The Engineer's Place for News and Discussion®

 Previous in Forum: Call for Chapter for 2 NYC books on Climate Change and Pharmaceuticals Next in Forum: Selecting a Wind Turbine

### Subscribe to Discussion:

CR4 allows you to "subscribe" to a discussion
so that you can be notified of new comments to
the discussion via email.

### Rating Vote:

Associate

Join Date: Jan 2009
Location: Houston, TX
Posts: 54

### Measuring Low Frequencies

01/05/2009 8:57 AM

We have a vibration table whose calibration has come into question. There is an input frequency that comes from a program that dictates how many Hz is applied to the vibration. First, I want to measure this vibration with an external instrument using an accelerometer. My problem is that I cannot find an accelerometer that can measure less than 2 Hz. Any suggestions?

Pathfinder Tags: accelerometer calibration frequency HZ vibration
Interested in this topic? By joining CR4 you can "subscribe" to

Check out these comments that don't yet have enough votes to be "official" good answers and, if you agree with them, rate them!
Guru

Join Date: Mar 2007
Location: City of Light
Posts: 3311
#1

### Re: Measuring Low Frequencies

01/05/2009 12:25 PM

You can build one with a mass between two preloaded spings and an LVDT for displacement acquisition or with a cantilever and a strain gage bridge. How big is the amplitude of your table? Frquency is not all in fact you want to measure accelerations.

Associate

Join Date: Jan 2009
Location: Houston, TX
Posts: 54
#2

### Re: Measuring Low Frequencies

01/05/2009 2:42 PM
 Duration (s) Avg Amplitude Max Amplitude Min Amplitude Frequency RMS 5.5 0.03 -5.8 -9.64 2.16 5.93 6 0 -0.11 -0.22 118.26 0.16 6 0 -0.12 -0.18 124.88 0.15 6 0 -0.03 -0.16 130.78 0.13 6.1 0 -0.1 -0.15 133.33 0.12 6.1 0 -0.04 -0.24 133.06 0.11 6 0.13 -30.74 -48.78 26.75 31.39 3.1 0.02 0.45 0.03 103.74 0.18 3 0.02 0.38 0.02 104.05 0.17 3.1 0.02 0.43 0.02 102.5 0.16 3 0.02 0.46 0.03 8.49 0.15 3 0.02 0.49 0.04 55.35 0.15 3.3 0.02 1 -0.01 7.22 0.15 3.1 0.02 0.39 0.03 54.07 0.15

This is the type of data I am working with...I am working with limited information here ( I am not even sure the units of measurement here). I will repost when I can gather more. I think I will go with an accelerometer - I have the financial resources to buy one. But I like the cantilever/strain gage idea. I have done something like that before.

Guru

Join Date: Sep 2006
Posts: 4669
#10

### Re: Measuring Low Frequencies

01/07/2009 9:45 PM

Nathaniel,

Your measurement at 2.16 Hz consists of only ~2.5 cycles at this frequency, assuming "Duration" specs the time this frequency is applied. This isn't much of a sample, all things considered. You may wish to increase your duration a bit for lower frequencies. Is there some reason for this such as a concern about overheating your shaker table's driver/amplifier (assuming it is a "voice-coil" type of driver)?

Something you might also try is to sweep your table's driving frequency and create a plot of frequency vs amplitude (aka a "Bode plot"). This will clearly show you where your table is at resonance as well as other, frequency-dependent behaviors. Your vibration sensor may have its own, internal frequency and amplitude dependencies as well, so you need to keep this in mind as well.

Do you have a second, possibly higher-quality, calibrated vibration table against which to test your present sensor? You have several potential sources of error and you need to know what they are and how they're contributing to your overall measurements.

(Note: Your RMS amplitude at 26.75 Hz is considerably higher than the surrounding values. You may be near a resonant frequency here. Your low frequency amplitude may be due to resonance at this frequency, as well. Load your table with a mass and re-take your measurements. See how they differ from your current values. You should see heightened/reduced amplitudes at other frequencies.)

Accelerometers are the most common means for calibrating vibration tables and such are used by NIST. You may wish to use a triaxial accelerometer to determine the purity of your table's real-world vibrational modes, even if the table is designed to move in one dimension only. For example, seen below is a Bode plot in three axes along with a reference using a calibrated laser vibrometer:

This next figure shows how rocking motion in one axis can introduce errors in the laser reference channel, should you use this method:

Associate

Join Date: Jan 2009
Location: Houston, TX
Posts: 54
#11

### Re: Measuring Low Frequencies

01/08/2009 4:31 PM

This actually is data taken from one of our operations - we don't want to crack a mold.

