Continue to Site

Welcome to EDAboard.com

Welcome to our site! EDAboard.com is an international Electronics Discussion Forum focused on EDA software, circuits, schematics, books, theory, papers, asic, pld, 8051, DSP, Network, RF, Analog Design, PCB, Service Manuals... and a whole lot more! To participate you need to register. Registration is free. Click here to register now.

Register Log in

Transistor biasing using Microstrip Lines

Status
Not open for further replies.
M

mdocampo1

Junior Member level 2
Joined
Jul 6, 2010
Messages
20
Helped
3
Reputation
6
Reaction score
2
Trophy points
1,283
Location
Philippines
Activity points
1,435
I need to design a 9 GHz microstrip transistor oscillator. I plan to use ATF 34143 as my transistor. The datasheet of the transistor has S-parameters that are based on the DC operating point of the transistor. Currently I am familiarizing myself with the Ansoft Designer as my simulator.

My question is, how can I bias a transistor using only microstrip lines?

Thanks in advance :)
 

You can try to design a (lambda/4) microstrip line ended by a radial stub which acts as a shunted capacitor when properly designed.

From my experience, this is a narrow bandwidth biasing circuit, depends on your design goals.
 
I'll read about this quarterwavelength + radial stubs network. Thanks!

Another question, if I use this biasing network, will it affect the S-parameters in the given datasheet?
 

S parameters drom datasheet have been measured with bias Tee/Test fixture and are related to the transistor itself due to calibration/deembedding.

Are you saying that you are going to use this set of S parameters to design your oscillator?
 

yes sir?

though I also have a doubt since if I put my bias network I think the S-parameters will change, right?
 

I think ll be more serious to get the transistor model somewhere.

No, S parameters don't change, and you don't care about that because if I understand you want to copy S parameters from datasheet to a S2P file and put it in the simulator and create your own design all around...

Well, why not but at 9 GHz could be a tough job because not easy to tune something after manufacturing.
 
yes sir, I already got an S2P file for a certain operating condition (Vds = 3V Ids = 20mA), that's why I want to bias my transistor to that operating condition, is that wrong sir?

also sir, that's one of my "future" problems since after fabricating my design, it will be hard to tune.
 

Yes, transistor S parameters are only valid for a specific bias point.

Yes, I mentionned that future tuning could be hard due to "high frequency" operating condition. Try to assess the value of lambda.
 

You can simulate your osc circuit in ADS, and there are some exapmles in ADS.
 

I observed that many of you guys use ADS for microwave circuits.

Unfortunately, we don't have ADS here, we only use Ansoft Designer since we have a license for it.
 

From my understanding (Please correct me if I'm wrong), radial stubs serves as bypass capacitors in a way that it provides a short circuit. However, this will provide connection between the DC and RF power so there will be a quarter wavelength line that will make this short circuit appear as RF open, and thus providing isolation from the DC supply.

I also saw the properties of a radial stub.

The radius of the radial stub determines the attenuation frequency while the angle affects the bandwidth.

The information I've got are mostly about the DC-RF isolation issues. I did not see anything or at least an equation regarding the DC biasing using the microstrip line.

My question is, is it right to represent the microstrip lines as an impedance then use Kirrchoff's Voltage Law to compute for the DC operating point?
 

You need a bias circuit independent from your 1/4 lines. Take a look at this **broken link removed** and the attachment to get some ideas. With respect to bias circuit, L1 & L3 of Fig 2 of AN 1175 basically are your 1/4 lines.
 

Attachments

  • Fujitsu HighPower GaAsFET Bias Considerations.pdf
    98 KB · Views: 298
  • Like
Reactions: tony_lth

    tony_lth

    tony_lth

    Points: 2
    Helpful Answer Positive Rating
    V

    Points: 2
    Helpful Answer Positive Rating
I forgot to check the stability of my transistor. Using the K-delta criteria:

K = (1 - |S11|^2 - |S22|^2 + |delta|^2) / 2|S12 * S21| > 1
|delta| = |(S11 * S22) - (S21 * S12)| < 1

for unconditional stability.

When I computed my K and delta, I got K = 1.156 and |delta| = 0.4277 that satisfies the criteria. Meaning, at 9 GHz, my transistor is unconditionally stable.

How can I make this unstable at 9 GHz?

I saw an example from Pozar's Microwave Engineering book that he added an inductor at the gate and then use a common gate configuration to make his transistor unstable at 4 GHz. He did not mention anything on how to do that, he just added that inductor and used a common gate topology.
 

You are designing an oscillator not an amplifier. If your circuit is unconditionaly stable then this is not an oscillator...

Could you provide the related circuit. Something should be wrong in some L, C values or in the loop...

PS : When you perform a small signal stability (K&delta) analysis, take care of the frequency range). Simulation from 0 Hz to few harmonics is needed to be sure that any start up of oscillation can occur (for amplifiers).
 

Sir I only calculated the K delta from the given S-parameters of the transistor at 9 GHz. I found out that it is unconditionally stable at 9 GHz so I need to do something about it to make it unstable at 9 GHz, which I am not familiar with.

The K delta was computed manually. I did not use any simulator for that sir because there are equations based on Pozar's Microwave Engineering book.
 

Ok, there is a misunderstanding.

I was thinking you were computing K from S parameters of your complete oscillator. There is no reason to calculate K of your single transistor. Try to do your oscillator with this transistor and see what happens. If the selected transistor is unconditionnally stable at 9 GHz doesn't mean you can't build an oscillator with it!

Calculating K at 9 GHz frequency from of the transistor doesn't give you significant information for the incoming design.
 
owwww. thank you sir! I was very worried a while ago because of this. thanks sir! I'll be studying the biasing network for my oscillator. GREAT HELP sir :D
 

Status
Not open for further replies.

Similar threads

Part and Inventory Search

Welcome to EDABoard.com

Sponsor

Back
Top