Author Topic: Building a vibration tester?  (Read 15018 times)

0 Members and 1 Guest are viewing this topic.

Offline ftransformTopic starter

  • Frequent Contributor
  • **
  • Posts: 728
  • Country: 00
Building a vibration tester?
« on: August 16, 2012, 10:59:22 pm »
What goes into vibration tests of PCB's?
I have an old motor that I thought about converting into a vibra table. The idea is to transfer the rotational motion of the motor into a linear up and down motion of a rod with a wooden surface attached. By varying the motor speed one can vary how fast the surface moves up and down. I would attach the motor to another shaft by a rubber belt as not to shake the motor.


What are the requirements for vibration testing? Are you supposed to test at certain frequencies or do you just shake it as much as possible?
Is this more complicated then I thought? I've read about unbalancing a motor and fastening it directly to your surface in order to create vibration but wouldn't this destroy the motors bearings?
« Last Edit: August 16, 2012, 11:02:24 pm by ftransform »
 

Offline EEVblog

  • Administrator
  • *****
  • Posts: 39271
  • Country: au
    • EEVblog
Re: Building a vibration tester?
« Reply #1 on: August 16, 2012, 11:07:54 pm »
I was thinking about building my own vibration table, maybe from one of those vibration exercise machines?
As I no longer have access to vibration gear from work.

Generally you want to do sweep testing across a range of frequencies, designed to incite component resonance.
There are standards for this, and ones designed to simulate air and road transportation etc.
Response is generally plotted on a Digital Signal Analyser, although PC based systems are very popular these days.

Commercial Shakers come in various types. Ling Dynamic Systems is a major player, now bought out by B&K

Dave.
« Last Edit: August 16, 2012, 11:09:37 pm by EEVblog »
 

Offline ftransformTopic starter

  • Frequent Contributor
  • **
  • Posts: 728
  • Country: 00
Re: Building a vibration tester?
« Reply #2 on: August 16, 2012, 11:09:33 pm »
Would you build yours by unbalancing a motor or by my method?


A electric vibrasander might work for some smaller projects, the one I have has a speed control pot on it.
The big motor I had in mind could probably vibratest a rack mount server though...
« Last Edit: August 16, 2012, 11:11:38 pm by ftransform »
 

Offline EEVblog

  • Administrator
  • *****
  • Posts: 39271
  • Country: au
    • EEVblog
Re: Building a vibration tester?
« Reply #3 on: August 16, 2012, 11:16:35 pm »
Would you build yours by unbalancing a motor or by my method?

I've always used proper permanent magnet elecrodynamic shakers, so don't know about hacking regular motors.
My first thought is to at least use an existing vibration plate system, so those exercise shakers look like they might do the job.
I've been tracking a few on ebay, you can get them cheap.
No idea how linear they are though, what range they work over, or how easy they are to drive.
You always have an accelerometer on the shaker plate so you know the absolute level you are putting into the device.
Then secondary accelerometers on items or boards of interest within the product.
Although just a swept frequency shaker table on it's own would be useful to see if a product survives and to find possible weak points.

Dave.
 

Offline ftransformTopic starter

  • Frequent Contributor
  • **
  • Posts: 728
  • Country: 00
Re: Building a vibration tester?
« Reply #4 on: August 16, 2012, 11:20:02 pm »
That does sound like a more neat solution unless you need to test something very heavy.
Its always good to see snake oil come in handy!
 

Offline ftransformTopic starter

  • Frequent Contributor
  • **
  • Posts: 728
  • Country: 00
Re: Building a vibration tester?
« Reply #5 on: August 16, 2012, 11:43:46 pm »
Hmm I think my idea is shafted... the motor I have is 1750 rpm, meaning that it only rotates 30 times a second, enough for a 30 hertz vibration....
I see the commercial machines go to 500 hertz at least.
 

