Phase Delay

[navneet]

Hi Lynn,

I am trying to measure very small phase values between a stimuli and response.

I have tried all networks and have not noticed any response from channel delays.

So it got me interested why isn't it showing the inter-channel delays.

Of course this is good for me!

I remember when I saved the two signals in a text file and then measured the phase difference between them, there was a large inter-channel delay.

But then when files are saved - signal inegrity is lost as all signals are referenced from t = 0.

So I measured immediately after acquisition and it did not show any inter-channel phase delays, exactly what you postulate over there.

With this I have been able to veirfy some RC networks and a simple circuit consisting of two resistors. 

Feeding the voltages directly to two differential input channels gave phase difference of zero. This is exactly what I want.

I have also been able to verify a lot other networks, each is absent from the inter-channel delays. So experimental evidence is sound - but am curious how this is "cleverly" avoided in the extract tones subVI. I split the two signals and send them to separate extract tones subVI.

[Henrik_Volkers]

Here is a quick test vi for the extract single tone detection

Play around with the noise , samplerate , number of samples and phase(s)

The histogram will build up if you choose 'run continously' , not nice but quick&dirty 😉

[navneet]

Hello Henriks,

thanks for the quick example. I dont have 2011 though, could you please downconvert it?

I will only be able to look at it at work - I am going tomorrow.

But the attached image answers some important characteristics as I see it.

As we see - the extract tones does not consider the start phase (3deg). It fundamentally looks at the delay (0.03deg)

and thus computes the delay as 0.03 deg, and not 3.03deg.

I have not gone into the mathematics of it, but it appears it does not get affected by the start phase?

All in all, I have done numerous verifications so I am flowing with it. I just wanted to see how does extract tone avoid this pitfall.

[navneet]

Ah

I forgot.

After acquiring the two signal via DAQmax, I send it through a "extract portion subVI". This extracts waveforms after a small delay (200 ms) to account for delayed reading after write (using a sequence). Will this be making the difference? Something like doing the job of aligning waveforms?? It just clicked to me now.

[Henrik_Volkers]

What version do you have?

Can you send your vi? Or the part from capture to phase detection?

and the extract tone does consider the start phase,  what is displayed is the difference in phase (defined  30m deg and measured 0.0300003 deg in the pic 😜 )  and what is shown in the histogram is the difference (or error) of the algorythm due to noise between the analytical signals and the result of the two tone extractions from reference and delayed signal.

[navneet]

Hi Henriks,

Now I am going to make a short trip to work!

I want to see whether its the extract portion of signal express VI that is making the difference.

I believe it is making a realignment effect, thus the removal of phase delays. What do you reckon?

Will message you parts of my codes when am at work.

[navneet]

Lynn and Henriks.

Attaching here a part of the code showing the acquisition and computation of phase difference.

Checked with and without the extract portion of signal VI. No difference.

Henriks, I am using LabVIEW 2009 SP1...my bad...will check the example you attached.

[Henrik_Volkers]

If you post a picture do it in the correct format!!

Just rename a 2MB BMP to jpg is rood !!

I will have a look if you post it as a PNG (best for grafic) or jpg

Even better post the vi (probably smaller)

Sorry

[Henrik_Volkers]

here is a 2009 version

[navneet]

Hi Henriks,

Sorry!

I will pm you the section. 

Hope you dont mind.