Rotary Balancing

[DiscoBall]

I believe the link to the Dev. Zone that Raymond is reffering to Two-Plane Balancing Using LabVIEW PDA and NI CF-6004 CompactFlash Data Acquisition Card http://zone.ni.com/devzone/cda/tut/p/id/4381. Raymond, are you wanting to modify this VI for use with an encoder? 

[RL]

Joshua,  that is the link that got me back on this project.  I don't think I have the skills to modify the VI to input an encoder instead of a 0 degree marker.  I was checking to see if anyone knew of an encoder VI that I could adapt or if someone knew of another VI that would be applicable.

Raymond

[DiscoBall]

Sorry, I didn't noticed that the link was posted before (my bad). Modifying for an encoder shouldn't be so bad. The hard work of detecting an edge has been done. It looks like you would just need to divide out the number of pulse per revolution of your encoder. The output of the Analog Tach Process for PDA.vi gives you the pulse position in seconds. Say your encoder gives 4 pulses every revolution, you will need to use every 4th value of that pulse position array and build a new array to output from the Analog Tach Process for PDA.vi.

However, the 4452 is restricted to Traditional DAQ according to the readme found here: http://ftp.ni.com/support/softlib/multifunction_daq/nidaqmx/8.8/readme.html. So, you would need to change the DAQmx Base functions over to Traditional DAQ functions and lose the PDA functions to make this example work for your card and computer...

Message Edited by Joshua B. on 12-29-2008 11:44 PM

[kyle_vi]

RL,

Did you come up with a vi you can share?  I need something to balance a fan..

Thanks

[RL]

I am waiting for some hardware that was back ordered.  The preliminary balancing was done with standard data acquisition and an Infrared tachometer which was used to establish a zero point.  I don't think that what I am doing will be of much use to you because it is item specific.  This is a cantilevered blower where the only place metal can be removed is on the hubs.  So far the measurements where done in g rather than ounce inches.  The distance between the zero points was subdivided by 360 degrees to find the angular position of the imbalance.  A third accelerometer was added to confirm the movement between the two planes.  This accelerometer may prove to be superfluous - we'll see if it contributes meaningful information.

Raymond

[kyle_vi]

Still looks like something useful.  I need it for initial weight adjustments for a molded part.   I would make a change to the mold to add weight where needed.  Looking for a balancer like datastick bal-2000 right now.  But it would be great to do it with Labview.. 

[peper2001]

Hi,

I'm also interested in one plane and two planes balancing of shafts using labview. I got the VI "2 planes balancing PDA" (refered in this topic) that i have modified to be used with non-pda functions.

I have two questions regarding this VI:

1)  I don't understand what append after the filtering of the time-series. There is a processing using complex number that I don't understand. Could someone explain what is going on in this part of the VI?

2) How could I improve the performance of the blancing. I'd like to balance lightly unblanced shaft and the VI do not seem to be able to do that so far. I've tried to change the order and the bandwidth of the filter but it doesn't work any better. 

Thanks in advance!

[RL]

Datastick Bal-2000 does have a higher price than I could afford.  They however offer the services of a training consultant, Derek Norfield, who wrote  Practical Balancing of Rotating Machinery.  I find that the book has a comfortable progression from basics to more advanced balancing solutions.  I  have spoken to Mr. Norfield and he impressed me with his knowledge. 

I opted for a different solution before speaking to him and I suspect that it has already cost me way more in time to go a cheaper route.  Balancing may be straight forward but invariably a situation occurs where persistence alone just does not lead to a satisfactory result.  Standing on the shoulders of informed people is so much easier.

Raymond

[RL]

"2) How could I improve the performance of the blancing. I'd like to balance lightly unblanced shaft and the VI do not seem to be able to do that so far. I've tried to change the order and the bandwidth of the filter but it doesn't work any better. "

Are you certain that you have a software problem?  Are you able to check the raw signal to determine the noise level?  I have run into problems because sufficient noise was making it past the filter to give very erratic results in the latter balancing iterations.

What are you using for hardware?

Raymond

[peper2001]

Concerning my hardware, i use an USB NI-9164 with a a NI-9234 card (4 analog channels). My tachometer is a laser with a reflective tap on the shaft, giving a square signal as the shaft rotates. I also use two 500mv/g accelerometers, one on each bearing.

I've check the raw signals but I don't know how to assess the noise level. All I can say is that they do not look like a sine wave at all.

I've also looked at the filtered signals (the related subvi is "bandpass get first order") but, with 60 taps, it is still very noisy. I have to increase the taps up to 300 to get a nice signal.

Any suggestions? I also tried to use "extract single tone information" to get the phase and amplitude of the unbalance but it doesn't seem to work any better.

Thanks

Christophe