How to improve the efficiency of frequency multiplier with a single diode?

Hello,

I am trying to make a frequency multiplier with a single diode. Could you please kind enough to help me. The thing is I am inteersted in getting 10.5 GHz from 5.25 GHz Rf power. I am using a schottky diode, which has a Conversion Loss of 14 dB. And when I attach it to the matching networks on either ends, the loss increases while the return losses no doubt improve.

Is there any trade off between return losses and conversion loss. How can I improve my multiplier, that is work efficiently? Is theer any mathematical relationship to let me know how much power levels one can achieve at particular diode?

Thank you

Re: Harmonic Mixer

ok strange subject name for a multiplier

a step recovery diode works well for multipliers in the uWave bands
I use one in a project to multiply from 1296MHz to 10.368GHz (x9)

another way is to use a MMIC like a NLB-310. Overdriving it at 5.xx GHz will produce harmonics up the band, just filter off the one you want.

cheers
Dave

Re: Harmonic Mixer

Thanks for reply. But why not a varactor diode? I have read it gives more output power, this is what I require, Maximum output power at 2F. Do you have any design examples? Like how do I bias the diodes etc? and any reference which can help me in selecting the diode.

Thanks again

Re: Harmonic Mixer

MobiNaz said:

Thanks for reply. But why not a varactor diode? I have read it gives more output power, this is what I require, Maximum output power at 2F. Do you have any design examples? Like how do I bias the diodes etc? and any reference which can help me in selecting the diode.

Thanks again

Click to expand...

Yes you could use a varactor diode, I have used them with great success on lower freqs eg. tripling from 432 to 1296 MHz works well ~ 50% efficiency. The diode I use is a power varactor BAY96, pumping 15W in at 432 gives ~ 7-8 W at 1296MHz. It wont work at 10GHz

Offhand I dont know what varactor diodes are available (if any) for 10GHz operations you will have to do some searching for that.
You also didnt mention what sort of power levels you want at 10GHz .... milliwatts, watts??
None of the diode multipliers will be overly efficient and probably less than 50% and decreasing as the freq increases.

cheers
Dave
VK2TDN

Re: Harmonic Mixer

Thanks for your reply. Well, I'll be pumping the diode with around 10 to -70 dBm. Is it possible to get an efficiency of around 50% at 10.5 GHz and upto 35GHz?

Re: Harmonic Mixer

Dear MobiNaz:

If you use a good Schottky diode as a frequency doubler, the efficiency will be poor. A better idea is to use a pair of such diodes; they act as switches controlled by the input signal. You can build even a wideband multiplier but its efficiency is usually ~ 13 dB. You will still need a post-amplifier to obtain a useful signal for e.g. a mixer.
Typically, a two-diode doubler needs an input > 15 dBm to operate well. Any lower input will strongly diminish its efficiency.

A subharmonically- pumped mixer can work as a harmonic mixer, but its conversion loss is usually >10 dB higher than a "straight" mixer, and >10 dB higher for each next harmonic.
There are varactor frequency multipliers; they utilize "parametric" action and are better in efficiency, but very demanding in design. They need THREE tuned circuits (or more) named "idler (s)" and they usually work best at a single frequency. A good frequency doubler with a varactor diode, from 5 to 10 GHz can have a conversion loss of <6 dB, but typically needs > 100 mW input power to generate ~30 mW at twice the input frequency.