What are some alternative methods for measuring waveforms in hardware testing?

[Mothic]

As part of a masters thesis project I will be attempting to examine how well a piece of equipment transmits a particular waveform. I was planning to transmit the waveform via an antenna, then receive and record the transmitted waveform via another antenna connected to an oscilloscope, then compare the two. (e.g. Signal waveform, pulse-width, noise etc.)

Some said to me "I doubt the sensitivity, resolution and accuracy of an oscilloscope will be sufficient for the purpose. How will you know what measurement artifacts are produced by the oscilloscope? You need to think more about what measurements you need to do (and what is available to do it). There may be issues of calibration to think about as well."

! I've heard someone else suggest an oscilloscope should be fine.

Since hardware testing is beyond my regular experience... can someone help me out here?

What options does one have for measuring waveforms beyond oscilloscopes or spectrum analysers? Is there a difference? What do I need to keep in mind?

 

[vk6kro]

An analog oscilloscope doesn't add any artifacts. Radio transmitting stations routinely use oscilloscopes to monitor their transmissions.

If it has adequate bandwidth, you may be able to examine the carrier waveform, or you may be able to demodulate the information off the carrier and examine this to compare it with the intended modulation.

This will depend on the type of modulation being used.

Care must be taken when receiving the signal as it will be possible to collect a very large signal and possibly damage the equipment.

Also, overload of the receiving equipment must be avoided, possibly by attenuating it sufficiently.

If the modulation is Amplitude Modulation, then the detection must be arranged so that the diode's threshold voltage does not cause distortion.

 

[Mothic]

An analog oscilloscope doesn't add any artifacts. Radio transmitting stations routinely use oscilloscopes to monitor their transmissions.

Ah Excellent! Thanks for the informative response, i'll look into options for analogue oscilloscopes, but will they be able to record the signal for me without generating artefacts in a ADC conversion?

If I've only got access to digital oscilloscopes.. Could I measure the signal without any issues if its of a high performance? (e.g. Sampling Rate > 2 x Max Freq, Suitable bandwidth, etc).

Also when you say Radio transmitting stations do this... are they feeding the signal via cable straight into an oscilloscope? Or transmitting it antenna to antenna?
E.g. Will that cause more issues / artefacts for me?

 

[skeptic2]

What is the source of your signal, a signal generator or something else?
How much power does it put out?
What kind of antenna are you using, dipole, loop or something else?
How far apart will the antennas be?

 

[Bob S]

What is the power level of the received signal, compared to kTB? T = temperature in kelvin, B = bandwidth in Hz, and k = Boltzmann's constant, 1.38 x 10^-23 Joules/kelvin? What is the noise figure of your input amplifier? .Any additional noise is due to your equipment. Spectrum analyzers allow you to measure power levels, usually in dBm, decibels below 1 milliwatt.

Note: at 293 kelvin and 1 MHz bandwidth, kTB=1.38 x 10^-23 x 293 x 10⁶ = 4 x 10^-15 watts = 4 x 10^-12 milliwatts = -114 dBm.

 

[Mothic]

What is the source of your signal, a signal generator or something else?
How much power does it put out?
What kind of antenna are you using, dipole, loop or something else?
How far apart will the antennas be?

The source will be one of those Ettus USRP devices.
The power output will be 100 to 200 mW.
Antennas will range from a basic omni-directional antenna then later a directional one.
Distance is to be determined (e.g. part of the work is to measure / calculate this).

 

[vk6kro]

What frequency will you be using?

 

[Mothic]

What frequency will you be using?

At first something in the 1 - 250MHz region, then 2.4 GHz-2.5 GHz later.
Its still being mapped out as a project.

 

[skeptic2]

The path loss between the transmitting antenna and receive antenna depends primarily upon distance, frequency and antenna gain. Without those values it is impossible to estimate how much signal you could expect to see and whether an oscilloscope would be able to show it.

A spectrum analyzer will be much more sensitive and you still will be able see the waveform of the signal by setting the bandwidth to zero.

 

[jsgruszynski]

There are many ways of doing this.

Generally what you are talking about doing is NOT done in one-shot. The communication channel testing is separate from modulation/waveform validation. You should be breaking this into controllable pieces instead.

You can usually use an oscilloscope if it's not too primitive - for most digital modulation schemes you would use an "eye diagram" mode common to medium to high end oscilloscopes to characterize the modulation integrity. You can also do full bit-error-rate testing (BERT) once the eye shape looks OK. This is how most digital modulation systems are debugged and validated.

You can use a spectrum analyzer in some cases because the frequency domain view of simple pulse modulations are very distinct. For more complex modulation it become less useful however.

You simulate a channel with a direct cable connection between transmitter and receiving simulating propagation losses with various attenuators and/or filters. Never the real air channel until you have a near complete commercial release of HW. There are government regulatory reasons aside from the control/simplicity reasons.

Antenna and T/R system testing is generally down without modulation using a VNA in the frequency domain. This can always be applied to a modulation scheme to estimate its impact on signal integrity. Again, "dummy" loads and channels can be used, particularly using anechoic chambers.