Hello, I am testing the AD9680-1000 eval board and need to compute the absolute power in dBm into the ADC. I have a measurement of power at the input to the board @ the RF input SMA connector. I assume there is some insertion loss and a matching network on the front end that attenuates the input before RF is provided to the ADC. Can anyone help translate power into the board to power into the ADC?

Hi Siyi,

Please refer to Figure 66 of the Datasheet RevA. It is shown below for your convenience.

The figure shows the front end network used in the evaluation board AD9680-1000EBZ. You will notice that for the AD9680-1000EBZ, the balun used on the board is the Marki Microwave BAL-0006SMG. If you inspect the datasheet for this balun, you will notice that there is ~6dB insertion loss along the balun. This is by design. The BAL-0006SMG is a transmission line balun and is sensitive to matching. Hence we had to use a matching network comprised of resistors R1 and R2. In the case of the AD9680-1000EBZ, you will notice that these are 25ohms a piece. This matching network is needed to get the desired wideband match (which provides the flatness in the band) as well as AC performance (SNR/SFDR). Unfortunately the matching network has an insertion loss of ~6dB across it. Then there is the series resistance R3 which also attenuates the signal a bit. This is why you have to output a big amplitude (~16-17dBm) to get a signal -1dBFS at the ADC input.

If the ADC output is showing -1dBFS in the FFT, it means that the ADC is sampling a voltage that is 1.5151Vpp differential (based on a 1.7Vpp differential input). This means that the single ended signal swing is 0.7575Vpp, which translates to 268mVrms (you can use any Vpp to RMS calculator). There are many lookup tables that show you how to translate Vrms to dBm. From my source (http://www.minicircuits.com/pages/pdfs/dg03-110.pdf), this calculates to 1.5729dBm into a 50ohm load.

Hope this helps,

Umesh