Right.If this power is at its maximum,
the corresponding impedance is the complex conjugated of the output-impedance of the actual circuit (the amplifier).
Is this right?
Correct.The same method should work to find the input-impedance of a circuit (Source-Pull?).
Show me Spectre's netlist.But with activated load-pull,
the port delivers always 0 W (RFIN is in the order of E-171 V).
What am I doing wrong?
If you don't use r0, you don't have to define r0 as design variable.why I have to fill in 50 Ohm?
At 3 places:
Z0 in the analyses form,
resistance in the port-properties,
r0 in the variables.
Hi,
if I understand this right, then Load-Pull means, that a load with a tunable impedance is connected to the output of a circuit (for example an amplifier). Then this impedance is varied and the power which is delivered to this load is observed. If this power is at its maximum, the corresponding impedance is the complex conjugated of the output-impedance of the actual circuit (the amplifier). Is this right?
The same method should work to find the input-impedance of a circuit (Source-Pull?).
I read in the "SpectreRf Workshop" PDF about "Power Amplifier Design Using SpectreRf" the chapter about "Load-Pull Measurements". I simulated the example circuit and everything was fine.
My goal is to find the input-impedance of a simple rectifier (the impedance, at which the output-DC-voltage is at its maximum).
What am I doing wrong?
I tested this with only an inductor instead of the rectifier. Same result: without load-pull everything is fine, with load-pull everything is zero.
Additional question: Can anyone explain me, why I have to fill in 50 Ohm? At 3 places: Z0 in the analyses form, resistance in the port-properties, r0 in the variables. I thougt, that the impedance is varied. But where are these 50 Ohms used?
Please help me!
Thank you!
Sascha
Show me Spectre's netlist.
Code dot - [expand] 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 // Generated for: spectre // Generated on: Jul 20 11:10:32 2016 // Design library name: RFworkshop // Design cell name: EF_example_loadpull // Design view name: schematic simulator lang=spectre global 0 parameters pin=0 fin=1G theta=0 r0=50 mag=316.7m include "/data/wms_hf_u/ghz_rfid/v2/cadence/TECH/tsmcN90rf//../models/spectre/toplevel.scs" section=tt_lib include "/homes/lischer/diode.scs" // Library name: RFworkshop // Cell name: EF_example_loadpull // View name: schematic C4 (net2 RFIN) capacitor c=10p C5 (RFOUT 0) capacitor c=10p D5 (net2 RFOUT 0) sbd_rf w=4u l=2u nf=1 m=1 D4 (0 net2 0) sbd_rf w=4u l=2u nf=1 m=1 PORT1 (RFIN 0) port r=50 num=1 type=sine freq=fin dbm=pin fundname="RF" simulatorOptions options reltol=1e-3 vabstol=1e-6 iabstol=1e-12 temp=27 \ tnom=27 scalem=1.0 scale=1.0 gmin=1e-12 rforce=1 maxnotes=5 maxwarns=5 \ digits=5 cols=80 pivrel=1e-3 sensfile="../psf/sens.output" \ checklimitdest=psf lphb loadpull inst=PORT1 rho=mag rhostart=0 rhostop=990m + rhostep=110m phi=theta phistart=-180 phistop=160 phistep=20 + z0=50 { hb hb autoharms=yes autotstab=yes oversample=[1] + fundfreqs=[(1G)] maxharms=[5] errpreset=conservative annotate=status } modelParameter info what=models where=rawfile element info what=inst where=rawfile outputParameter info what=output where=rawfile designParamVals info what=parameters where=rawfile primitives info what=primitives where=rawfile subckts info what=subckts where=rawfile save PORT1:p RFIN RFOUT saveOptions options save=allpub currents=all
If you don't use r0, you don't have to define r0 as design variable.
Z0 in the analyses form is actually used in Z0 in load-pull analysis.
