Low freq. spectrum meas.

Which product is right for your exact requirements
Post Reply
Arthur

Low freq. spectrum meas.

Post by Arthur »

I am considering purchase of an ADC-216. I need to measure spectral properties below 1 Hz. It looks like the standard scales on the spectrum analyzer have a low frequency limit of 81 Hz full scale.

1. Does software exist to expand this range to about 1 Hz full scale?

2. Do you see any problems in doing this measurement with your software and equipment, other than it will take a lot of time?

Thank you , Arthur

Michael
Advanced User
Advanced User
Posts: 656
Joined: Thu Jul 07, 2005 12:41 pm
Location: St Neots, Cambridgeshire

Post by Michael »

Hello,

You will not be able to select 0-1Hz spectrum range.

You can select the lowest bandwidth and increase the spectrum bands. YOu can then zoom into the horizontal to filter out the 1 Hz range.

Spectrum updates for low frequency FFT's will be time consuming! There are a lot less cycles per second to work with so conversion time is increased dramatically!

Best regards,
Michael
Michael - Tech Support
Pico Technology
Web Support Forum

Guest

Post by Guest »

I'm having some similar problems looking at a 0-32 Hz range. Are the preset frequency bands the only options available? I can see how the x-axis can be customised but cannot see similar for frequncy. At the moment I am looking at a 0-74 Hz range but only using half the data!. Also when I up the number of spectrum bands to anything above 128 I lose the plot and no amount of waiting seems to help.

Michael
Advanced User
Advanced User
Posts: 656
Joined: Thu Jul 07, 2005 12:41 pm
Location: St Neots, Cambridgeshire

Post by Michael »

Hello Guest,

Which instrument are you using for this purpose?

I will try to run a test here to see if I can repeat your problem.

Best regards,
Michael
Michael - Tech Support
Pico Technology
Web Support Forum

Guest

Post by Guest »

I am using a 3424 Picoscope to measure an output from an accelerometer. I am setting the display to Volts the window to Hanning using the 'average' option with a range to 74Hz and Zoom x2 currently I only get a plot when the number of spectrum bands is 128 any higher and no matter how long I wait I get no plot. I am using auto trigger but timing is not important and I would rather have a plot made up from several averages as the vibration is gaussian and I want to try an eliminate any noise.

The laptop I am using is not the latest by any means, it's definitely only USB 1.0 and it's running windows 2000. Not sure if this might be something to do with it?

I am trying to replicate some PSD plots that were created by sampling between 0.39 - 30 Hz, using Hanning Windowing a 256 block size (results in a 0.39 Hz delta F (Copied straight from the document not sure 100% what it means!)) and plotted as g²/Hz (logarithmic scale) vs Hz. I understand that I may need to copy the data from picoscope into excel and manipulate it there but the results using 128 spectrum band lack resolution at the moment hence me wanting to increase the number of bands and sample rate.

The last plots we obtained using a Hardware Spectrum analyser were annotated 120mHz Banwidth, by my possibly flawed calculations if I were looking at a 32Hz range I would need approx 533 samples to obtain this level of resolution

From looking at the signal in the scope mode the peak seems to be approx + - 1V which is supposed to equate to 1g however setting the voltage range on the spectrum to +- 1V results in a barely visible trace.

Any advice help gratefully recieved!

Guest

Post by Guest »

Have you had a chance to try this yet?

Guest

Post by Guest »

I have managed to reproduce similar results on the bench with a pulse/function generator.

Using a frequency of circa 4 Hz on Spectrum with the following settings;
X-scale linear, y-scale dB, Window Hanning, No. of spectrum Bands 128, display mode Average. Frequency limit 76 Hz and zoom x2

I get exactly as I would expect a peak at 4 Hz and decending peaks at intervals up the frequency range.

If however I increase the number of spectrum bands (Increasing the resolution as I understand it) you either get no plot at all despite waiting or a flatline. I am not using the latest software (version 5.13.3) is there anything that I could be doing wrong? Also if I up the number of spectrum bands to the limit of 4000 ish picoscope crashes.

Guest

Post by Guest »

a little more progress,

By increasing the frequency range to 153Hz (zoom x5) and increasing the Number of Spectrum Bands to 2000 I seem to be able to get some reasonable loooking plots.

However this just seems like a work around to me since I'm only interested in the range 0.39 - 30 Hz so logically I should just pick the lowest range (76 Hz) then the appropriate number of Spectrum Bands which is the situation I started with!

Michael
Advanced User
Advanced User
Posts: 656
Joined: Thu Jul 07, 2005 12:41 pm
Location: St Neots, Cambridgeshire

Post by Michael »

Hello All,

We are looking into this problem and there appears to an error with the scope sampling mode.

We should get this fixed in time for the next software release.

Thank you for your patience.

Best regards,
Michael
Michael - Tech Support
Pico Technology
Web Support Forum

Post Reply