We use ADI-RFSOM(picozed) hardware.
When we pass our TX signals(BPSK modulated signal with 2.4 MHz bandwidth) to the AD9361, and observing the o/p at TX ports, the o/p power level varies for different frequencies
For same settings in HW and same signals sent to the AD9361 data port, the TX power for BPSK modulated signal is
-3dBm at 70MHz
+7dBm at 800MHz
+2dBm at 2.4 GHz
Why this variation is happening.
How we can get the power levels as mentioned in the data sheet
The board was designed to work across a wide tuning range, so it is expected for the frequency response not to be flat from 70MHz to 6Ghz.
In order to get datasheet performance, narrow RF front ends…
So, many of these questions are not for us.
But more for the AD9361 design forum.
The characteristics for AD9361 are defined/supported by…
Which hardware version are you using ? https://wiki.analog.com/resources/eval/user-guides/adrv936x_rfsom/hardware
Moving this post to Linux Software Driver forum for comments on power variation across frequency for SOM module.
It could be because of matching.
Thank you for the reply
The HW we use is of type ADRV9361-Z7035
The actual part number on the board(picozed HW) is
Picozed part number: AES-PZSDRCC-FMC-G with
SOM carrying the part number "BD-Z7PZ-SDR2-G"
Is this the expected variation in power level for the hardware we purchased.
In order to get datasheet performance, narrow RF front ends need to be designed and tuned to the desired frequency. (FMCOMMS2 is an example for 2.4 GHz)
On revision F of the board the RF performance was improved by changing the PCB material to Tachyon100G, but there still is a ~15dB variation across the whole tuning range. Below is the response of a RX-TX loopback test
We are making SDR system based on ADI-RF-SOM.
We have planned this module as field deployable module.
Currently we use AD's AD-PZSDR2400TDD-EBZ as RF front end for 2.4 GHZ band.
Later we are going to make our RF front end with few frequency bands
When the power level variations are more, it will be challenging to design the RF front end.
In the user guide of picozed, the TX power is mentioned as +6.5dbm to 8 dbm, these specifications mislead us for the TX power, similarly may be the case for the RX side and also for each channels of TX and RX
Whether the reference measurements are available for individual channels of TX and RX paths, so that we can cross verify our measurements. We have to do our data link verification