The impedance which you have found by applying Load-Pull technique is NOT the complex conjugate of the output impedance of actual circuit.Optimum impedance is generally pretty different than complex conjugate of the output impedance of the circuit.Load Pull technique is very useful for Power Amplifiers and it's generally measured and simulations are pretty far from the reality due to model inconsistencies,uncertainties and rough model approximations.
I guess you're trying to find Input and Output Impedances under large signal conditions.In order to do this, you may try some simulation techniques such as PSS, HB or LSSP.They will give you a right aspect for your circuit.
Wrong.The impedance which you have found by applying Load-Pull technique is NOT the complex conjugate of the output impedance of actual circuit.
Optimum impedance is generally pretty different than complex conjugate of the output impedance of the circuit.
Can you understand a definition of S22 in LSSP ?I guess you're trying to find Input and Output Impedances under large signal conditions.
In order to do this, you may try some simulation techniques such as PSS, HB or LSSP.
They will give you a right aspect for your circuit.
Can you understand LoadPull Simulation ?Load Pull technique is very useful for Power Amplifiers and it's generally measured and simulations are pretty far from the reality due to model inconsistencies,uncertainties and rough model approximations.
No.This I don't understand.
Where is this Z0 needed during the load-pull simulation?
Is it maybe only needed as scaling factor for the smith-charts?
It is because PORT2 is load.In the Analyses-Form, PORT2 is the used Load Instance.
But in the Properties-Form of PORT2, "dc" is choosen as "Source Type".
Why?
Right.PORT1 is "sine".
That means, that RFIN is a sine-signal
Wrong.and therefore RFOUT is also a sine-signal.
It is because PORT2 is load.Why is PORT2 not "sine" as "Source Type"?
Your previous goal is different from the following.My goal is to find [color]the input-impedance of a simple rectifier[/color] (the impedance, at which the output-DC-voltage is at its maximum).
You should do SourcePull Simulation.Therefore I thought, that the load-pull-method is better to simulate the rectifier with different impedances in the port (which acts as an equivalent for the antenna).
Sorry, this is wrong.Correct.The same method should work to find the input-impedance of a circuit (Source-Pull?).
If you would like to sweep load impedance, you have to set "type=dc" for "analogLib/port".And why is a DC-type PORT used at the output in the Spectre example?
PORT2 (RFout 0) port num=2 type=dc
lphb loadpull inst=PORT2
+ rho=mag rhostart=0 rhostop=990m rhostep=110m
+ phi=theta phistart=-180 phistop=160 phistep=20
+ z0=50 {
hb hb autoharms=yes autotstab=yes oversample=[1]
+ fundfreqs=[(1G)] maxharms=[5]
+ errpreset=conservative annotate=status
}
PORT1 (RFin 0) port r=50 num=1 type=sine freq=fin dbm=pin fundname="RF"
lphb loadpull inst=PORT1
+ rho=mag rhostart=0 rhostop=990m rhostep=110m
+ phi=theta phistart=-180 phistop=160 phistep=20
+ z0=50 {
hb hb autoharms=yes autotstab=yes oversample=[1]
+ fundfreqs=[(1G)] maxharms=[5]
+ errpreset=conservative annotate=status
}
If I do a normal hb- oder pss-simulation (normal means: no load-pull)
the simulation results are as expected.
But with activated load-pull,
the port delivers always 0 W (RFIN is in the order of E-171 V).
It seems that Cadence Spectre always treats "port with type=dc" regardless of actual "type" if it is specified in Loadpull setting.But what about my problem regarding the simulation of the input impedance?
Why is RFIN (nearly) 0 V?
As workaround, try to invoke combination of "portAdapter" and "port with r=50 type=sine".
Here "port with r=50 type=sine" is required to set driving signal source.
No.
Z=Z0*(1+Gamma)/(1-Gamma), Here Z can be Zload or Zsource
Gamma=rho(Z)*exp[j*phi(Z)]
So Z0 is needed to determine Z.
Quote Originally Posted by Dr. von Rosenstein View Post
My goal is to find [color]the input-impedance of a simple rectifier[/color] (the impedance, at which the output-DC-voltage is at its maximum).
Your previous goal is different from the following.
Quote Originally Posted by Dr. von Rosenstein View Post
Therefore I thought, that the load-pull-method is better to simulate the rectifier with different impedances in the port (which acts as an equivalent for the antenna).
Quote Originally Posted by pancho_hideboo View Post
Quote Originally Posted by Dr. von Rosenstein View Post
The same method should work to find the input-impedance of a circuit (Source-Pull?).
Correct.
Sorry, this is wrong.
Source-Pull determines output-impedance of driving source signal.
Quote Originally Posted by Dr. von Rosenstein View Post
In the Analyses-Form, PORT2 is the used Load Instance.
But in the Properties-Form of PORT2, "dc" is choosen as "Source Type".
Why?
It is because PORT2 is load.
Driving direction of Loadpull is forward not reverse.
I can not understand what you want to mean at all.I am going crazy!
.....................................
Yes, I have selected the PortAdapter
as "Load Instance" in the Choosing Analyses form!
No.OK, that was what I mean with "scaling factor".
Right.If I want to measure the real circuit using a 75 Ohm System, than I should type 75 into the "Choosing Analyses" form so that I can compare the simulation results with the measurements. Right?
Right.SourcePull varies the impedance (of the power-source) which is seen by the input of the circuit.
Right?
Zin should be equal to complex_conjugate(Zsource).If one looks at the power which is delivered to the circuit, one can find a maximum of that power.
I thought, that the corresponding impedance-value which leads to that power-maximum,
should be the input-impedance of the circuit.
Is this wrong?
Right.The input signal of the Spectre Example has a frequency of 1 GHz.
So, the output signal has also a frequency of 1 GHz plus some harmonics.
Right.I thought, that to do a LoadPull-Simulation, one has to vary the impedance/Gamma at the frequency, one is interested in.
Wrong.Therefore a sin-type Port is needed, because dc-type can only dc.
Simply you can not understand spectre primitive "port" at all.That I thought and that is the reason why I am confused.
Quote Originally Posted by Dr. von Rosenstein View Post
I am going crazy!
.....................................
Yes, I have selected the PortAdapter
as "Load Instance" in the Choosing Analyses form!
I can not understand what you want to mean at all.
Quote Originally Posted by Dr. von Rosenstein View Post
OK, that was what I mean with "scaling factor".
No.
It is a reflection coefficient.
Quote Originally Posted by Dr. von Rosenstein View Post
If one looks at the power which is delivered to the circuit, one can find a maximum of that power.
I thought, that the corresponding impedance-value which leads to that power-maximum,
should be the input-impedance of the circuit.
Is this wrong?
Zin should be equal to complex_conjugate(Zsource).
Simply you can not understand spectre primitive "port" at all.
Do you mean you can now resolve the following issue ?Until yesterday,
PORT1 was the Load Instance in this form, now it is the PortAdapter.
But with activated load-pull,
the port delivers always 0 W (RFIN is in the order of E-171 V).
No.Z0 is a reflection coefficient?
Yes.Do you mean the Zsource-Value, at which the maximum power is delivered from the source to the circuit?
Do you mean you can now resolve the following issue ?
I am going crazy! I tried this and I see a sinus-signal at the net between the Port and the PortAdapter and at the net between the PortAdapter and the circuit. Good! BUT: I there is no sweep! It is always the same sinus-signal! For all rho(Gamma) and phi(Gamma)! It only depends on the values for rho(Gamma) and phi(Gamma) which I set at "Global Variables"! Yes, I have selected the PortAdapter as "Load Instance" in the Choosing Analyses form!
We use cookies and similar technologies for the following purposes:
Do you accept cookies and these technologies?
We use cookies and similar technologies for the following purposes:
Do you accept cookies and these technologies?