Using Tektronix Oscilloscope to Measure Phase

[quantumfoam]

Hi guys,

I am working on a project that involves measuring the phase angle between two sinusoidal voltages. The circuit under examination is an RL type (the resistor was in series with the inductor), where the phase of voltage across the inductor is measured relative to the voltage input of the circuit (it has zero phase).

The oscilloscope I used to measure such was the Tektronix MDO3014 Mixed Domain Oscilloscope. The two probes in use were the Tektronix TPP0256 type. The probe connected to Channel 1 measured the voltage signal across the inductor while the probe connected to Channel 2 measured the voltage input signal of the circuit.

After obtaining the phase between the two signals, I realized that the measured phase was much smaller than expected. While using a resistor of 32.2 Ohms, an inductor of 10 mH, and an input voltage of 100 Hz, the oscilloscope gave a phase between 1-8°. After calculating the phase angle of the voltage across the inductor as shown in the attached image below, the phase angle was clearly off from that measured by the oscilloscope (I calculated it was supposed to be 79°.

Am I missing something in the set-up of the oscilloscope to measure phase between two voltage signals? I made sure to calibrate the probe on Channel 1, and I attempted to do so with the probe in Channel 2 as well. Since this discrepancy occurred for frequencies (100 Hz and higher) that were well within the measurement range of the oscilloscope, I am wondering if it has anything to do with a set-up I missed in executing.

I noticed that the measured frequencies for the input voltage and inductor voltage were the same, implying the frequencies weren't distorted. I measured the phase from the Channel 1 signal (inductor voltage) to that of the Channel 2 signal (input voltage). I also used both an Agilent Arbitrary Function generator and the Arbitrary Function Generator from the oscilloscope. When applying the two different function generators separately, both gave me the same wrong angle (1-8°).

 

[Tom.G]

Check your arithmetic in the last step. What number do you get as the divisor?
Check that the 'scope is set to trigger on one channel, not Alternate or Chop... or better yet, use External Trigger on the supply voltage if you can.

 

[jim hardy]

The oscilloscope I used to measure such was the Tektronix MDO3014

wow i come from the days of vacuum tube analog 'scopes. Grew up on a Tektronix 545.

not Alternate or Chop..

(Assuming these mean the same today as fifty years ago)
Alternate and Chop set the vertical display mode not the trigger.

Yes, TRIG must be from one channel or the other
but at his low frequency he needs display set for CHOP not ALT.
@Tom.G i fear we're becoming dinosaurs. Sadly there's no mention of those terms on that preposterous oscilloscope's front panel..
https://www.tek.com/datasheet/mixed-domain-oscilloscopes

125 trigger settings, not 3 ?
He'll have to delve into the menus and figure how to make that beast behave like a plain oscilloscope..
Or find a more basic one better suited to the simple task at hand.

"Gee whiz guys, i need a TACK hammer not a jack-hammer !"

old jim

 

[jim hardy]

I'd start about here in the instruction book
https://www.tek.com/oscilloscope/mdo3000-mixed-domain-oscilloscope-manual/mdo3000-series-3#

 

[quantumfoam]

Check your arithmetic in the last step. What number do you get as the divisor?
Check that the 'scope is set to trigger on one channel, not Alternate or Chop... or better yet, use External Trigger on the supply voltage if you can.

I am assuming you mean the divisor that is in the inverse tangent function. The divisor is 2π. Thank you for responding and your feedback!

 

[quantumfoam]

wow i come from the days of vacuum tube analog 'scopes. Grew up on a Tektronix 545.
View attachment 215203(Assuming these mean the same today as fifty years ago)
Alternate and Chop set the vertical display mode not the trigger.

Yes, TRIG must be from one channel or the other
but at his low frequency he needs display set for CHOP not ALT.
@Tom.G i fear we're becoming dinosaurs. Sadly there's no mention of those terms on that preposterous oscilloscope's front panel..
https://www.tek.com/datasheet/mixed-domain-oscilloscopes
View attachment 215204

View attachment 215200
125 trigger settings, not 3 ?
He'll have to delve into the menus and figure how to make that beast behave like a plain oscilloscope..
Or find a more basic one better suited to the simple task at hand.

"Gee whiz guys, i need a TACK hammer not a jack-hammer !"
View attachment 215205

old jim

The oscilloscope is not mine, that's for sure. I could never own one of these without buying a relatively okay used-car first! I have read through the triggering options that the oscilloscope has. Unfortunately, I didn't have much knowledge on the subject of triggering. I did look up any useful videos to help me understand. So by properly triggering my input signal voltage (Channel 2), the oscilloscope will place any other waveforms (such as that in Channel 1:that voltage across the inductor) relative to the triggered voltage signal (Channel 2)? Also, although I put a low frequency in my example shown above, ultimately I want to probe the circuit's response at high frequencies (up to 20 MHz) Thank you for responding and for your feedback!

 

[jim hardy]

So by properly triggering my input signal voltage (Channel 2), the oscilloscope will place any other waveforms (such as that in Channel 1:that voltage across the inductor) relative to the triggered voltage signal (Channel 2)?

