Download link for the Article via ANALOG DEVICES How to Model Statistical Tolerance Analysis for Complex Circuits Using LTspice by Steve Knudtsen https://www.analog.com/media/en/technical-documentatio

Hi Andy Thank you very much, I really appreciated your help. I thought about to place a voltage source in series with either input pin to simulate the offset. But before to do this, I did this questio

Yes, there have been several ways discussed recently, but the easiest is: .param Rtol 0.01 <===== Define tolerance as 1% .param R1val 10k*(1+gauss(Rtol)) <===== Gaussian with 1% standard deviation .pa

In a previous and now locked message ak za asked: 2) in monte carlo simulation with mc function, the value created by mc are uniform distribution. how can i have a normal distribution(bell shape distr

Hello Neil If it's a .model, either add bv={x}, with .param x=mc(...), or directly bv={mc(...)}. As always, the curly braces are the key. I assume even in subcircuits the .model definition will involv

FYI, I believe it is intentionally like this for a reason. While developing a simulation, it is very useful when repeated runs of the simulation behave the same. This way you can compare them and obse

Hello David, > By "controlled current sources" do you mean to add a current source across every node-pair at which I want to measure the resistance? Yes. If you don't want draw them into the schematic

I don't think you can cycle through the stepped waveforms. There are other ways to select which step to display. Some might be a little faster than others. I start with a right-click in the waveform w

Hello, > 1. ".step param run 1 50 1" , did this mean this simulation could run from 1~50 times , normally repeatedly ? This simply means LTspice will run 50 times. The parameter "run" will be increase

Hello, You could create random parameters with the common tolerance, e.g. named RTOL and CTOL. .param RFAC=(0.99+0.02*random_function) ; +/-1% .param CFAC=.... .param R1k=1k*RFAC .param R100=100*RFAC

Greetings Mercel 1) I agree with you that it is well known that paraleling discrete devices is not a good idea..... 2) But the idea of batch buying all diodes from a single wafer poses a different twi

albe, While we wait for one of the "experts around here", found you some interesting stuff: ...our messages: https://groups.yahoo.com/neo/groups/LTspice/search/messages?query=Monte%20Carlo%20ltspice .

Philipe I have uploaded an example.. temp/monte_carlo_1.asc Comments? AG

jitz wrote: It is better to upload the schematic (.asc file), not a picture of it. It looks like your schematic might be right already, or almost right. It does a Monte Carlo set of simulations with r