What you are suggesting is actually my next step. I will create a program that will run the vibrating table at certain frequencies for certain durations. I will compare this with other tables that we use, and plot (just as you said) and create a calibration curve based on those plots.

I like that triaxial accelerometer...never used one.

Guru

Join Date: Dec 2006
Location: Germany 49° 26' N, 7° 46' O
Posts: 1952
#12

### Re: Measuring Low Frequencies

01/09/2009 4:45 AM

Hi Europium,

this seems to be a good example of a bad accelerometer mounting?

In the laser-vibrometer distance measurement no resonances are to be seen.

So the a-meter or its mounting has the resonances. A-meter: not very likely.

?

RHABE

Associate

Join Date: Jan 2009
Location: Houston, TX
Posts: 54
#3

### Re: Measuring Low Frequencies

01/05/2009 2:42 PM

btw, thanks for the post.

Guru

Join Date: Mar 2007
Location: City of Light
Posts: 3311
#4

### Re: Measuring Low Frequencies

01/05/2009 3:33 PM

Without units it is impossible to make the choice.

At low frequencies the problem is the zero shift evolution and the possible signal loss when using piezo accelerometers.

2
Guru

Join Date: Dec 2006
Location: Germany 49° 26' N, 7° 46' O
Posts: 1952
#5

### Re: Measuring Low Frequencies

01/06/2009 1:23 PM

Hi,

there are a lot of accelerometers that can measure from true DC to upper limit of some Hz to some KHz.

Best of best (since now 50 years not changed fundamentally in the mechanical design - a rare example) is the Sundstrand Q-Flex, from Honeywell-Sundstrand, Redmond.

Do it yourself: If your rotary table is a big one then you can take a good big pendulum provide the best bearing you can get (zero hysteresis) attach a fine resolution potentiometer (resistive) or LVDT as suggested above or angle encoder readout to some storage or display - thats it.

I designed many different ones since now 30 years, here are the flexible suspensions:

no bearing friction, mass and bearing integrated into one piece of high quality material.

Readout in example with a cylindrical mass was inductive the others capacitive.

RHABE

2
Guru

Join Date: Jan 2008
Posts: 1611
#6

### Re: Measuring Low Frequencies

01/06/2009 2:15 PM

Let us get up into the 21st century with the subject matter (no pun intended). The coil and spring based accelerometers of yesteryears are dead, just they do not know it yet. No joke, nor exaggeration. If you go to Home Depot or similar home improvement store, you can buy inexpensively an Inclinometer, measuring the angle to the horizontal. It does that by measuring gravity on a silicon chip level, made by the same technology as any old transistor, just etching out a cantilever on the small scale.

You can go straight to the maker of Analog Devices, or go to the distributor: www.Digikey.com Subject: MEMS Accelerometers and Gyroscopes. Read the specifications, they are written in a language you like. It measures STATIC gravity and at any frequency up to its limits, which may be 6 kHz. In my catalog they are pg 580.

Instead wiring up your own, buy an Evaluation Board under \$100. It will provide all you need, including connector to your frequency counter or oscilloscope. Good luck!

Guru

Join Date: Dec 2006
Location: Germany 49° 26' N, 7° 46' O
Posts: 1952
#9

### Re: Measuring Low Frequencies

01/07/2009 3:30 AM

Micro-mechanical accelerometers gained a market share but are far away from good quality miniature accelerometers.

One reason is the flexure - much better design and material in macro.

Next reason is the capacitance of the distance measurement. Not so easy (impossible) to measure fast and accurate (5KHz, 10-5) with minimum noise and minimum nonlinearity this gap that is defining the capacitance.

The electrically conducting plates that are the capacitor to be measured have to be parallel, have to be guarded around to prevent nonlinearity from fringe fields and have to have "constant" distance if at zero g condition.

This gap is near 3µm in MEMS accelerometers, how to get this stable to 10-5?

Next: axis orientation.

Next: drift of zero and scale factor.

Everything depends on specifications!

RHABE

Associate

Join Date: Jan 2009
Location: Houston, TX
Posts: 54
#7

### Re: Measuring Low Frequencies

01/06/2009 2:24 PM

Thank you all for the replies. I will move forward in this project, keeping all recommendations in mind.

Guru

Join Date: Feb 2006
Location: Pakistan: 33.2N; 72.16 E
Posts: 1778
#8

### Re: Measuring Low Frequencies

01/06/2009 3:19 PM

Regards.

How is by a digital stop watch.

Measure time & workback for Freq.

__________________
"To pity distress is but human; to relieve it is Godlike." ~:~ Horace Mann ~:~
Interested in this topic? By joining CR4 you can "subscribe" to