Hi, I'm currently trying to upgrade my Spectrsys simulation and need some advice on a couple of components that I'd like to "improve" over the existing models.

1) RF switch: All of the RF switch models I can find in Spectrasys are passive. In other words they do not include the effects of compression, TOIP etc. I think that the solution to this problem is to include a zero gain amplifier stage to simulate these effects. Question; which is the best model to use? Secondly should it be before or after the switch model? The losses through the switch are less than 1 dB and normally I would not worry about this but as I am discovering Spectrasys does need to be treated carefully otherwise you can get some odd results.

2) DC variable RF attenuator: Same sort of problem, the attenuator I am using has an odd P1dB performance. The P1db performance is control voltage dependent (no suprise) but it is very non linear. It has a high P1dB compression at top and bottom of the control voltage range and is low in the middle. Any ideas on how I should go about simulating this? I did consider an equation based model but can't get my head around the equation to model the Vcontrol vs P1dB

Many thanks

Patrick

1) RF switch: All of the RF switch models I can find in Spectrasys are passive. In other words they do not include the effects of compression, TOIP etc. I think that the solution to this problem is to include a zero gain amplifier stage to simulate these effects. Question; which is the best model to use? Secondly should it be before or after the switch model? The losses through the switch are less than 1 dB and normally I would not worry about this but as I am discovering Spectrasys does need to be treated carefully otherwise you can get some odd results.

2) DC variable RF attenuator: Same sort of problem, the attenuator I am using has an odd P1dB performance. The P1db performance is control voltage dependent (no suprise) but it is very non linear. It has a high P1dB compression at top and bottom of the control voltage range and is low in the middle. Any ideas on how I should go about simulating this? I did consider an equation based model but can't get my head around the equation to model the Vcontrol vs P1dB

Many thanks

Patrick

2) In version 2008.01 new parts were created for building user models. The part you want to use is called, 'NonLinear Block'. You can create your own user model with the attenuator and this block. From the new user model you can create a new part and place it in your own part library. Please refer to the shipping example \Examples\SPECTRASYS\User Models\Non Linear Filter User Model.wsx that shows the same process you want to follow except using the attenuator instead of a filter.

Also you may just want to use the variable gain amplifier which is a DC controlled attenuator followed by an amplifier. See the shipping example \Examples\SPECTRASYS\Amplifiers\Simple Variable Gain Amp.wsx for more information.