as good as there programmers
you and me my pal only see the shinny body work while under the bonet is
anything from a ram jet to a birds nest... with wheels


so i guess it is best to make a few designs you know best
like the one you have

and look thru the sims till you find one that gives good results

after all i can dig thru a box of exactly the same parts and only find a handfull that match the needs never mind the spec

as do simulators in general.....!!!!

the best simulators is maths and metering

 

[huojinsi]

By far,i very believe Cadence virtuoso.The reason is i only use virtuoso to design and haven't use other simulation tool!

 

[|IAngel|]

Cadence is a buggy program. This being said, it's all about the models. I've designed some pretty "wild" (weak inversion, UBER low voltage, nA) type circuits and they were very close to the models, and therefore the simulations.

It's all about the models for accuracy. Have a good kit designed for Cadence, you'll get wicked results. Have a bad kit, you are in trouble.

That being said, I've read H-Spice is more accurate with proper models than Spectre, but with BSIM3.3 and Spectre, you can get quite awesome simulations.

 

[sunking]

it is a good tool for analog circuit design.
maybe the option is not rightly set. check it agian

 

[eng_Semi]

Cadence is the best

 

[Teddy]

The question is little confusing. If you do schematic entry it has nothing to do with the simulation. Then you create netlist in analog environment.

I suppose you use Spectre for your cadence sims. And so if you have different results it could be issue of spectre models.

What you can do is to switch from spectre to Hspice and use it instead of spectre for simulations. Or you can dump the Hspice netlist and re-run sims of your design in Hspice and compare.

In general if the models are good you should get similar results

 

[flushrat]

Yes, I believe.
Check you schematic. i.e the bulk connection ...

 

[flamingo]

I believe them all.
In my experience, there are indeed some little differences of results given by different simulators, but they are in the range of target. Check your schematic carefully, and you can try some relatively simple blocks (current mirror, single-transistor amplifier, etc.) to validate first. Gook luck.

 

[sankudey]

I agree with |IAngel|.....for MOS...Bsim3v3 and for BJT/HBT VBIC model work good for cadence...some times also the kit provider may refer/prefer some simulator name(s)....you could check that also....

 

[beckchm]

I think they are all correct.
There are few differences of results given by different simulators. The problem is that your design does not match the simulators .Check your schematic carefully.

 

[gevy]

Simulation “truths” from Allen:
- (Usage of a simulator) x (Common sense) ≈ Constant
- Simulators are only as good as the models and the knowledge of those models by
the designer
- Simulators are only good if you already know the answers