High Speed Digitizer Automatic Gain Control

[maxwellb]

Hi,

I'm working on an application using a PXI 5112 digitizer to measure an RF signal with a highly dynamic amplitude. I've been playing with an automatic gain control program to help deal with this. I acquire 7000-sample bursts of 100 MHz data and dump them to host computer memory. I can acquire these bursts at about 200 Hz if I don't mess with the gain. Each time I do adjust the gain, there is a ~.05 second delay. After each burst, before the next burst is acquired, the average peak-to-peak amplitude of the burst is measured. If this amplitude is greater than 80% or less than 20% of the current vertical range, the vertical range is adjusted.

I generally have it working, but one thing I've noticed is that, in certain amplitude ranges, my algorithm results in rapid vertical range changes, at up to about 20 Hz or so, which is about as fast as it is possible to change the vertical range. There must be some kind of mechanical switching happening within the board, because I start to hear rapid audible clicks coming from the device. Will it damage my board to make these switches flip at that rate?

Does anyone else have experience implementing something like this?

Thanks,

Max

[Jim_S]

Hello Max,

After looking over your post, the short answer is yes this will wear out the board eventually.

When looking over the NI Scope Help Documentation, the 5112 has Electromechanical Relays that change when the vertical parameters are changes. Some of the other reason for clicking are resetting the device and self-calibration. The electromechanical relays have a life span of 1-2 million operations so I would avoid if possible frequent changes to the vertical parameters. Changing the vertical range across the 0.5 V and 5 V levels is the most common cause of relay changes so I would avoid this as well.