[FPV_GTp]

would be nice to have one of these machines $$$$$$$$ High-speed turbo balancing machine [nomedia="http://www.youtube.com/watch?v=TMA5RFpGRbg"]YouTube - High-speed turbo balancing machine[/nomedia]



Amazing with all that technology they still remove metal filings of rotating assembles being balanced and it mixes with the return oil lubrication system while the shaft assembly is being spun to operating speeds to be filtered.

[neilw20]

Rik.

Most of the noise comes from the drive system.
The old analogue bandpass filter suffers a large phase shift for a small change in frequency. The phase gives you the position which once the Q is high enough to get rid of the noise, the angular error is significant.
A small angular position, once the balance becomes finer becomes an error at close 90 degrees.

Don't worry about speed regulation.
Just spin it fast, release the drive belt and process on the spin down.

You can use a Kalmen filter http://en.wikipedia.org/wiki/Kalman_filter to resolve signals at 0 and 90 degrees, say, call them X and Y.

I reality you simply multiply the signal by plus/minus 1 at the correct times for 0 and 90 degrees, and save them, along with the current speed.

sqrt(X^2 + Y^2) gives you the vector for each measuring plane.
Multiply X and Y by 1/(speed squared), to get rid of the speed problem.

Stuff all the vectors in and array for discrete speed steps, and process as percentiles, and throw out the wild values caused by silly resonances.
You need 0 and 90 degrees signals optically from the shaft, and the signals for each measuring plane.


For this to work it needs to be a high tune system where you measure the unbalance FORCES. Not deflection.
Piezoelectric transducers against a large mass, work well.
You can resolve the vector to the balancing plane(s) from the measuring planes by simple moments about the measuring points treating the shaft as a beam.

[Rikkepic]

Hi Neil,

Tnx for your reply.

I tried to do build it in digital environment, but it introduced even more phase lag. The idea was to use a DSP, and implement a 100th order low pass FIR filter in it, convoluted with a delta pulse ( in order to get a band pass) and then multiply the samples with that filter. In this system the phase shift was not constant, so that was the big issue, and the reason why I got back to analog domain.

The first thing I did was making the signals visible, so i can see all the frequencies present:

picture1

and then zoom in on the frequency of interest:
picture2


I measured the signal with an usb Handyscope, and plotted the signals with Matlab. This is the FFT at 7800rpm, where i got the best result.

Then, i made a simulation of a band pass filter in LTspice
picture
I took as example a LT1363, and made a band pass filter with it at 130 Hz (i.e. 7800rpm). The red line is 1st order, de green one 2nd.

Some people here in holland drive the shaft first with a compressor. Disadvantage is that u will always need a compressor. So that's why I thought to hold the turns stable, and make a band pass filter on that stable frequency.

The method you described we here know as the least squares method, where you draw an optimal line trough points (correct?). The question i'd now like to ask is why you use this theory/method. Can you show some measurements and analysis you did? It looks very interesting, but also quite complex(). Do you have pictures of set-ups? I wonder why you use such a complex method, and what the information is you get (because that's not quite clear to me).

Please let me know what u think, i could use some help

Cheers,

Rik

p.s. turbinecrazy: If u like me to start me a new topic, please let me know. i don't want to "steal" your topic or something like that.

[neilw20]

Rik,

Don't process in real time. You have got no hope.

I see you said you 'tried' to do it digital. This definitely needs planning and no trial and error stuff.

In real time, about all you need do is sum the X and Y data, in the correct sign and multiply by 1/speed squared, which filters it, and store in the correct array for each speed interval. Needs heaps of memory.

Now when you process each speed interval, use percentiles, to toss out the wild data.

Better still store ALL the RAW data and process it later. You can get rid of wild rubbish better that way.

Process it ALL LATER, when you apply the brake, if there is one.
You can't get phase lag if you time tag it properly, which is how it is filtered too.

A comment: If your web pages were simple, without all the fancy stuff, it could be translated by Babelfish by just pasting the links.
Be really nice if http://babelfish.yahoo.com/ worked on your web pages. I use it all the time, for non-English stuff, but it won't work past your home page.
You can't expect a translator to process silly embedded stuff, and a whole heap of your page is lost by the fancy, useless menu stuff.
It is getting rid of the silly stuff why all my projects work properly.

[kulk_sd]

Rik.

Most of the noise comes from the drive system.
The old analogue bandpass filter suffers a large phase shift for a small change in frequency. The phase gives you the position which once the Q is high enough to get rid of the noise, the angular error is significant.
A small angular position, once the balance becomes finer becomes an error at close 90 degrees.

Don't worry about speed regulation.
Just spin it fast, release the drive belt and process on the spin down.

You can use a Kalmen filter Kalman filter - Wikipedia, the free encyclopedia to resolve signals at 0 and 90 degrees, say, call them X and Y.

I reality you simply multiply the signal by plus/minus 1 at the correct times for 0 and 90 degrees, and save them, along with the current speed.

sqrt(X^2 + Y^2) gives you the vector for each measuring plane.
Multiply X and Y by 1/(speed squared), to get rid of the speed problem.

Stuff all the vectors in and array for discrete speed steps, and process as percentiles, and throw out the wild values caused by silly resonances.
You need 0 and 90 degrees signals optically from the shaft, and the signals for each measuring plane.


For this to work it needs to be a high tune system where you measure the unbalance FORCES. Not deflection.
Piezoelectric transducers against a large mass, work well.
You can resolve the vector to the balancing plane(s) from the measuring planes by simple moments about the measuring points treating the shaft as a beam.

I have read at places about wattmetric instrumentation for balancing machines. It is claimed that it given most accurate results. Please let us know more about it, if somebody has some information.