Offline PuterGeek

  • Regular Contributor
  • *
  • Posts: 88
  • Country: us
    • SolutionsPLUS engineering
Re: Building a vibration tester?
« Reply #6 on: August 17, 2012, 02:01:12 am »
This is actually pretty hard to do. We had a vibrating table at my last job and thought it was really torturing our products until we put an accelerometer on it and we couldn't even get 2 G. It was basically an eccentric attached to the platform which was mounted on isolation mounts which allowed it to vibrate in X and Y with wasted energy going into the Z direction.

It is pretty challenging when you look at the energies involved. Limiting the energy to one dimension means it has to be strong enough in the other dimensions to survive and small voice coil vibration actuators are driven by multi kilowatt amplifiers. This translates into heavy and expensive even when you are buying used.

We tried to come up with something but couldn't find a combination that could vibrate a few pounds up to a few Gs. We ended up trying to get a used unit but these are thousands of dollars. When I left the job they still hadn't bought anything nor paid for testing. The products will pass a basic vibration test but are failing after hundreds of hours of vibration. Rather than spend money they just keep the marketing shuffle going.

I would be interested in anything anyone comes up with even if can only test very light objects (ounces rather than pounds). I expect the amplifier is the biggest challenge but the motor is non-trivial.
 

Offline ftransformTopic starter

  • Frequent Contributor
  • **
  • Posts: 728
  • Country: 00
Re: Building a vibration tester?
« Reply #7 on: August 17, 2012, 04:37:55 am »
This is actually pretty hard to do. We had a vibrating table at my last job and thought it was really torturing our products until we put an accelerometer on it and we couldn't even get 2 G. It was basically an eccentric attached to the platform which was mounted on isolation mounts which allowed it to vibrate in X and Y with wasted energy going into the Z direction.

It is pretty challenging when you look at the energies involved. Limiting the energy to one dimension means it has to be strong enough in the other dimensions to survive and small voice coil vibration actuators are driven by multi kilowatt amplifiers. This translates into heavy and expensive even when you are buying used.

We tried to come up with something but couldn't find a combination that could vibrate a few pounds up to a few Gs. We ended up trying to get a used unit but these are thousands of dollars. When I left the job they still hadn't bought anything nor paid for testing. The products will pass a basic vibration test but are failing after hundreds of hours of vibration. Rather than spend money they just keep the marketing shuffle going.

I would be interested in anything anyone comes up with even if can only test very light objects (ounces rather than pounds). I expect the amplifier is the biggest challenge but the motor is non-trivial.

At what frequencies did you do the vibration testing in your job?
 

Offline EEVblog

  • Administrator
  • *****
  • Posts: 39271
  • Country: au
    • EEVblog
Re: Building a vibration tester?
« Reply #8 on: August 17, 2012, 04:44:44 am »
This is actually pretty hard to do. We had a vibrating table at my last job and thought it was really torturing our products until we put an accelerometer on it and we couldn't even get 2 G. It was basically an eccentric attached to the platform which was mounted on isolation mounts which allowed it to vibrate in X and Y with wasted energy going into the Z direction.

Usually test platforms for electronics are just Z axis only.
You then test the product by strapping it to the platform in different orientations and doing the required sweeps in each axis.

Dave.
 

Offline Jon Chandler

  • Frequent Contributor
  • **
  • Posts: 539
    • Throw Away PIC
Re: Building a vibration tester?
« Reply #9 on: August 17, 2012, 07:35:13 am »
One problem with as eccentric on a motor is that you have no control of amplitude.  You can't vary the amplitude by changing speeds.  In fact, the force goes up as the square of the speed, so at higher frequencies, the vibration level is much, much greater.

I'm certain you can break boards this way, but trying to quantify the results would be tough.

Ir'm not familiar with the standards off the top of my head, but the requirements may call for constant displacement, constant velocity or constant acceleration over the range of interest.  To approximate these tests, you need to be able to control the frequency and amplitude of vibration.  Since resonances can be high Q, you need pretty tight control of frequency.  If you have jumps in the excitation frequency, you may miss resonances entirely.
 

