About conversion loss of the 4th harmonic mixer

[jocko_lee]

Hi! everybody

I'm simulating a 4th harmonic mixer with an anti-parallel diode pair, LO=8.5GHz;RF=34-36GHz;IF=DC-2GHz.

When i simulate it as down-converter,the conversion loss is about 8 to 10dB,but when use it as up-converter,the conversion loss changes to 2 to 4dB.

but i think the conversion loss should be the same whether use it as up-converter or down-converter.I was puzzled.Is my rusult natural? or there is some thing wrong with my design or simulating method.

Thank u for helping me!

 

[jiripolivka]

I do not know about simulations of this kind. Harmonic mixers are pumped by a lower LO power; in the non-linear elements of that mixer, harmonics are generated, and only one of them takes part in mixing function.
Typically, a good mixer at the "first" harmonic LO has 6-8 dB conversion loss. At second harmonic, twice as much, 12-15 dB. At third harmonic, you can expect 18-20 dB. At fourth, 23-26 dB conversion loss.
The condition to make a good harmonic mixer is to pump it with a maximum subharmonic LO input power, to generate the desired harmonics.
Your data show that your simulation is simply wrong. My numbers are real, measured on good harmonic mixers.

 

[tony_lth]

You can ref Hittite website, the conversion loss of most of its 2nd Harmonic mixer is 10dB.

 

[jiripolivka]

I like your query: you try to simulate a harmonic mixer at 4th LO harmonic. I replied that to my knowledge, your simulation is wrong.
Then you counter what you find at Hittite- again, they make harmonic mixers at 2x LO frequency. OK, they achieve CL of 10 dB- but guess how?
They integrate a LO amplifier in mixer chip, to optimize 2nd harmonic generation.

I do not see how it solves your problem with 4th harmonic?
I know real mixers, and my data are taken on real harmonic mixers.

You can buy from Hittite but no harmonic mixer they have runs on 4th harmonic. Your simulation is plainly wrong.

 

[cnm]

I like your query: you try to simulate a harmonic mixer at 4th LO harmonic. I replied that to my knowledge, your simulation is wrong.
Then you counter what you find at Hittite- again, they make harmonic mixers at 2x LO frequency. OK, they achieve CL of 10 dB- but guess how?
They integrate a LO amplifier in mixer chip, to optimize 2nd harmonic generation.

I do not see how it solves your problem with 4th harmonic?
I know real mixers, and my data are taken on real harmonic mixers.

You can buy from Hittite but no harmonic mixer they have runs on 4th harmonic. Your simulation is plainly wrong.

Two different guys, man.

I don't see what is wrong with having an LO driver amp. A very common practice.

 

[jiripolivka]

Yes, using a LO driver amplifier is a good idea; in harmonic mixers, LO pumping level should be as high as possible, and its adjustment also can optimize the mixer to operate best at a selected harmonic.
I do not see any good approach in modeling as you have not adequate data. No diode manufacturer for instance offers data on a particular diode operating at some harmonic frequency. The best approach is to try a real mixer and adjust optimum conditions.
Harmonic mixers are almost never used in applications where conversion loss is required to be low. They are quite lossy and they are used in applications where it does not matter like with counters to operate at harmonic multiples, with spectrum analyzers to see a response at a higher frequency than the analyzer can, etc.
It is also difficult to calibrate such mixers as the LO/x signal source sees an unmatched load when used with a harmonic mixer.