While doing bump test on various structures I have observed that on some of them I would get either a "Waiting for trigger" or "Overload" message after bumping. I varied trigger level, does not help.
I am using CSI and 4x4 for bumping. Rubber mallet does not fix the problem either. Same settings work on other structures.

Has anyone experienced something like that?

Thanks.


Thanks

Try to change Autorange to ON if it's OFF.

If hitting the CSI with a 4X4 don't help, then I wouldn't know what to suggest!

I remember that the CSI Analyzer’s manual mentions how to adjust the parameter so that you avoid unnecessary impacts due to the frequent appearance of these two messages: “Waiting for trigger" or "Signal Overload".

Dave,
I don't know if your problem is due to a CSI setup, but make sure you have changed from a hanning window to an exponential or flat window. The hanning window cuts off all of the energy from the impact.
Ron Brook

... OR forget the trigger and use a multiple random impact method. Start beating the object (This also relieves tension in the person doing the beating.) with no particular rythm and start the averaging. Use the Hann window (often called a hanning window, but the guy's name was Hann.). You get more energy into the modes this way.

Set the number of averages (>1 like 8 or 16 - this also gives reason to beat longer.) to smooth it out.

If you are using a two channel analysis forget the above.

Exponential averaging will add apparent damping to the resonances, but if you use software to extract the modal parameters it probably compensates for this.

Dave,

If I am doing an impact test without a force hammer and 2-channel analyzer, then I use the rapid hit method as Bill suggested. Finding an optimum trigger level can be frustrating, so I rarely use that method. I strongly recommend that you measure a spectrum before you start impacting. This will help identify whether you really input energy into the structure in the frequency range of interest. I do not use wood for impacting, unless that was the last resort. It is not very effect for large structures where low frequency energy is needed. Everyone has their weapon of choice!

If you really want to do the triggering method, then you might try setting for analog integration (CSI products) and try velocity or displacement instead of the more typical acceleration units. Another suggestion: monitor waveform and then set trigger level slightly higher than highest peak from normal background. If machine is running, then use Negative Averaging method.

Walt

Thanks to all.

It appears that activating "Autoranging" helped eliminating the original problem while running the standard single channel bump test. Setting up proper triggering level, of course, is also very important. Therefore, taking a TWF while bumping prior to actual bump test will help.

But for assessing resonance frequency only, normal spectral data acquisition using Hann (thanks Bill) window can be used with random bump excitation.

David

For those on the board with dataPAC's. The triggering function works very well. The analyzer brings up a separate screen with a vertical overall magnitude bar. On the side of the bar is a pointer that is movable with the up and down arrows. The user can impact the structure and see at what sensitivity setting the analyzer will wait for the triggered event.
I have always used my force hammer, even for single channel impacting. The tip is very soft and there is a 20 lb. additional mass for the energy necessary to excite low frequency issues.
I don't subscribe to continuous impacting as it tends to cause 'ringing' in the data. This will occur if your impact occurs before the existing impact time window has finished. This can be a problem on low frequency testing.
Also, ALL windows effect the data. If modal testing is what is called for, then use it. Usually, with impact testing, one is only interested in identifying modes that could coincide with fundamental frequencies or their predominant harmonics.
Also, make sure you have used a pretrigger. This also insures that you don't lose any of your impact energy.
I also use 1 volt/g accelerometers. They will give you more energy output.
If you are going to do impacting, do it right and your results will be much better;

Use an impact hammer with a very soft tip and a good sized mass attached to the back of the hammer head

Use a 1 volt/g accel

Make sure your analzer has a pretrigger set up (5% delay usually is sufficient)

Use an exponential window to maximize your energy from the impact.

Set the trigger and use it to average your hits. Four triggered hits is good enough.

Usually when people have trouble with the trigger, it's due to using too hard of a tip. This spreads the energy over too wide of a frequency range and doesn't put the energy at the frequencies of interest.

Good Luck!

Since I discovered the wonders of the Commtest vb3000 or the latest vb7 it makes all this expensive triggered data capture piezohamme kit etc redundant. Just set to continuous time wave form, hit the structure with a mallet in a regular pattern then down load to the Ascent software and 'slice the TWF'to give a great FFT Spectrum. You can pick any pulse to see if it varies. A great tool lowering the cost of vibration analysis

Are you sure that you are using a CSI 4 x 4?

That is correct. CSI's hardware option with no instrumented hammer is officially
called CSI2130-4x4.

David,

Where do you get the 4x4 from CSI or the lumber yard? CSI must offer a Calibration Certificate and carrying case!

Walt

I agree with Mr. Bill other than I usually set up for several averages maby 200 amd stop it when it looks OK. I also use Peak Hold most of the time.

CSI is gone by the was side so maby the bashing should stop.