You know, we actually have a set of headphones from CSI. (He was a good salesman) Thanks for the idea of actually using them.

Aubrey

We use the headphones too, and wouldn't be without them. Really valuable.

You also readily 'hear' cable or connector problems, and it will drive you nuts if you don't get it fixed pronto! With an accelerometer, you are listening to acceleration. But all you have to do is use a Wilcoxon 793V to listen to velocity (the integration is done on-board the sensor before the signal returns to the meter). I will try this -- and compare to acceleration -- Thursday and get back to you. As for "listening" in addition to the other tools I use, I'd lick it if I thought I could actually taste a problem!

I have never heard that hearing (ears) works with vibration velocity. Human audio response is based on sensing dynamic pressure (the ear drum). It has been shown that sound pressure close to a sound radiating surface is proportional to surface velocity. Perhaps that's were the association comes from. In the case of listening to bearings, it would appear that high frequency acceleration response would be better than attenuated velocity.

Rusty,
Let us know what you "hear" with the velocity sensor compared to accelerometer.

Walt

How would you use the headphones with a datapac?

Thanks, Joe

Joe,

I have the directions at home I will try to look up. I also have the prints for making a 3 switch modulator to hook the earphones in ot. Basically all you need to do is tap into the cord (cord dedicated to earphoes only) and parallel a phono jack to plug earphones into. There is some circuitry involved,. but it really isn't much. Shoot, you can buy them on grease guns now.
We used to use them 6 or 8 years ago, so it couldn't be that difficult. I think they still sell them (the adaptor), you might check.
I'll see what I can dig up at home.

Dave

quote:

Originally posted by Walt Strong:
....It has been shown that sound pressure close to a sound radiating surface is proportional to surface velocity....
Walt

Yes Walt, that's precisely the basis of my suggesting an "edge" for velocity. If you imagine placing your ear very close to the equipment, you would be hearing the sound radiated by the vibration velocity of the sound-radiating surface. So velocity better represents the "true" sound.
Of course, we all know that acceleration is a well established and effective route for monitoring "events" carried to the sensor at high frequencies.
But in some, but not all cases, there could be other useful sounds to hear at different frequency ranges, eg low frequency things going on in the process-fluid-flow circuit.

Note added, July 27 am. Key points to evaluate, when listening to velocity as compared to acceleration are:
1) Did you totally "lose" the high frequency activity related to metal to metal impact etc, or is it still there but now masked to some extent by other sounds?
2)How does the nature of the sound, to the ear, change, if at all?
3) Did you hear anything else of interest / potential use?

This message has been edited. Last edited by: <Ron Hartlen>,

Rusty,

Did you get a chance to compare listening between accel and velo?

quote:

Originally posted by Ron Hartlen:
Listening to velocity (given adequate HF response) has a bit of an edge on acceleration. It is actually velocity that the ear responds to. So, if the sound has a mixture of frequencies and/or broader band, a velocity signal would give a truer replication of what the ear would hear, whereas acceleration has a bias for higher frequencies.

Ron,

I am in agreement that human ear responds to velocity. Velocity in this case means air pressure waves velocity. But pressure waves velocity produced by a headphone is proportional to the VIBRATION signal ACCELERATION if hooked up to an accelometer. That is what a headphone does.

So, it appears to me that we have exactly what we need, namely we are listening to the acceleration signal (what are intersted in) which is being converted into velocity - a proper physical parameter the reciever (human ear) is most sensitive to.

This message has been edited. Last edited by: David_G,

Ron,

So, it appears to me that we have exactly what we need, namely we are listening to the acceleration signal (what are intersted in) .[/QUOTE]

David G.

Two things for you to think about a bit more:
1) Why do you say that acceleration is what we are interested in? Keep in mind that at very high frequencies, where vibration response is mass limited, when you listen to acceleration you may be in effect listening to forces. These forces may or may not be doing damage, ie perpendicular impact with no sliding (elastic collision) would cause acceleration, but may not do any damage.
2) Go back (conceptually) to the point of measurement with a microphone, velocity measurement and acceleration measurement. Since it's vibration velocity that radiates the sound, the spectrum of the microphone and the velocity signal will be very similiar, ie both representing the "true" sound. But the accelerometer will distort it, overemphasising the high frequencies.
Hope this helps!
RTH

I hadn't really thought about wether I'm listening to acceleration or velociy with the CSI headphone adapter... I honestly don't know which it is.

BUT, when we use headphones, we aren't hearing either one... we are hearing the pressure caused by the movement of the speakers in the headphones. There is a question as to what is causing the speakers to move (raw acceleration signal or velocity signal), but we are hearing the speakers.......

thevibeguy makes a valid point. It should be noted also that when wone looks at acceleration (or listens to it) one is displaying a signal. Hopefully, this signal faithfully represents the acceleration of what one has interest, but this is not always the case.

