Wrong Phase using the Lock-in Tool kit

[Epsilon]

    Hi!

We have been using the NI Lock-in toolkit. To test it, we have connected it to a system under test, that simply consists of a R/C low pass filter. This is useful, since we can calculate the exact response in advance.

The amplitude (gain) response is correct and is always the same, however, when looking at the phase response, the first section is looking good but then it produces seemingly random and erratic results everytime.
The phase response jumps around and produces 180 and 360 phase shifts.

Has anyone ever had that problem? Any explanations for this behavior? Any remidies?

Thanks,
Markus Tarin
movimed.com

[Berto]

Hi Markus,

I'd like to know what Lock-In Toolkit you're using.  Is it the one that you can get at this link:  http://digital.ni.com/express.nsf/bycode/lockin?opendocument&lang=en&node=seminar_US

or the newer example code that can be found here (this one is based upon DAQmx):

http://sine.ni.com/apps/we/niepd_web_display.display_epd4?p_guid=042319214C65627AE0440003BA7CCD71&p_...

Let me know which one you're using and if the alternate helps out at all.

Thanks,
mike

Message Edited by Berto on 02-20-2006 09:06 AM

[Epsilon]

    Hi!
Thanks for the reply.
I am not sure, if we have downloaded the startup tool kit, yet.
We have been working with two different versions of the lock-in tool kit, one for traditional DAQ and one for DAQmx.
I keep on wondering, what difference it makes, which hardware we are using, since we can even simulate the data on the computer and we are getting the same results. Therefore I would conclude I could potentially use any suitable hardware. We are intending to use a S-Series board down the road.

I will definitely look into the links and see if there is anything that we missed.

Thanks so much.
Markus Tarin

[malef]

Markus,

could you please state your setup of controls when you simulate the data?
Perhaps you just have to insert "Unwrap Phase.vi" into the program.

Manfred

[Epsilon]

Hi Manfred!

Our setup is very straight forward. We are using a generic M-Series board to generate the stimulus signal as follows:
Output frequency for x-number of cycles - measure gain/phase response.

To sample the response, we are using an S-Series board. All of it is done using DAQmx drivers, which again, I don't think makes any difference, since we are experiencing the same problems with a 'pure' software simulation.

Our unit under test is comprised of a resistor and a capacitor, which represents a simple low pass filter with predictable gain/phase response.
It appears to us that the erratic phase response at higher frequencies is a result of 1pi versus 2pi phase wrap or flip, which may not be handled correctly by the lock-in tool kit.

We were thinking of "unwrapping" the phase, but I wanted to find an explanation for this behaviour first, to ensure that we are not dealing with a software bug.
Also, where can we find the unwrap phase.vi?

Thanks,
Markus

[malef]

Marcus,

Unwrap Phase ist part of the advanced analysis library and can be found in
Functions - Analyze - Signal Processing - Time Domain
Please state your setup for pure software simulation.
I dont have time to measure it, but if simulation gives same results I may have a look at the simulation.

Manfred

[Epsilon]

Hi!

Thanks for the information. We will try out the unwrap pase.vi to see, if that solves our problem, first.

Best,
Markus