Phase tracking to maintain 90 degrees between 2 analogue inputs for resonant frequency

[djnckln]

Dear all,

I am using labview for some experimental work. I want to vibrate a cantilever at its resonant frequency, I am both measuring and exciting the cantilever using piezo transducers all through Labview. I want to keep the excitation and measured signal at 90 degrees phase difference and measure the frequency which will be at resonance. I am struggling to keep the signals at 90 degrees phase difference in the loop. I am new to using Labview and would appreciate any guidance. I have attached the program. 

I am using the spectral measurement VI to obtain the phase in degrees of both analogue inputs and have tried to add and maintain 90 degrees to the drive signal. 

Thanks in advance for any help.

[Kevin_Price]

You may as well know this sooner than later.

Not gonna happen.  Not under Windows.  And not with buffered tasks.  Something like this would almost certainly require an FPGA implementation and require many realms of expertise that are *very* far removed from the reliance on DAQ Assistants and Express VI's that are shown in the posted code.

Don't shoot the messenger, just trying to save you a lot of wasted time pursuing the unachievable.

-Kevin P

[Henrik_Volkers]

My first go would be a pure analog solution. If you realize a (positive) feedback to the amplifier (with limited output 😉 ) you will end up in oscillation. (DIY amplifier always will oscillate, while DIY oscillators refuse to do so :D)

The feedback network could involve some sort of PLL to lock it at 90° ....

If the cantilever is the piezo aktor itself and don't need that much voltage , I would try to feed it with a negative impedance ... did that to test quarz resonators , but don't have experience with cantilevers .... two transistors and some R+Cs ...

[Kevin_Price]

Henrik's suggestion to use analog circuitry sounds *way* better than anything I said.  Kudos to him!

-Kevin P

[Henrik_Volkers]

Today I happened to talked to a colleage who knows more about AFM cantilever etc...  I have no experience with this ...

One point he mentioned was, that the resonance frequency of the cantilever usually don't change, so it's just to find it once ...

So it seems that no fast controller is needed to track the resonance. Analysing the signals and looping until the resonace freq point is found can be done in LabVIEW.... (and he did it by sweeping the generator from coarse to smaller ranges , find the peak(s)  and set the desired freq. (usually on one slope , not on the peak resonance, so one can detect the direction of the (amplitude) change..

However you mentioned you want to look at a frequency change.. so some properties of the cantilever change during the measurement task. Seems you don't have a AFM 😉

How fast do they change?

How fast do you need to track that change?

To detect the phase .. have you tried the tone detection vis ?  (extract single tone information.vi) ..  feed it with more than 10 periodes .. also work with DFT, but gives better results ...