I am using the AD8343 (frequency mixer DC - 2.5 GHz) for a 2 stages mixer. The first stage is a 640MHz + 400MHz -> 240MHz and is working well. The second stage is a 240 MHz + 242 MHz -> 2 MHz so it brings high-frequency signal to "normal" low signal, ready to be acquired with an ADC. The matching/tuning of the second stage is way more challenging than the first one because of this low signal. While I reached the +7 dB gain described in the datasheet at the output of the first stage (with a -10 dBm RF and -10 dBm LO), the output of the second stage never reached more than -20 dBm, that is to say 10 dB losses with regards to the RF signal.
I carefully followed the instructions of the datasheet for the differential input: a parallel capacitor gives a -20 dB peak reflection coefficient (S11 with the network analyser behind a balun). The output load is then adjusted with a potentiometer. Again, this procedure gave good results with the first stage.For low frequency (let say < 5 MHz), the matching of the output should -- I assume -- not necessary be 50 ohms but more a high impedance. Despite the datasheet show interesting examples for down-conversion, the lowest output frequency is still 70 MHz and I am missing advices for tuning at lower frequencies.
Thanks for your inputs !Étienne
Sorry for not getting on this sooner. I believe you have done all the right things. Having AD8343 operate down to such low IF is as challanging as you have seen. AD8342 should be a much better choice for your application, particularly as your 2nd down-converter to the ADC. AD8342 data sheet provide the example down to 1MHz IF between page 19 and 20.
The output stages of AD8342 and AD8343 are a bit different, though they’re both having open collector outputs. AD8342 output stage internally uses pull-up resistors and better matching to support broadband performance, but AD8343 does not offer these. Therefore the similar external output network may not be able to help AD8343 get down to 1MHz.
I hope these would help.
I'm using the AD8343 at 2.5GHz in an FMCW-Radar so that IF frequencies from a few kHz to several MHz are generated. The circuit from the datasheet didn't work, I simply added two 68Ohm resistors from OUTP and OUTM to VPOS in order to get it running. Otherwise only higher IF frequencies are possible ( > 50MHz). The OUTP and OUTM are connected over a capacitor to a differential op-amp, whose input resistance is around 250Ohm. Of course I don't know, if this is the optimum choice and how it changes the noise figure, gain and IIP3. Perhaps somebody could provide a hint. Unfortunately only the S-Parameters of this mixer are available, but no spice model so that I can't perform a simulation.
I recommend to add a note in the datasheet, because the mixer is declared to work from DC to 2.5GHz.
Are you sure AD8342's output has internal pull-ups????