That's how an analog scope from the 60's would have done it. I cannot assure you there's not some esoteric reason why that computer-system-disguised-as-an-oscilloscope does something different.
Tektronix has a legacy of intuitive panel controls so i think what you said will in all probability turn out to be so .
But check it a frequency where you know what to expect for reassurance. Do you get 45 degrees around 512 hz?

In my day for checking phase difference i used "Chop" mode whenever possible because i knew it was drawing both traces simultaneously during each sweep , so there was no possibility of a trigger error.

As you learn that scope you'll develop your own methods. It is a truly astounding instrument.
Being a slide rule sort of guy I'd have to go back to school and take a minor in DSP to really appreciate its capability .Thanks for getting back to us - let us know how your lab goes maybe post a picture ?

old jim

 

[quantumfoam]

That's how an analog scope from the 60's would have done it. I cannot assure you there's not some esoteric reason why that computer-system-disguised-as-an-oscilloscope does something different.
Tektronix has a legacy of intuitive panel controls so i think what you said will in all probability turn out to be so .
But check it a frequency where you know what to expect for reassurance. Do you get 45 degrees around 512 hz?

In my day for checking phase difference i used "Chop" mode whenever possible because i knew it was drawing both traces simultaneously during each sweep , so there was no possibility of a trigger error.

As you learn that scope you'll develop your own methods. It is a truly astounding instrument.
Being a slide rule sort of guy I'd have to go back to school and take a minor in DSP to really appreciate its capability .Thanks for getting back to us - let us know how your lab goes maybe post a picture ?

old jim

I will definitely keep you guys updated if I find anything that helps solve this. I'm going to the lab today to investigate whether what I said works. Thank you very much for your help and understanding of the situation. I really appreciate it.

 

[quantumfoam]

Update:
So I went ahead and tried the above. Unfortunately, still wasn't executing the right phase angle. However, I was lucky enough to stumble into one of the new EE professors on campus yesterday. He came over and helped me in obtaining a phase angle. Strangely, as my professor explained, using a single inductor (10mH) and resistor (32 Ω) wasn't enough to measure the phase angle. I didn't quite understand his reasoning to be honest, but I did observe what he did differently. Instead of using just one inductor, he placed two (totaling to 20.7 mH). He also changed the resistor out and replaced it with two resistors (one was 20.83 kΩ, the other was 495.3 kΩ). I should also note that he used a peak-to-peak voltage of 1.2 V. By changing the frequency to 100kHz, the oscilloscope measured a voltage phase shift of 58.4° across the inductor, just 0.3 % off from the value calculated! Here are the photos of the circuit ( I changed out the position/values of the resistor but the phase shift still occurred. I even changed the inductors out for others with value around the same value) and the oscilloscope. As soon as I tried replacing the inductors with an air core inductor of my own creation (I measured it to be 6.4μH) while also switching out resistors to large (1kΩ-1MΩ) or small (22 Ω), the effect was that of yet another highly inaccurate phase angle (one of the trials, I used a resistor of 47 Ω, air core inductor of 6.4 μH, and a frequency of 1MHz, I should have measured a phase shift across the inductor of about 40-45°...I only measured 5.5°). Let me know what you guys think about this.

 

[Tom.G]

I think you need to show photos of the 'scope that includes the waveform, frequency, and trigger conditions, also state if the probes are x1 or x10; along with a photo of the functioning circuit taken at the same time as the 'scope photos with all the connection, probes, clip leads shown. Also include a schematic of the circuit showing where the 'scope and signal generator are connected.

The photo 'circuit1.jpg' shows one scope probe appearently connect to an otherwise unconnected inductor lead. I see no path between the resistors and the signal generator ground. Is it hidden somewhere?

using a single inductor (10mH) and resistor (32 Ω) wasn't enough to measure the phase angle.
Hmm, If you had the right frequency, I don't see why not, X_L at 500Hz is 31.4 Ohms.

 

[quantumfoam]

I think you need to show photos of the 'scope that includes the waveform, frequency, and trigger conditions, also state if the probes are x1 or x10; along with a photo of the functioning circuit taken at the same time as the 'scope photos with all the connection, probes, clip leads shown. Also include a schematic of the circuit showing where the 'scope and signal generator are connected.

The photo 'circuit1.jpg' shows one scope probe appearently connect to an otherwise unconnected inductor lead. I see no path between the resistors and the signal generator ground. Is it hidden somewhere?

using a single inductor (10mH) and resistor (32 Ω) wasn't enough to measure the phase angle.
Hmm, If you had the right frequency, I don't see why not, X_L at 500Hz is 31.4 Ohms.

Hi Tom. G,
I apologize for not getting to you sooner. I have not had much time to work outside of school or get in the lab. However, today my EE professor and I went ahead and conducted some tests on the oscilloscope for some basic AC resistor circuits. We determined (with calculations/simulations) that there was some kind of unaccounted for inductance and capacitance in the probe and/or the oscilloscope. We had the exact same problem when it came down to measuring the voltage and current rms values with a digital multimeter (it was able to measure such values). I believe, after some research, that we need to calibrate the oscilloscope. The university got these scopes fairly recently, and with many people not having a clue as to what is wrong, I have a feeling that they were not calibrated in the beginning. I plan on calibrating the oscilloscope and then compensating the probes. Then I will analyze the basic AC resistor circuits and see if anything changed.