Origin of the 50 kHz "interference" produced by ADALM2000

Dear Developers,

For the past several months we have been working on building a wireless communication system by using two ADALMs as the Tx and Rx along with other external hardware (mixers, filters, amplifiers, antennas, etc.). Although we were able to achieve some functionality such as video streaming, there is one vulnerability with the system: it detects the packet using the hardware trigger of M2K. That is, during the operation, the transmitter constantly sends non-cyclic buffers/packets and the receiver detects the packet by setting the trigger voltage level to be above the noise floor but below the signal amplitude. Then, the detected packet/buffer can be read for demodulation using getSamples or getSamplesRawInterleaved.

Although the system initially works very well when the voltage of the signal is much higher than the noise floor, we encountered a problem when the signal is small enough to have only 20-30 mV peak-to-peak voltage. In this case, the trigger level is set to somewhere between 20 and 10 mV and we initially thought it's a safe trigger level because the noise samples we received using ADALM is within 10 mV peak-to-peak range.

However, after several instances of "false alarm" triggering events, we finally figured out that there is some 50 kHz "interference" coming from the ADALM that "leak" into our analog input and the interference has a non-negligible amplitude (typically more than 10 mV ). I have attached below some images to show the samples of the "interference" that is received by ADALM, but it is the video that is attached in the following link that made us realize the "50 kHz" is actually coming from ADALM itself rather than elsewhere (link shared through goolge drive).

https://drive.google.com/open?id=1L-9mDO1Y0OjtFbk1WfHOhBLlFsVX5Wh8

(we made our own interface board so that the ADALM inputs and outputs can be accessed through the SMA connectors)

Another (actually this finally convinced us) observation that wasn't show in the video is that:

1. Connect ADALM (not powered) analog input with the scope, don't see the 50 kHz interference

2. Plug ADALM USB cable into the computer, initially there is still no 50 kHz interference

3. Wait for a few seconds, the interference appears.

Do you guys have any insights regarding what the origin of the 50 kHz pulse/wave is? It seems to us that something (not sure what it is) is "initialized" after ADALM is powered on and that thing is generating 50 kHz pulses/waves.

It is okay if we are not able to identify or get rid of it as we are planning to add more amplitfiers to our system, but we are still curious about its origin.

Thanks in advance for any insights you could offer!

Steven