This blog post covers my experiences using the MSOX3034T to measure low power radiated RF signals.  The challenge this time is to have the Keysight 'scope help me learn about RF signals broadcast by a simple handheld RF remote control.  In this case, a remote control for a consumer ceiling fan.  I have previously studied this remote using a Tektronix MDO4104-3 with a built in spectrum analyzer.  So, though I know a fair bit about the RF signal generated by this remote, I'm using the MSOX3034T to see what it can tell me. I'm using this particular remote because I know its carrier frequency (304 MHz) is within the 350 MHz bandwidth of the analog channels on the 'scope.

 

My measurement set up

I connected a Diamond RH799 70 MHz - 1 GHz adjustable antenna to the channel 4 analog input.  Knowing that the remote I am studying has a carrier frequency of 304 MHz, I adjusted the antenna length to 24 cm, based on guidance from the chart printed on the antenna. I also changed the input impedance on channel 4 to 50 Ohms. I'm expecting a fairly low amplitude signal, so vertical sensitivity was set initially to 10 mV/div.  The horizontal time base was set initially to 5 ms/div, then adjusted to 2 ms/div once signal reception was confirmed.  The antenna connection is shown below.

 

Diamond RH799 connected to analog channel 4

 

The MSOX3034T captured the RF transmission from the remote control quite nicely.  I like the rich intensity-based amplitude detail visible in the screen snap shot below.  It reminds me of old CRT displays  Using cursors I was able to measure the burst duration at 12.2 ms.

OOK frame from remote captured by MSOX3034T

It can also be surmised that the modulation scheme used by the remote is On Off Keying (OOK) based on the bursts of carrier separated by dead time.  It looks like there are 13 bits transmitted in each code frame.  One more important detail to discover is the frequency of the carrier. Perhaps we can zoom in on a carrier burst and get a frequency measurement.  A quick check of the sample rate at this time base setting gives me pause.  Even though the analog front end on this 'scope has a 350 MHz bandwidth, I don't expect to get accurate frequency readings of a roughly 300 MHz signal that has been sampled at 200 MSa/s.  Let's take a look anyway just to confirm that.

OOK frame zoomed to 20 ns with 200 MSa per second sampling

Suspicion confirmed.  The carrier is under-sampled at this time base setting.  Notice the automatic measurements for frequency and counter are both incorrect.  One possible remedy is to decrease the main time base setting which will in turn increase the sample rate.  I will be looking for a sample rate greater than about 600 MSa/s to correctly acquire a carrier signal around 300 MHz.  I went to the limit with a time base setting of 2.0 ns/div resulting in a 5 GSa/s sample rate.  This produced the very clean carrier signal shown below.

RF carrier sampled at 5 GSa/s

The automatic frequency measurement looks fine at 303.87 MHz, but the counter measurement at 53.899 MHz doesn't make sense.  Perhaps the amplitude is too low to drive the counter?

 

How does the Tektronix MDO4104-3 handle this measurement challenge?

The Keysight MSOX3034T provided all the detail I was looking for about the structure of the RF remote control signal.  I was able to measure frame duration, determine modulation scheme, determine the number of bits per frame, and measure the carrier frequency.  All of these measurements took about 3 minutes to complete.  The same information was extracted from the MDO4104-3 in about 2 minutes from a single screen.  See the image below.

MDO4104-3 RF burst analysis

Here is one way to assess the two oscilloscopes in this measurement case.  Both instruments provided the measurement insights I was looking for with about the same effort.  The Keysight 'scope costs about $9000 less than the Tektronix 'scope.

Is the built in spectrum analyzer on the Tek 'scope worth it in this scenario?  Probably not.  If the same outcomes can be achieved at a much lower cost then the fancier features can't be justified.  That is true in this use case, but there are many other use cases where a built-in spectrum analyzer is very helpful.

 

Best regards,

 

Mark A.