Previous blog posts of my MSOX3034T Road Test:
Keysight InfiniiVision MSOX3034T RoadTest | Unboxing & First Impressions
Keysight InfiniiVision MSOX3034T RoadTest | Analog Specs & Basic Functionality
Keysight InfiniiVision MSOX3034T RoadTest | Low Speed Analog Experiments
Keysight InfiniiVision MSOX3034T RoadTest | Digital Channels & Serial Decoding
Keysight InfiniiVision MSOX3034T RoadTest | Bandwidth & Update Rate Measurements
Hello Everyone,
In this blog post I will show you some quick experiments I did to test how well the Keysight InfiniiVision MSOX3034T mixed signal oscilloscope handles high speed and RF related signals.
1. MIPI CSI 2 Camera Interface
First I wanted to see if the scope can be used to debug MIPI CSI 2 camera interface.
The MIPI CSI stands for Mobile Industry Processor Interface Camera Serial Interface, and an interface used between camera modules and SoC.
It used in mobile phones, gadgets and devices like the Raspberry Pi:
I also used it in my Stereo Vision and LiDAR Powered Donkey Car project, with hardware interface implemented in a Xilinx FPGA.
On the hardware level, the so called D-PHY interface (v.1) is used, which support data rates up to 1Gbps / lane. In the case of Raspberry Pi, one clock and two differential data lanes are used:
I was interested to see if the MSOX3034 is capable to show us the signal on one of the data lanes.
To probe the signal, I just used the exposed part of the connections of the ribbon cable. For ground I used the metal casing of the HDMI connector.
(Initially, I though to build some kind of an adapter for easier probing, but unfortunately I did not really had the time and parts for that.)
The camera used is a Raspi Cam v.1 with 5 Mp resolution. The software running was Motion, which streamed video from the camera.
This is what I got when trying to capture the stream:
The image frames are sent using burst of high speed differential signalling. The voltage level is between 264mV to 340mV, so it is a quite narrow band.
When we zoom it a little bit, first we can see a single packet, than the individual bits of an image frame:
(note: I was too lazy to check in what format, RAW or compressed, the image is sent 
)
The signalling frequency looks to be 400 MHz, and the scope is clearly capable to show the signal:
2. High Speed USB
Next, I wanted to see if the 350 Mhz bandwidth (~470Mhz measured) of the scope is enough for capturing USB 2.0 Hi-Speed 480 Mbps signaling.
This is what one of the captures Low Speed Analog Experiments blog post looked like:
The above capture show a packet of a file transfer from an USB stick.
This time, I wanted to take a look one something else. So, I ended up using a Logitech C920 webcam, the video being streamed using Cheese.
To capture the signals I used the same adapter I fabricated last time:
This is what I got:
The signals from the two differential data lines looks reasonably good. It looks like enough information is retained to be able to decode the signal.
The output of the math function looks a little bit noisy. At this speed, I would not necessary rely on it to analyze the signal.
3. NFC Antenna Tuning (continued)
In the Low Speed Analog Experiments blog post I attempted to measure the inductance of an NFC antenna, and based on that I calculated the required tuning capacitors in order to tune the antenna to 13.56 MHz.
The resulting inductance and capacitance values were 1.6 μH and 168 pF.
To continue this experiment, I though it would be a good idea to use the Frequency Response Analysis functionality of the scope to measure the actual resonance frequency of the scope.
So, I ended up designing the bellow circuit, and also done a simulation in LT Spice. As expected with the old tuning capacitors of 133 pF, the resulting resonance frequency should be around 15.5 Mhz.
(68pF = 137pF / 2)
Next, I built this circuit and did a Frequency Response Analysis measurement using the scoe:
To my surprise the resonance frequency of was around 12.5 MHz, instead of the expected 15.5 MHz:
After a bit of wondering what is happening, I vaguely remembered the probes has some capacitance. According to the datasheet it should be around 11pF.
So, I added the probe capacitance-s to our initial model, and run an another simulation in LT Spice.
The resulting value should be around 13.5 MHz. So, maybe we have more additional capacitance somewhere...
At this point I got a little bit lazy, and I though to just try capacitance values to see if the NFC reads gets any better.
The initial tuning capacitors were 133pF, and I tried 147pF and 168pF. Both, the 147pF and 168pF works better than the initial 133pF. The reading are now consistent, and there are definitely fewer NFC reads happening.
4. Looking at the output of the HackRF One
Finally, I wanted to take the look on a RF signal generated by the HackRF One SDR. The HackRF One is a Software Defined Radio device capable receiving and transmitting signals (in half duplex mode) with the frequencies from 1 MHz to 6 GHz,
As I just got my HackRF One since about two weeks ago and did not have the chance to experiment too much with it, I though try out something readily available from the internet.
After some searching I found this GNU Radio Companion project that implements Wide Band FM Transmission of a WAV sound file:
The output is frequency modulated signal with 152 MHz carrier frequency.
The output of the HackRF One was connected to the Channel 4 of the scope, using a BNC cable and an BNC to SMA converter. The input termination of the scope was set to 50 Ohm.
This is what I got, when I run the example:
(Note: we probably shouldn't run this on the HackRF One with an antenna connected without some kind radio license)
In the next blog post I will take a quick look on the Function Generator, Voltmeter, Frequency Counter and other miscellaneous features of the scope. Then, I will finish the Road Test with a summary on what we did, comparison with competitors and finally some conclusions.
Stay tuned!
Top Comments