Page 1 of 1

RF Output - Gibbs phenomenon ?

Posted: Thu Jun 10, 2021 9:30 am
by pmrp
Hello Red Pitaya community !

I am just beginning to play with this wonderful tool (RP 125-14), mostly with basic RF signal generation & acquisition.

My project involve sending pulse-echo signals for ultrasound NDT applications. However, while playing with the oscilloscope app to generate 1us pulses (using 1kHz PWM with 0.1% duty cycle), I noticed a very high amplitude oscillations with a central frequency around 20MHz (see the screenshot below).

The setup is very simple: the OUT1 is directly plugged to the IN1 via a BNC cable & adaptors. All other inputs/ouputs are isolated with 50ohm terminators. Swapping between inputs and outputs does not change anything in the result.

Is it the usual behavior ? I would not expect the Gibbs phenomenon to be that strong, given the bandwidth of the RP's outputs. Would that come from the impedance of the BNC cable or something else ?

Image

Thank you by advance !

Re: RF Output - Gibbs phenomenon ?

Posted: Thu Jun 17, 2021 1:02 pm
by kasaudio
Hello.

Is out1 also 50 ohm terminated?

Re: RF Output - Gibbs phenomenon ?

Posted: Thu Aug 12, 2021 10:08 am
by redpitaya
it is, and all the specs for all the boards are easily accessible, also for comparison reasons, here:
https://redpitaya.com/rtd-iframe/?ifram ... dware.html

Re: RF Output - Gibbs phenomenon ?

Posted: Tue Jan 18, 2022 6:41 pm
by Kaltobi44
Did you find a solution for that ?

Re: RF Output - Gibbs phenomenon ?

Posted: Wed Jan 19, 2022 8:34 am
by redpitaya
I would suggest connecting a load via a T connector near the input. It can be that 50 Ohm is not needed, in some cases 1KOhm is enough.