Vibration Isolation at 100 Hz

I had this problem, but it was at 60 Hz, problem was I was running at 60Hz so the dame thing shook itself to death. So I removed the source of the vibration. If you find the source, you may find it cheaper and easier to remove then it is to dampen it out.

100Hz sounds like your running speed is 50Hz and something is exciting the shaft every 1/2 rev. or 2 X RPM. If that is true, then you may have a bearing problem, too much oil and too much play between the shaft and the bearing will allow oil whip and that will show up at 2X RPM.

Let me know what you find out.

i can't remove the the thing that is making the vibration because we use it for temperature control of our unit (there is NO WAY that we have time/money/energy to find/use something else)

the vibration is definitely coming from the 2 brushless muffin fans (that are used on the TE cooler, as you state, I'm certain that it is caused by an imbalance within the fan motors, but it is and OEM unit we integrate and am sure that the vibration issue we are experience is only being seen by us.)

An air gap between the TE unit and my chasis has been the only way to totally eliminate the noise to this point.

fwiw: the instrument is a lab based unit measuring 8" x 16" x 10"

Surely it is unlikely that 2 fans would stay in phase precisely. Is there any heterodyning?

If it is exactly 100Hz and this is not the fan rotational speed then you can probably assume that that this is 2x line frequency and has nothing to do with bearings or imbalance. You could then be looking at some sort of stator eccentricity (uneven air gap between rotor and stator). This would become worse when the motor is "twisted" when bolted down to an uneven surface.

When I have had this sort of problem, I have been able to physically separate the units and then pipe the air through concertina type plastic hose such as is used for tumbler driers (but is available in different diameters). If such an arrangement is not out of the question for your customers, you may be able to keep the thermoelectric coolers close to the units (as may be necessary to keep any control loop under control), and just pipe the room-temperature input air. I don't know why, but I have found that sucking through such a pipe sometimes gives different results from blowing.

Plastic screws are transmitting the vibration.

Yeh...well they would

Suggest you use large adhesive pads, or Room Temperature Vulcanising Silicone Rubber as an adhesive. Aim for large surface area and low density foam.

I'd guess it's a lot of trial and error, I've no idea what mass you are trying to support or the magnitude of vibration.

Maybe hang it on rubber straps from a frame work (sky hooks?)

Sorry if this is no help...

thanks for the support..........

i decided to try and attack the problem at it's origin....i ordered new fans (how i'll integrate these into production is a problem i'll deal with if they work) , supposedely low noise, but i figger that, that is audible, but they claim that their bearings are good, i think the bearings are the key in this case.....

we are also running tests to determine if the noise we are seeing in the lab is actually having any effect on performance...........

fwiw: even with the fans off, we have noise in this region, which points to the 56-60 hz that surrounds us

You could use a double layer isolation. You should mount a flexible sheet (thick rubber) with plastic screws on the top of the first layer, then mount the second layer of isolation and fasten the cooling system to the sheet with differently located screws.

The plastic screws will transmit the vibration as you have determined. You likely need to construct a mechanical filter. This will involve eliminating the screws to isolate the cooling system from the instrument. The compliance of the connection and the mass of the cooling assembly will determine to what frequency the isolation is effective. A compliant connection can be made with rubber isolation mounts, elastic bands, foam rubber, springs, or any number of structural elements. It's all compliance, mass, and damping.

I would recommend some items:

1.show up your function block of test condition..

2.If possible, sow up your test data for spectrum via micro phone or vibration sensor.

reasons:

1. axial fan will have stall problems if pressure drop too high and cause axial vibration.

2. You can compare the test data of acoustic data and vibration data to see what's going on.

thanks for the replys......

I have tried various types of rubber/foam to try and de-couple the fan vibration from my unit but the only foam that has worked so far is too soft to be able to use (it is egg crate looking packaging foam). If I compress it at all while trying to mount the unit the vibration transfers thru again.

SO, what I need is a simi-stiff material that does not allow 100Hz to pass.

If I can find a stiff enough foam/rubber that eliminates, or at least dampens it considerably, that I can also use for mounting, my day is done.........and it is beer thirty.......

There is some fancy math going on to generate the output and my guru here analized things and made me a chart that showed the frequency of the noise to be 50Hz

BUT

We then tried to determined the frequency by mounting the unit on top of a speaker and then driving said speaker with a known Hz until the spike moved to the location that we see when the fans are running. This was at 100Hz.

When we drove the speaker at 50Hz there was no noise in the region that we see the fan noise.

I am/was concerned that it is acoustical noise, because we are running a Fast Fourier Transform but can't figure out how to determine whether it is acoustical or vibration. That being said, I'm not sure that it even matters because.......

If I seperate the fan assembly from the unit by only a hairs width, the noise goes away, so whether it is acoustical or vibration the only way to remove it, thus far, is totally isolate the fan assembly from the unit. So if it is sound, it is only strong enough to affect the unit when it has a path other than thru air.

I should get my new fans today, but probably won't get a chance to intall them until monday, so i have all weekend to think about this..........

Even with plastic screws, the fans are still hard-mounted to the chassis. To eliminate the vibration pass-through, you need to completely isolate the fans.

The drawing on the right illustrates the concept. The rubber blocks completely separate the fan from the cabinet and dampen the vibration. Ensure that the screws connected to the cabinet have at least 3/4" of rubber between them and the screw connected to the fan cage.

You may need to experiment with different types of rubber to find one with the right resilience to absorb your particular frequency of vibration.

You can also find commercial anti-vibration mounts on-line. A quick google search led me to these. The pins/bolts are molded into the rubber, but do not touch internally. The company probably has a variety of models for different frequency bands.

Sounds like the 100 Hz is a resonance based on your standard mounting.

This may not be the answer you are looking for, but if you can mount it on springs that may be you best shot at isolation. You design the natural frequency of the mass of the unit on springs to be such that vibration engery is not trasmitted.

The other approach is to add damping to the system. Sounds like foam has given you both a softer stiffness support with added damping.

Here is the theory, based on industrial machinery, but the equations are the same.

Fabreeka is a good company and may be able to help.

(I have no association with Fabreeka other than being their customer occasionally)

http://www.fabreeka.com/tech/isolation_theory.pdf

Hii,

Please check www.las-par.com . This company will help you about your situation.

I hope you will solve your problem.

Thank you

update:

I tried using different fans with different bearing types, sleeve, ball, patented gizmo, etc. The vibration still shows up (it moves around in frequency a bit). Unfortunately, the best performing fan from a vibration standpoint, does not have adequate air flow or pressure for the cooling application.

i still have not found a material (foam/rubber) that is dense enough to use for mounting but soft enough to damp the vibration. I have a vendor sending me some of their damping foam, but am not expecting too much.

I can almost totally eliminate the vibration from affecting our signal by laying the fan assembly on a soft piece of foam (which lays on top of the machine). But, i still haven't determined a way to attach it. Every attempt has failed thus far. The most promising was by using zip ties. Today I'm going to try using rubber straps.

I'm beggining to think you have a resonance of the structure at 100Hz, not so much that the excitation of the fan rotation is the main problem. By detaching the fans you radically alter the mass of the system which will alter it's fundamental frequencies of vibration upwards, moving them prabalby well away from trouble.