AnsweredAssumed Answered

AD9371 Trigger "Jitter"

Question asked by andrew.feldman on Jan 12, 2017
Latest reply on Jan 16, 2017 by sripad



I am looking for help in correcting a Mykonos loop trigger performance issue. 




I have built an application that uses MATLAB and AdiCmdServerClient.dll to talk to Mykonos. In MATLAB, I use the command




to set up loop trigger. My understanding (correct me if I am wrong) is that the above command sets up the ObsRx's and the Rx's to be loop triggered off of the Tx's, which I always set to IMMEDIATE triggering.


The Problem:


It is important for my application that the delay from Tx to ObsRx be constant, which is why I am using loop trigger. When RXTRIGGER.TX_LOOP and TXTRIGGER.IMMEDIATE are used, I expect the delay from Tx to ObsRx to fixed, but this is not always case. As a test, I connect Tx1 to ObsRx1 using an SMA cable and 7dB of external attenuation, then I program Tx1 to have 3dB of internal attenuation and ObsRx1 to have a gain setting of 255. The sampling rate of Tx1 and ObsRx1 is 307.2MSPS. Using MATLAB I command the Mykonos Tx1 to transmit a waveform, then measure the delay at ObsRx1. When only Tx1 is enabled and loop trigger is set up, the delay measured at the ObsRx1 does not change over the course of multiple measurements (which is what I want). But when Tx1 and Tx2 are enabled and loop trigger is set up, the delay from Tx1 to ObsRx1 is not the same for each measurement. My application requires Tx1 and Tx2 to be enabled, so this is an issue.


Furthermore, it is desirable (though less important) for Rx1 to be loop triggered off of Tx1 as well. I have found that, regardless of how many Tx's are enabled, the Tx1-to-Rx1 delay that I observe when I connect Tx1 to Rx1 is never consistent even when I enable loop trigger.


Any help with these issues is appreciated.