Connecting IQ signals from signal generator to TRF370417EVM

Paul:
Typically, the arb waveform generator will have an option for I/Q differential output so that all of the connections can be made appropriately. In your case you are using a arb function generator that is not equipped with that option. As a quick test, you can generate quadrature signals with the function generator and inject on just the I+ and Q+ connections. Note, the TRF370417 requires a 1.7V common mode on all four inputs to operate properly. For the two connections to the instruments, you will need to add this offset to the signal. On the unused ports, you will need to add this voltage externally. I have done this previously by “kludging” in a 100 ohm to ground and a 100 ohm pull-up resistor to 3.4V external. This will provide a similar 50 ohm termination to the BB port and provide the 1.7V common mode voltage. When only using the one of the two input ports you will take an automatic 6 dB insertion loss hit which will impact your gain and IP3 performance.
A better alternative is to use a true differential quadrature source. If you do not have access to the proper test equipment, consider utilizing one of TI’s DAC to achieve the same thing. The TSW3084, for example, is a full featured DAC which includes the DAC3484 with the TRF3705. The TRF3705 is a newer modulator that has a 0.25V common mode to match up to that DAC. If the frequency capabilities of the TRF370417 are required, then a regular DAC EVM (e.g. DAC3482) can be configured for differential output. The common mode voltage will still need to be addressed properly.
--RJH

Thanks for your response RJ.

Not wanting to take the 6dB insertion loss, and not having a TSW3084 available, I ended up ordering a couple AD8476-EVALZ boards to convert my single ended I and Q signals into differential I+/I- and Q+/Q- signals. Each of these boards were ~$32 from Digikey and came populated with SMA connectors, which was quicker than making up a small board for the AD8476 chip myself. Adding a 1.7V offset to the single-to-differential converter board seems to have allowed me to meet the input requirements of the TRF370417 for I and Q signals.

My only lingering concern is that the spec sheet lists the output impedance of the AD8476 is only ~0.1ohm, whereas the TRF370417 specifies that it should be 50ohm. As far as I can tell this is not having a negative impact on my test setup, but I fully admit that there may be effects I am not experienced enough to recognize.