07-03-2007 11:14 AM
07-03-2007 02:23 PM
Thanks for your posting,
By using an acelerometer or two, you will be able to measure the stimulus (shaker) and the response (vibration transferred thru the unit under test) with the two accelerometers. By analyzing the vibration, you can determine whether to increase or decrease the vibration level. You may want to perform vibration level measurements, or vibration measurements at specific frequencies. Either way, the output of your analysis should be a scaler value which you can use as a process value for a PID algorithim or other control algorithim. The output of the control algorithim would control for example the magnitude of the vibration signal fed to the shaker.
So the diagram is really 1) measure vibration leve 2) pass level thru control algorithim and 3) adjust amplitude of signal fed to shaker.
Does this make sense?
07-03-2007 02:40 PM
Yes that does make sense. I understand the basics, having worked in a Vib test lab for a number of years. I am more interested in knowing if the specific hardware that I have at hand is suitable for the task. Can I turn the SC-2345 into an actual Vibration Controller? I see the need for an output module to feed a signal to the amp. I want to have a closed loop system, where the control continuously checks the signal on the shaker head, and corrects the signal going to the amp. My goal is to be able to A)sweep a sine wave at either a constant displacement, constant acceleration, or a combination using a crossover frequency
B) run a random profile with a controlled frequency spectrum.
Basicly, we have almost all of the hardware here except for the output module. Since I orked in Vibration/Modal analysis for a long time, I was asked if I could come up with a vibration system, and was hoping to use the NI equipment that we have on hand.
07-06-2007 05:34 PM
01-04-2008 12:08 AM
01-07-2008 06:56 AM
I am getting closer, but have just ordered the hardware, and have not yet started writing the program(s). When I have anything, I will post here.
02-10-2012 11:28 AM
I know this is a complete long shot but has anyone actually figured out how to do what Preston proposed. Either have the acceleration stay steady as you sweep through frequencies or the ability to sweep through accelerations at steady frequencies such that you can directly adjust acceleration, not just the amplitude of your signal, so that you do not have to guess how much you have to change your output signal to step up a certain ammount in accleration.
02-10-2012 11:46 AM
The only example on the net that I have found involved the use of shift registers. The code should be attached below, I was just wondering if there was a more elegant way of creating this system. Code was written by Dave Duimistra at GVSU in 1997 so the code is a little outdated