Design method for Colpitts oscillator at 2 GHz
I'am currently designing a Colpitts oscillator around 2GHz. For this I am using the design method described in the following book:
Microwave Circuit Design using linear and nonlinear techniques by Vendelin,Pavio and Rohde.
In this book(p.408-413) they derive a formula to calculate the resistance wich has to be put in series with a inductor and the necesary capacitors for the oscillation frequency.
Formula for Rs(series resistance):Rs=gm/w^2C1C2. (with w=pulsation freq.)
Now my problem: As I am using a GaAs HEMT ( CGY50 from Siemens), and in the datasheet they do not indicate a value for gm at 2GHz ( and not even for any frequency) I made a PCB and plan to measure the S-parameters ( for a certain drain current) and measure a this bias point the "gm".
But is there a simple method to measure gm of the GaAs HEMT?
Also if somebody has the small-signal parameters of this device, I would really appreciate it if you could share them with me. So I can use this parameters to simulate the oscillator. I ask this because they are not mentioned in the datasheet and I even mailed to Siemens if they could look them up for me. Up to now they didn't even reply to me.........
Greetz E-goe
There are S-parameters on page 5 in the data sheet.
Hi VSWR
I have read this datasheet too but for me it's not clear which bias current they use.
I think is biased with Id=60mA, Do you agree?
I also need the small signal parameters of this device(wich I didn't find anywhere up to now).
Greetz E-goe
If the goal is only to make a Colpitts oscillator, don't waste time to simulatie, try it in the lab. The only thing you have to take care is that if fmax is large enough or not.
Besides that, you can(for real application, this is a MUST) add a buffer as a output driver to isolate the LC tank. And a high impedance high frequency probe helps a lot, because spetrum cause a 50ohm load unless you transform this impedance to a high impedance.