Offline robrenz

  • Super Contributor
  • ***
  • Posts: 3035
  • Country: us
  • Real Machinist, Wannabe EE
Re: Building a vibration tester?
« Reply #10 on: August 17, 2012, 01:32:53 pm »
Just thinking out loud here:

3HP variable speed router as your power source.  Mine will go 30000 rpm (500Hz) and with some mods to the  speed control will do 1800 rpm (30Hz).  make an adjustable eccentric with its own bearing supports that can do from 0 to .1" radial offset = 0 to .2" PP amplitude. table is supported by parallel flexures to contrain motion to z axis only. Connect the eccentric to the table with a fairly long connecting rod to minimize lateral loading of the flexures. hack the router speed control to be controled by the 50 ohm output of my signal generator so the sweep range and speed is easy to adjust. connect the router to the eccentric/bearing assembly with a flex coupling so the router bearings are not blown out by misalignment. Connect the router and eccentric and base of the flexure system all together. calculate and set the amplitude necessary for the max. acceleration at the max Hz.  Turn it on and hope it all stays together. ;D

HLA-27b

  • Guest
Re: Building a vibration tester?
« Reply #11 on: August 17, 2012, 01:41:00 pm »
 

Offline T4P

  • Super Contributor
  • ***
  • Posts: 3697
  • Country: sg
    • T4P
Re: Building a vibration tester?
« Reply #12 on: August 17, 2012, 02:04:10 pm »
Nice. Is that a steam or compressed air powered machine? I hear pressure ...
 

Offline SeanB

  • Super Contributor
  • ***
  • Posts: 16386
  • Country: za
Re: Building a vibration tester?
« Reply #13 on: August 17, 2012, 04:01:45 pm »
I would guess compressed air, probably via a 100mm feed pipe and some really big air motors. The little ones sold to move powder in a silo can easily tear themselves out of the sheet siding.
 

Offline FenderBender

  • Super Contributor
  • ***
  • Posts: 1115
  • Country: us
    • The Solid State Workshop
Re: Building a vibration tester?
« Reply #14 on: August 17, 2012, 08:28:40 pm »
Would you build yours by unbalancing a motor or by my method?


A electric vibrasander might work for some smaller projects, the one I have has a speed control pot on it.
The big motor I had in mind could probably vibratest a rack mount server though...

That's what they use in cellphones and the like for vibration. Not sure how it would work on a large scale.
 

Online nctnico

  • Super Contributor
  • ***
  • Posts: 28503
  • Country: nl
    • NCT Developments
Re: Building a vibration tester?
« Reply #15 on: August 17, 2012, 10:55:41 pm »
What goes into vibration tests of PCB's?
I have an old motor that I thought about converting into a vibra table. The idea is to transfer the rotational motion of the motor into a linear up and down motion of a rod with a wooden surface attached. By varying the motor speed one can vary how fast the surface moves up and down. I would attach the motor to another shaft by a rubber belt as not to shake the motor.
A motor is a bad idea. To much wear & tear on gears and very little control over the shape of the vibrations. You probably want to use something like this:
http://www.amazon.com/Aura-AST-2B-4-Pro-Bass-Shaker/dp/B0002ZPTBI
There are small lies, big lies and then there is what is on the screen of your oscilloscope.
 

Offline olsenn

  • Frequent Contributor
  • **
  • Posts: 993
Re: Building a vibration tester?
« Reply #16 on: August 18, 2012, 01:44:30 am »
Why build a vibration tester... that's what Lindsay Lohan is there for!
 

