How to get the desired Noise Figure Circle and Gain circle for Low Noise Amplifier.
I designed simple low noise amplifier for 2DB Noise Figure and Maximum gain ( given in David .M.Pozer book Example 11.5 page no559 third edition)for the following S-parameters at 4GHz, used FHX35LG HEMT Transistor.
S(1,1) S(2,1) S(1,2) S(2,2)
mag angle mag angle mag angle mag angle
4GHz 0.6 -60 1.9 81 0 0 0.5 -60
F(min) = 1.6 dB, Noise Resistance of transistor = 20 Ohm, optimum reflection coefficient of transistor = mag 0.62 angle 100
I used AWR MWO software tool and simulated , didn't get proper gain and noise figure circles. Here with I am attaching the desired noise figure circle and gain circle and obtained NF circle and gain circle. I don't know where I am doing mistake, please can you help me get the desired NF circle and maximum gain.
hy,
first of all, i don't see any port for analyze, because for a circle is necesarry to have a port for meassurement.
Also, i don't see which is the reason that you put some MLEF after resistance?i don' know if is good to put there.And also i see that the source is 0V.So i don't know if the design is good.Maybe you should consider to replace youre design, because just for example gor gain circle toy must have 2 ports and you haven't.Try to read some help, user guide in MWO.
hope i was usefull
Dear Sir,
It is a example given in MICROWAVE ENGINEERING by DAVID.M.POZER (Page no 559, Example 11.5), I try to get it , but couldn't get it. MLFE is used for open circuited stub, if we add port , I didnt get anything. As you told source is zero , If I make it source to some value , nothing effect I found. If we change the DC bias ,then earlier s-parameters value will change.
- First, you re anot providing any DC bias to the drain;
- Second, DC voltage at gate is zero, is that what you want?
- Third, from the two above, looks like you just cared for small signal behaviour, you forgot DC bias;
- Fourth, I am not used to Microwave Office, but I believe there is some kind of test ports specific for S-parameters simulation, without having to add 50 ohm resistors in source and load terminals;
Good luck.
Dear Sir,
I modified the design as you told, am getting better results than the earlier. But I didn't get the expected results. Here with I attach MWO file
But you still have no bias nor bias network...
Dear sir,
I added dc bias to our ckt , am getting nearer to result. But I didn't get the desired result.Here with I attach the MWO file .
Now you have DC at the drain, but remember that for sinusoidal signal that voltage source is a short-circuit! You must add some kind of AC choke between DC voltage source and the drain terminal. Usually an inductor or, if you are working at a single frequency (or at least within a small bandwidth) a quarter-wavelength line does the trick.
Moreover:
- You still have no bias at the gate;
- Drain voltage is 4 volt, which is the maximum absolut rating for that device (dangerous in reality, not so in simulation). A more conventional voltage of 2V is adviseable.
etmabreu sir,
As you told sir, I added DC bias at the both gate, drain with 2 volts and quarter wavelength. I am sorry , couldn't get the desired result, but somewhat better ,better result than the earlier. Here with I am attaching the MWO file
btv_murthy, if you look at the transistor's datasheet, you will find it is a depletion type MOSFET. Hence, your bias voltage at the gate must be negative.
In fact, you must choose a suitable (negative) gate voltage so you can get the Ids quiescent current you need.
Are you sure those microstrit lines are a quarter wavelength long? They just seem too identical to a microstrip line you have at the output matching network. If you prefer you can place an inductor instead of a microstrip line for this purpose. Perhaps your simulator includes some component like "DC feed" or "RF choke" which are meant to simulate as perfect conductors at DC (bias) but perfect open circuits at AC (signal).
Dear Sir,
Those are quarterwavelength microstrip lines only. DC feed or RF choke is not compatible then I used the inductor , what inductor value I have to substitute . I have substituted L = 50nH. Herewith I am attaching the MWO file
If you do not have a Transistor that has the same S-Parameters then you will not get the
same results. That example says it's a GaAs FET, also you will need to have a FET with the
same /similar Optimal Noise figure. If you do not have this then you should not expect the
same results !
Your Amp might have a BW of 100MHz - 4GHz and the Amp in the Book might have a
3GHz - 5GHz operating range... There are a lot of things that can cause different results.
btv_murthy, those 50nH are fine, since their impedance is around 1200 ohm at 4 GHz, so high enough compared with the common impedances of the microwave circuits (around 50ohm).
You just need to check the DC voltage at the gate, make it to some negative value that gives the desired Ids you chose during design.
Doing this, if your input and output matching networks are properly designed, and if your transistor is accuratelly modelled, then the circuit simulation should behave fine.
Element_115 Sir,
I argee with you, didn't use the GaAs FET transistor. I used HEMT transistor and changed the S -parameter at 4 GHz. If you know the GaAs FET transistor for the 4 GHz , Can u help me to tell, what is number of transistor for GaAs FET for the s-parameter. Like I used FHX35LG for the HEMT transistor.I can search for it.
Dear Sir,
I changed the HEMT transistor to GaAs FET transistor i.e f101002A. I copied that f101002A file, changed the S-parameter values at 4GHz,then also I am getting same results as earlier getting .
Dear Sir
I made it negative value at the gate, have to get nearer to desired result. I used the GaAs FET i.e F101002A transistor . Here with I am attaching the MWO files
Ok, did you change the Noise Opt values too ? If you did, then you can do the hand calculations
that are in the book and you should get the same answers. Now your simulation tool might be
taking in to account things that the Book does not discuss because of the complexity.
- - - Updated - - -
Ok, did you change the Noise Opt values too ? If you did, then you can do the hand calculations
that are in the book and you should get the same answers. Now your simulation tool might be
taking in to account things that the Book does not discuss because of the complexity.
Element Sir,
Earlier I was not changed the Noise optimum values, now I changed the noise optimum values at 4GHz, then I simulated . Still little near results has to come.Here with I am attaching the MWO files