I hope this thread doesn't fade away before somebody has a chance to try listening to velocity and aceleration, to compare and let us know what you found.
To recap: even if we don't know precisely what the data-acquisition system and headset are doing, there has to be a difference. ie if you feed a different source signal into the process, you'll hear something different at the headset. And I'm suggesting that velocity should give a result more representative of what your ear would hear directly, although the high frequency "content" from the accelerometer approach will probably be "buried" a bit in the other sounds.

I will try and do this tonight... I have a new MP3 player that has a "line in" jack so I should be able to record the velocity/acceleration signals and post them for you. The little digital recorders I've tried in the past (using a microphone input) seem to auto-gain the signal, which is no good if you're trying to compare 2 signals. We'll see if this works.

I also work in a paper mill and we use a 100mv accelerometer. We have a Beloit and a Tampella machine with dryer speeds of .9 to 1.5 hz. We use a SKF CMVA 60. I have had extremely good success useing acceleration enveloping with the 4th filter (5KHz-40KHz) to find cracked inner race. The 3rd filter (500Hz-10KHz) also works but not as well. I've also found that all the other defects ie. race, ball, cage show up better in velocity, whether with a accelerometer integrated or a velocity transducer. I have also used velocity enveloping with a lot of success, but because of time involved, I use one Acc Enveloping and one velocity spectrum on each dryer with a 100mv Wilcoxon 797 accelerometer with a Fmax of 160 Hz. I also use headphones. I can hear the hairline crack on inner races in velocity better than acceleration, but can hear absolutely nothing on a bearing that is practically destroyed. I hope I've been of some help. Let me know if it works for you.

I would have to wonder if the headset would create a parasitic load on the signal and effect amplitude accuracy.

quote:

Originally posted by rustythevibeguy:
As for "listening" in addition to the other tools I use, I'd lick it if I thought I could actually taste a problem!

Rusty,
That is a technique I have never considered.

Aubrey,
If you are confused , it means you are learning new things. That is good.

The whole key to good detection is understanding what you are seeing, hearing, feeling, or (in Rusty's case) tasting. There are various processing techniques that give various results. It all depends on what I am looking for as far as machine defects. There are many factors that we can not control (structural transmission, stiffness, damping) when we doing diagnostic vibration routes. The only thing we can control is the vibration data that we look at. I am a big fan of trying different measurement techniques to see what works best. So far, I have found no perfect solution.

In general, I have to side with the 100mV accels, but I still keep velocity and displacement sensors on hand. I also have various microphones for "noise" hunting. They are all tools that have a proven place in certain situations.

Take Care,

As for velocity vs. acceleration, keep in mind I am not trying to listen to the "noise" coming from the machine, as you would if you are standing next to it. I am listening to the vibration signal from the vibration sensor.

I too wonder about a "parasitic load" on the signal if you are trying to come up with a homemade adapter. I would think this would not be a problem with the CSI adapter.

I did try recording with my MP3 player using the "line in" port attached to the CSI adapter. Works great and automatically encodes to MP3. The bit rate is selectable (98, 120, 160 bits/sec) but since it's encoded, you do lose some of the signal. But I am not trying to record a signal to post process and analyze. I just want to be able to send a customer the sound files and say, "Here is a good bearing, and here is the bearing I think is bad.... what do you want to do."

And it would be cool to "collect" examples of faults to maybe use in a training class some day.

I did not have my velocity sensor (793V) with me, so still don't have a velocity/acceleration comparison.

When a transducer signal gets converted into air pressure waves and we listen it with a head phone set what we want to have is be able to distiguish an abnormality, such as, "a click" sound on a background noise. If this click is standing out more profoundly when the raw signal is acceleration, then this is the best unit to detect a problem, if velocity, then velocity raw signal is the best.

I just expressed previously an opinion that acceleration, which gets larger proportionally to the frequency will be the best to make this sound standing out. Yes, acceleration is proportional to the force causing it, but that is exactly the indication of a problem.

Of course, an experiment is the king. It is impoirtant to make it representative, with various defect types and various background vibration.

David

Hello Ron,

I am pleased to hear you share the same dream that I have: portable laser vibrometers at affordable prices. I hope I see it before I retire. I have faith it will be possible actually.

When I was a sophomore at college way back when, we had read a story by Isaac Asimov about a certain technician Aub in the umpteenth century. This man somehow discovered how to write numbers side by side and performed arithmetic operations by hand. This was contrary to the teachings of the time which forced use of some equipment that fit in the palm of a man on which he would enter numbers and multiplication, division etc would be performed automatically. I had sighed and wished we had some equipment like that to help us with calculations in exams. Half a year after that I bought my first Casio calculator that performed the four arithmetic operations only and did not know that any number divided by zero is not defined so if you tried it, the calculator would start counting to infinity like a stop watch. I still own it!

So you never know!
Regards
Ibrahim

quote:

Originally posted by VibraTek:
Hello Ron,
.
.
.."This man somehow discovered how to write numbers side by side and performed arithmetic operations by hand.....Ibrahim

Ibrahim,
This reminded me of something that puzzles me. How would one take square roots, by hand, using Roman Numerals?
ps Glad to see that at least one person connected with the main point of my original post on this thread!