Hi,

Im very new to ADS (used cadence for 15 years...)

I need to insert a high pass filter (at baseband) in an simulation I copied from "examples".

1) Since I could not find a timed HP filter I made a new block wich

summed the BB signal minus a low pass filtered version of the BB signal. mathematically this is OK

but is this the way to do it in ptolemy?

2) I got warnings about having to long Tstep.

Changing the Tstep in FloatToTimed was not correct

since it seems to change tho whole sampling process for FFT etc.

So instead I UpSample of 4 before the filter and DownSampe by 4 after. Is this OK?

Will this UpSample be a zero order hold or a first order hold?

/CG

Im very new to ADS (used cadence for 15 years...)

I need to insert a high pass filter (at baseband) in an simulation I copied from "examples".

1) Since I could not find a timed HP filter I made a new block wich

summed the BB signal minus a low pass filtered version of the BB signal. mathematically this is OK

but is this the way to do it in ptolemy?

2) I got warnings about having to long Tstep.

Changing the Tstep in FloatToTimed was not correct

since it seems to change tho whole sampling process for FFT etc.

So instead I UpSample of 4 before the filter and DownSampe by 4 after. Is this OK?

Will this UpSample be a zero order hold or a first order hold?

/CG

For more accurate high pass filter modeling, we recommend that a cosimulation with an A/RF high pass filter is done.

2) You are correct that just changing the Tstep in the FloatToTimed converter is not the correct way to correct the warning you saw. Upsampling the signal is the right way. Which upsampler did you use and how did you set its paramters? For example, using the USampleRF with Type=SampleAndHold will result in a zero order hold.