News:

Let's find out together what makes a PIC Tick!

Main Menu

RF passives simulation software?

Started by trastikata, Dec 02, 2022, 03:14 PM

Previous topic - Next topic

trastikata

Hello all,

I am testing some RF chips and I was wondering if you can recommend software that can give me impedance and frequency response from schematic depicting a network of passive elements (caps, resistors, inductors).

Thank you

JonW

What frequencies and bandwidths you are operating over?  I use keysight Pathwave (ADS) but its horrifically expensive.  Just make sure you use the S-parameter models for the R,L & C and include vias, traces plus the substrate dielectric constant as the higher in frequency you go the more errors you will introduce from the component parasitics and trace impedances. 





trastikata

Quote from: JONW on Dec 02, 2022, 04:18 PMWhat frequencies and bandwidths you are operating over?  I use keysight Pathwave (ADS) but its horrifically expensive.  Just make sure you use the S-parameter models for the R,L & C and include vias, traces plus the substrate dielectric constant as the higher in frequency you go the more errors you will introduce from the component parasitics and trace impedances. 

Thank you for the advice JONW. I'd like to test some 433 MHZ chips - LoRa, Sigfox and some other narrowband transceivers. For the test PCBs I'll be mostly following the recommended design applications but still want to see what is happening.

JonW

I suggest you get a VNA so you can measure the S-parameters into the devices and the antenna.   433M is high enough that parasitics and trace impedances will have an impact on the circuit performance.  A rule o thumb using 0402 is to add 0.2nH to the ideal capacitor to account for the inductance of the pads and a 0.1pF capacitor in parallel with an inductor to again account or the pad capacitance etc.  This should be accurate to 2GHz using ideal L & C.  Unless you are using attenuators avoid resistors as these will be lossy.  For narrow band ISM bands you should easily be able to reactively match using L & C and transmission lines.  DK of FR4 is typically 4.5 and you will need to keep the thickness between 0.2 to 0.8mm otherwise the trace widths are wide or far too narrow.

Some VNA are pretty low cost now but you will need a cal kit or make a simple one yourself.  The alternative is to rent one.




Yasin

There is RFsim99 program. It's very old but effective. If you can't find it let me know. I can share.

HAL

Hello Gentlemen

There is a site http://tonnesoftware.com/ by Jim Tonne - WB6BLD which has several programs and many papers which Jim has written and that I have used in "real world" RF designs.  They work very well.  I recommend giving them a try.....

Hal

Wimax

It depends on what you want to do, if you want to determine (by simulation) the performance of a pcb, characteristic impedance of the tracks, parasitic phenomena due to vias, couplings etc. perhaps an electromagnetic simulator is indicated.
A high-frequency circuit simulator can be useful to study the behavior of a network by having models of the components you want to use or S-parameters provided by the manufacturer or measured.
The VNA, properly calibrated and used, can give you the S-parameters of your 2 or more ports network from which to derive other parameters if necessary.
At the level of electromagnetic simulators it seems to me that a light version of Sonnet is available (a 2D and a half). I know a few high frequency circuit simulators but they are commercial products really, really expensive ( Microwave Office or ADS).


trastikata

Quote from: JONW on Dec 02, 2022, 05:06 PMI suggest you get a VNA so you can measure the S-parameters into the devices and the antenna. 

JONW, followed your advice and bought a NanoVNA. Just got familiar with the device and started playing with the VNA, took some measurements.

Would you consider the following antenna and module good match?


JonW

#8
Looks like a 2x1 switch in the module trace.  Anything better than 10dB S11 is considered good.
You will also get some degradation from the SMA launch but it looks OK to me.
Presume you calibrated the cables out?

trastikata

Quote from: JONW on Jan 30, 2023, 07:18 AMLooks like a 2x1 switch in the module trace.  Anything better than 10dB S11 is considered good.
Presume you calibrated the cables out?

Thank you for the information JONW.

The cables were indeed calibrated - all modes Open, Short, Load, Isolated and Through.

John Drew

#10
There's a very useful Windows program for the VNA (NanovnaSaver). It makes everything easier to use. It's free.
There's also a version for Android.
Clearer readouts, greater flexibility. I use it whenever a trailing USB connector can be used. The TDR function has helped me a number of times in the past tracking down bad connectors, water filled coax etc.
John