[SOLVED] Mixer Modeling in RF Systems

Hello everybody,

I am trying to model a minicircuits mixer MAC85 in ADS as part of a Rx chain. I am confused by some notations on datasheet and how to apply them to my case.

(**broken link removed**) Datasheet of mixer


1-) In ADS, mixer component asks for port reflections in polar while datasheet is providing the RF/LO and LO/IF isolations. How can I go from isolation to reflection of ports?? ( I have an idea going from VSWR to reflection coefficient but im not sure how to do it for all the datapoints.)

2- )My RF_Freq is the received signal while LO signal is my original signal before transmission feeded into mixer by a power divider. Hence I am aiming to go to baseband(someone experienced on this?), with some doppler frequency shift. Can anyone show me a way or some sources on how to tackle the simulation ?

Kind Regards,

ktr

ktr said:

How can I go from isolation to reflection of ports??

Click to expand...

Consider 3 ports S-parameter for frequency conversion device.

ktr said:

( I have an idea going from VSWR to reflection coefficient

Click to expand...

Right. You can determine magnitude of S11, S22 and S33.

ktr said:

but im not sure how to do it for all the datapoints.)

Click to expand...

What do you want to mean ?

ktr said:

Hence I am aiming to go to baseband(someone experienced on this?),

Click to expand...

What do you want to mean ?

ktr said:

with some doppler frequency shift.

Click to expand...

There is no relation between mixer and doppler shift.

Anyway you can build your own custom model by SDD, FDD or Verilog-A.

I think model written by FDD is suitable for your purpose.

"Right. You can determine magnitude of S11, S22 and S33."

Click to expand...

I determined S-parameters(11,22,33) of the device, but my data item only accepts a single s-parameter file. Any advice on how to combine them into a single file, although I dont have the S12, S21, S31 etc.

What do you want to mean ?

Click to expand...

I mean S-parameters are a long touchstone file list for given freqeuncies but the generic mixer device in ADS has these. I dont know what they mean.

It has SP11, SP22 and SP33 as complex and single values for ports.

What do you want to mean ?

Click to expand...

This is my setup to test mixer and when it downconverts the remaining signal should just go to baseband, but I am not sure what type of simulation I should run to get correct results at output.



When I say modelling, im talking about this primitive approach though, not the extended spice model or sth. I should say, verilog-a or fdd are not things that I know.

Regards,

ktr

ktr said:

Any advice on how to combine them into a single file,
although I dont have the S12, S21, S31 etc.
I mean S-parameters are a long touchstone file list for given freqeuncies
but the generic mixer device in ADS has these.
I dont know what they mean.

Click to expand...

I can not understand what you want to mean at all.
Maybe no one can understand it.

However use DAC(Data Access Component).

ktr said:

but I am not sure what type of simulation I should run to get correct results at output.

Click to expand...

If you truely require baseband signal processing, there is no option except for Transient Analysis.
And you can not use system model prepaired in ADS, since they are frequency domain models.
You have to build your own custom models based on time domain descriptions.

ktr said:

I should say, verilog-a or fdd are not things that I know.

Click to expand...

Learn them.

Many system models prepaired in ADS are written by FDD.

SDD and FDD are behavioral modeling ability of ADSsim native.
On the other hand, Verilog-A is generic and is not restricted to ADSsim.
However you can not utilize full ability of ADSsim if you use Verilog-A.

SDD and Verilog-A are time domain modeling.
FDD is frequency domain modeling.

Advice on FDD and SDD is taken, will spend time on them.

I used data components on ADS before, but they were mostly on 2port devices.

Sorry if I could not get my point through, I should do a better job, you are correct.



1- Those complex S11 etc numbers are what confusing me.

2- Now I have 3 seperate files for S11, S22 and S33 for my mixer. And I dont know how to put all 3 of them into a single data component(is that possible?)
(File input screen is for the 3port device that is standing alone in the right, not connected to the circuit)

ktr said:

I used data components on ADS before,
but they were mostly on 2port devices.

Click to expand...

Absolutely wrong.
Data components are for any numbers of port.

However use DAC(Data Accesss Component) not Data component.

ktr said:

1- Those complex S11 etc numbers are what confusing me.

Click to expand...

I can not understand what you are confusing at all.

ktr said:

2- Now I have 3 seperate files for S11, S22 and S33 for my mixer.
And I dont know how to put all 3 of them into a single data component(is that possible?)

Click to expand...

Use DAC(Data Accesss Component).

Absolutely wrong.
Data components are for any numbers of port.

Click to expand...

In Data Items section there are N-port devices, I used 2-port device there as a means of importing transistor s2p's many times, thats what I am saying.
But still, in that file there were all 4 S parameters of a 2-port.

I can not understand what you are confusing at all.

Click to expand...

I am confusing the thing: This mixer will work over a frequency band. So the ports of the mixer should have different reflection coefficients for different frequencies.
But, ADS asks me to write a single complex value. That is the confusing part, it should be useless entering a single value since it should only be meaningful for a single frequency.

What am I missing here?

use DAC(Data Accesss Component).

Click to expand...

This one still accepts only a single file.

ktr said:

But still, in that file there were all 4 S parameters of a 2-port.

Click to expand...

I can not understand what you want to mean at all.

ktr said:

But, ADS asks me to write a single complex value.
................................................................................................
What am I missing here?

Click to expand...

Use DAC(Data access component).
There "freq" is used as independent variable.

ktr said:

This one still accepts only a single file.

Click to expand...

Simply use multiple DACs.

DAC can treat Generic MDIF File.
Touchstone fomat is no more than subset of Generic MDIF.

Survey MDIF and Citi File.

I can not understand what you want to mean at all.

Click to expand...

If I had a single s-parameter file for a 3-port network with all the Sij elements in it, it would be simple.
But I have 3 seperate files, and I dont know how to implement that.

Use DAC(Data access component).
There "freq" is used as independent variable.

Click to expand...

I have no idea what that means.

DAC can treat Generic MDIF File.
Touchstone fomat is no more than subset of Generic MDIF.

Click to expand...

Should I feel ignorant for not knowing these? I learned about s2p format, and it was enough for all this time.

All I am trying to do is simulating a mixer, but I can't figure out how to make my seperate s-parameter files import into "something" ( whether this is a DAC, or a mixer model):bang:





Some guidance on what I should do would help.

ktr said:

If I had a single s-parameter file for a 3-port network with all the Sij elements in it,
it would be simple.
But I have 3 seperate files, and I dont know how to implement that.

Click to expand...

Simply combine 3 seperate files to one file as Generic MDIF or Citi file.
You can do this tasks by simple script such as AWK, perl, python, etc.

Or you can do this tasks by ADS's AEL.
Here { } operator is invoked.

ktr said:

I have no idea what that means.

Click to expand...

Surely search ADS example.
You can see many examples of DACs there.

ktr said:

Should I feel ignorant for not knowing these? I learned about s2p format,
and it was enough for all this time.

Click to expand...

s2p format has no value.

ktr said:

Some guidance on what I should do would help.

Click to expand...

It is too easy, so no one can explain more.