frequency multiplication, please help

hi

i want to build a low power transmitter and receiver which can send and receive data at 1.6ghz
now i have two idea going around in my mind
1. to design a pll based frequency synthesizer for 100 mhz and then using a frequency multiplier to multiply the frequency 16 times to get 1600 mhz final frequency

2. to directly design a pll based frequency synthesizer for 1600 mhz

please suggest me which is the best method to use or if you know any other method please let me know.

the main problem i am facing is inductors, when ever i think about it i feel like pulling my hair, how can i make inductor for nano henries or from where i can get femto farad capacitors?

also after some searching i found this article:h**p://petlibrary.tripod.com/rfmod.htm
can someone tell me which components are responsible for frequency multiplication there (other than two Schottky diodes, D1 and D2)?

i would really appreciate if someone can answer my question, i have already searched and read many articles for this but none helped me.

thanks

1.6GHz is not particularly difficult and you certainly don't need femto farrad capacitors!

You have not told the modulation technique you want to use, the schematic is AM but that probably isn't best for microwave data.

Considerations:
1. is you multiply the frequency, the whole signal is multiplied, including sidebands. If your original signal was 100KHz wide at 100MHz, when multiplied it will be at least 1.6MHz wide and probably much more. Also consider that frequency drift will be 16 times worse!
2. a PLL at 1.6GHz is relatively easy and you can directly modulate it so what you get out will be virtually identical to what goes in.

To add to your options, you can use a combination of both quite easily. Create a stable oscillator using a PLL at say 1.3GHz and mix it with a modulated oscillator at 300MHz. The two products will be at 1GHz and 1.6GHz. far enough apart that it's easy to filter out the 1GHz signal and keep the 1.6GHz one. At 300MHz the modulator should be easy to build using normal pinned components.

Brian.

thanks Brian for your response.
i am going for option two. but to get started i need to design a tank circuit for 1.6Ghz first. using the formula 1/(2 x pie x sq(LC)) i calculated that i need 1.5pF capacitor and 6.6nH inductor. now my question is how can i design 6.6nH inductor? I have 28awg wire (the wire from bus cable which is used to connect harddrive to motherboard) in hand. Do i need any other kind wire for it?
your response is solicited
thanks

Forget using that wire !

You need at least a solid copper wire and ideally one that is silver plated. At 1.6GHz the concept of using wire for coils starts to fail, simply because the wire start to become too short to physically and accurately mount. The usual method of making inductors is to use strips of copper on a PCB. The capacitor can be made from insulated solid copper wire too, just twist the wires together. You can then adjust the value by altering the tightness and length of the twists. It isn't very scientific but it works well.

Brian.