7 GHz upconversion mixer

i want to design a 7.5 GHz up coversion mixer with discrete schotky diodes or FET

can someone recommend me a circuit , or topology to begin with , is single FET mixer can be used

note the IF is about 2 GHz , and the carrier about 5.5 GHz


khouly

I would consider a single resistive fet mixers (see the original papers published by steve maas)....

Lance

yeah resistive mixer is good for transmitter , i was wondering to have high moderate power HEMT and use it , this can be also a choice , what do u think about it

i am using now NE4201s01 , which is a samll signal HJFET , so it can't ne used in high input power like o dbm and so

any resomendation , i am doing this in my master work

khouly

I've actually used the NE3210/4210 many times and had good luck in the 1-12 GHz range.

I've not had luck going with larger FETS and achieving better linearity... although i can't explain why... it is probably just a flaw in my implementations.

I've also been disappointed with the resistive mixers I've tried with avago enhancement mode discrete fets... again, not great implementations might be to blame.

Let us know how you make out...

Lance

did u use them in transimtter or receiver design ?

they are great in microwave receivers specially LNA and small singal mixers

khouly

I've used them in RX and TX designs. If you've got cases that are inherently good frequency plans for spurious you can do a lot with a single ended resistive fet mixer. I tried to build a singly balanced one once and failed, not sure what mistake I made but it was many years ago.

Lance

the main issue is when using the NE4201 as acctive mixer , it give a very good gain , in the samll signal
but as the IF signal is high , the gain drops alot , and the mixer becomes passive

what do u recommed ,gate or drain pumped mixer

i will try to design a ressitive one , and update u

khouly

I've just pumped the gate and built an IF/RF diplexing network at the drain (source grounded).

Lance

Added after 1 minutes:

oh, and you'll want to experiment with the gate bias (even though there is no average drain current).

Maas's papers give a lot of good info; in fact the main ones were for an x-band mixer originally so it should have lots of good relevant information.

Lance