Post Go back to editing

Strange spikes and signal dropout when using the ADALM2000 Network Analyzer

Category: Hardware
Product Number: adalm2000

Hello, I'm using Scopy to test an ADALM2000 with the connected BNC adapter board. I'm attempting to run a network analysis of a simple low-pass filter (values 17.7 kOhm, 4.7 nF, pole at 1.9 kHz), but I'm seeing some unusual behavior:

The magnitude and phase lines behave like I'd expect for a 1.9 kHz pole, but the plots are full of these "spikes" in the measurements. I used the buffer viewer to examine individual points and discovered that the spikes are actually measurements when the test signal disappears entirely, so the ADALM2000 is effectively measuring noise divided by noise, producing a random result. 

Buffer viewer of a good point:

Buffer viewer of the very next point, on a spike:

I've tried the following steps to debug:

  • Swap out cables
  • Remove and re-seat all connections
  • Use different scope probes
  • Increase signal level to 1 V (from 100 mV)
  • Add averaging and/or more periods of sample time
  • Change the number of points per decade
  • Change the start/stop frequency
  • Replace my R and C

I have no other electrical connections, the only components in the circuit are the R, the C, the probes, and the BNC->Clip lead connector. I used a third party scope to watch the waveform at the input and I noticed steps with a missing test signal corresponding to the timing of the spikes, as shown below. It's clear the ADALM2000 is still connected to the device under test, as the same waveform is present at the beginning/end of each step (the step up to 175 mV, snapping negative, and then settling to 0 V before sending the test signal), but the test signal itself is occasionally not sent. 

I'm running v1.4.1 of Scopy and v0.32 of ADALM2000 firmware, connecting to the device over USB. Any thoughts on what's going on, or what I should try?

Thank you.