improve the hysteresis loop

[siti]

Hi,

I am using DAQ NI6009 to measure a material's characteristics and perform in hysteresis loop. The sample rate of DAQ NI6009 is 48kS/s. I put samples to read = 45, rate (hZ) = 1k. But the loop is coarse as you can see from the attachment. how can I improve the loop to obtain smoother loop?

Thanx

[Dennis_Knutson]

Post either an image or the actual VI. Do not post something hidden in a file with a proprietary format (docx).

[siti]

I'm sorry for that. Here I attach in JPG format. The test has been done in frequency 50Hz. Hope you can help me to solve the problem. Thanx

[KrisnaWisnu]

Hi siti,

I am not familiar with your application, but if you increase the sampling rate and the samples to read, your result should look more smooth. Correct me if I am wrong.

Sincerely,

Krisna Wisnu

72377

Message Edited by Krisna Wisnu on 12-03-2009 09:21 PM

[siti]

Hi Krisna, Thank you for your reply.

I already tried for several values of samples to read as well as increase it. As a result, the trace/line of hysteresis loop become thicker for sure then the loop still coarse and moving a bit. The sampling rate of DAQ NI6009 is 48kS/s, how to increase the sampling rate?Is the best value of samples to read is depend on the sampling rate?

[KrisnaWisnu]

Hi siti,

If you can post the VI it would be easier to see what is happening. What is the frequency of your signal input to 6009? You can increase the sampling rate up to 48kS/s, if you have a loop in your VI, your sample to read should depends on your loop frequency.

Sincerely,

Krisna Wisnu

72377

[siti]

I attach here the VI. The input frequency is 50Hz.

I also have offset problem where the initial point is not at the origin (0,0). How to make the point is at the origin?

[KrisnaWisnu]

Hi siti,

I think 1kHz sampling rate should be sufficient for 50Hz signal, no aliasing occurred. I can't find any reason why the signal looked coarse except that because the signal itself is coarse. Meanwhile for the point is not origin, because you are plotting XY Graph, X components is from Analog Input Channel 0 and Y components is from Channel 1, and the first set of XY data is not 0, so that is why it is not pointed at the origin.

If you want to make it start at origin (0,0), you can offset the whole data. Example if the data begin with (-2, -0.035), you can offset it by adding +2 for X, and +0.035 for Y component.

Sincerely,

Krisna Wisnu

72377

[Henrik_Volkers]

Do you use 50Hz directly from the powerline, or do you use a generator?

Do you have additional frequency content in your signal ( any power/FU converter nearby)?

Sample both signals with highes possible samplerate for 1s (48kS/s : 48k points)  (or 2s) and make a FFT to look at your two signals.

'Clean' higher order frequencys are due to your hysteresis (n* 50Hz), all other will make your trace move around.

[siti]

Hi Krisna,

Thank you for your suggestions. Actually the measurements took the voltages at the primary side and the secondary side of the ring core to perform the hysteresis loop at the XY graph. But the waveform of the voltage taken is no static and moving a bit. Is that the cause of the coarse loop?