Issue with Output Frequency in Down Converter Custom Design

It looks like you are getting excessive LO-to -IF and RF-to-IF leakage. I have a number of questions that may help to figure out what is going on.

1. Are you getting these results on your board or on the customer evaluation board?
2. Do you have a customer evaluation board on hand to try and repeat the set-up on that board.
3. I assume that VCC is connected to 3.3V and that all of the ground pins are connected to ground. Is that correct?
4. Are you able to observe or measure the supply current? It should be around 130mA. What value do you measure.
5. Are you able to disable the device and observe the current drop to <<1mA.
6. What are the power levels on the RF and LO inputs?
7. Is the LO port ac-coupled?
8. Is the RF port dc-coupled?
9. How are the IF outputs configured? Are you observing the two single ended signals (which each have an output resistance of 25 ohms) or are you observing the combined single-ended level through a 1:1 balun?

10. What power levels are you observing at the IF output at the LO freq, at the RF freq and at the IF freq?

     

    Eamon 

There are a number of issues that I see with you circuit. The specified LO power that needs to be applied to the mixer is 0 dBm. The mixer is still specified to operate at input powers down to -6 dBm but the level you are applying is far two low. My guess is that this is the main issue. I was able to find a spec sheet for some of the filters you are using but not all. I'm assuming that FL3 and FL4 are both 15 GHz band pass filters. If you are targeting to apply 0 dBm to the LO pin of the mixer, you need to apply extra input power at LO_IN_0 to make up for the insertion losses of these filters. 

At the IF outputs you have two filters side-by-side. The differential output resistance of the IF output pins is 50 ohms. That means that each IF output leg as a single ended output resistance of 25 ohms. If your filters are 50 ohm filters (I could not find a spec sheet for the), there will be a mismatch between the mixer output and the filter input, resulting in additional power loss. 

Assuming that you get the LO drive up to the required level, if the LO is 15 GHz and the RF is 14 GHz/-50 dBm, I would expect and output power at each IF port of -13 dBm (i.e. 10 dB of conversion loss in the mixer, then the power gets split evenly between the two IF outputs). 

So as a next step, please turn up the LO power and see if the circuit starts working better.

Best Regards

Eamon