Offline EEVblog

  • Administrator
  • *****
  • Posts: 39271
  • Country: au
    • EEVblog
Re: Building a vibration tester?
« Reply #17 on: August 18, 2012, 03:38:28 am »
A motor is a bad idea. To much wear & tear on gears and very little control over the shape of the vibrations. You probably want to use something like this:
http://www.amazon.com/Aura-AST-2B-4-Pro-Bass-Shaker/dp/B0002ZPTBI

We used some of those Aura Interactor shakers at my old company to do lots of vibration type work.
Crude but effective. I didn't know they were still available!
The ones we used looked a lot different though, and possibly had a wider frequency range.

Dave.
« Last Edit: August 18, 2012, 03:46:15 am by EEVblog »
 

Offline bradleytron

  • Regular Contributor
  • *
  • Posts: 74
  • Country: ca
Re: Building a vibration tester?
« Reply #18 on: August 18, 2012, 04:17:36 am »
Not sure how much force you need to inject into the system under test but I helped develop an experimental physics apparatus to study Anderson Localization. We used a high wattage audio speaker, mounted facing upward and adapted by removing the cone and coupling the drive coil assembly to an aluminum plate. If force and the weight of the thing your testing are not issues it makes a cheap and very workable solution and can be driven over a wide range of frequencies.
 

Offline EEVblog

  • Administrator
  • *****
  • Posts: 39271
  • Country: au
    • EEVblog
Re: Building a vibration tester?
« Reply #19 on: August 18, 2012, 08:15:53 am »
We used a high wattage audio speaker, mounted facing upward and adapted by removing the cone and coupling the drive coil assembly to an aluminum plate. If force and the weight of the thing your testing are not issues it makes a cheap and very workable solution and can be driven over a wide range of frequencies.

I've been thinking of doing that, as the commercial shakers are simply more rugged purpose designed speakers with a shaft on them instead of a cone.
How did you secure the plate?
What was the maximum weight and G's you got out of it?

Dave.
 

Offline T4P

  • Super Contributor
  • ***
  • Posts: 3697
  • Country: sg
    • T4P
Re: Building a vibration tester?
« Reply #20 on: August 18, 2012, 04:08:49 pm »
why didn't i think of a vibration speaker!
 

Offline PuterGeek

  • Regular Contributor
  • *
  • Posts: 88
  • Country: us
    • SolutionsPLUS engineering
Re: Building a vibration tester?
« Reply #21 on: August 21, 2012, 05:29:35 am »

At what frequencies did you do the vibration testing in your job?

In theory 5 to 2000 Hz, 10 G minimum would probably work. I say probably because we never isolated the 'problem' frequency.

The product was a instrument panel for an industrial diesel engine powered equipment. The vibration frequencies are very complex with variables such as number of cylinders (mostly 4 and 6 but could be anything from 1 to 12), engine speed (700 to 3500 RPM), driven equipment (mostly pumps) and mounting location. And if all that isn't enough, there were several panel and mounting designs.

We used isolation mounts that mostly worked but without outside testing or equipment I'm not really sure how well they worked. As long as management thinks they can dance around any failures they aren't likely to let any cash go.

And on top of all that, I'm a EE and did most of the mechanical design because I was the closest thing we had to a mechanical engineer. Thankfully I didn't sleep thru my ME classes!
 

Offline PuterGeek

  • Regular Contributor
  • *
  • Posts: 88
  • Country: us
    • SolutionsPLUS engineering
Re: Building a vibration tester?
« Reply #22 on: August 21, 2012, 05:39:43 am »
We used some of those Aura Interactor shakers at my old company to do lots of vibration type work.
Crude but effective. I didn't know they were still available!
The ones we used looked a lot different though, and possibly had a wider frequency range.

Did they have much X-Y deflection?

We had looked at several of those. With the weights we were working with the back of napkin estimation was we needed a lot of them and a serious amplifier for power.

http://www.amazon.com/ButtKicker-BK-LFE-Frequency-Effects-System/dp/B0007P98K6/ref=sr_1_4?ie=UTF8&qid=1345527378&sr=8-4&keywords=Bass+Shaker looks like it might work pretty well and can handle up to 1500 watts. I wonder how much travel it has?
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf