Hi all,

I am trying to design a generator to convert motor vibration energy to electrical energy (around 100mW power required). I want this design can be applied to as many case as possible. But I have no idea in generally how much power I can get out from a vibrating motor and at which frequency ponit(s) should I design the genernator. I think Motor VIBRATION SPECTRUM may be a useful source to estimate.

I am wondering if you guys can give me some vibration spectra or data based on several types of motor, different power rates and/or different running velocity.

Also, any suggestion on how to estimate the amount of the energy/power a vibration source can generate is appreciated. If you have something in mind, please share with me if you want.

Best,

Roger

Typically the dominant frequency of vibration is 1x running speed. Sometimes if there is a foot problem or airgap problem, the dominant frequency will be 2x line frequency. Typical magnitude might be in the range 0.03 to 0.15 inches per second peak/0.

Does 100 mW mean 100 milliwatts? What are you trying to do? Can't you use a battery?

Some clarification is needed for me here. Are you suggesting that you will design a generator that will be excited by the vibration resulting from an improperly mounted motor? The higher levels of vibration from a motor would be generated by some sort of fault either in the mounting or the stator/rotor bars or bearings. And yes, there are dominant frequencies such as 1x and rotor bar pass but the acceleration energy is normally quite low in the order of 2-3g's. So, high 1x can be caused by unbalance, 2xLF--soft foot/variable air gap, stator problems, 6xLF in the case of DC motors with SCR firing cards. And not to mention the wonderful new AC frequency drives that can self destruct without the proper tuning.......

Thanks for your reply. I am brand new in this field, so can you two express your answers in more detail? Actually, our project intends to design a self-powered wireless motor sensor, which is mounted on the motor and transmits the sense singals wirelessly. The energy source might be vibration from a motor in normally mounted condition.


To electricpete: When you say "the dominant frequency of vibration is 1x running speed". What exactly "dominant" means? Is the magnitude of vibration the highest at this particular point of frequency? How about the 3dB bandwith around this frequency? I hope this bandwith is big enough such that the generator can convert enough energy even when its working frequency is slightly mismatch with the "dominat" freq. of the vibraion source.


To vibbase: When you talk about "acceleration energy is in the order of 2-3g" what kind of motor are you talking about? Is 2-3g (1g=9.8 m/sec square) possible when a motor is mounted properly?

Basicaly, I want to identify two things: Can 100+ mili Watts be generated from the normal vibration of a 1hp+ motor? At which point should I design my generator resonant frequency, such that it can harvest enough energy and can cover as much different running speed as possible.

Thanks,

Roger

Dominant frequency means the highest peak in the spectrum. 3db bandwidth is not applicable since the vibration occurs at essentially discrete frequencies for a constant-speed machine.

As you know, you need a component of force out of phase with displacement to deliver non-zero average power.

What kind of generator are you planning to use that will convert linear force and displacement into electrical energy?

A generator that converts vibration into energy already exists... It is called a velocity probe! The old style velocity probes had a spring suspended mass inside a wound coil. The machine causes the mass to vibrate, which excites the coil, giving an output relative to vibration. These things are still used all the time, I have some installed as shift rider probes on a GE turbine. Some of the older siesmic velometers produced as much as 1000mv per in/sec. I never thought about it, but I bet one of these would put out enough to power a black box with a ICP accelerometer and low power wireless eithernet sender... I don't think you could use the velocity probe to determine vibration and powe something with it at the same time... to many losses.. but I could be wrong.
Let's see... a typical smooth motor might vibration at 0.1 in/sec, so with a 1000mv/ips velocity probe, the output would be 100mv. I'm not sure what sort of amperage could be produced, but you need 0.1 watts/0.1V= 1 amp output needed. I doubt very seriously this is achievable. The velocity probe would have to have an electrical resistance of 0.1 ohm to get this type of amps out. Maybe you could use a transformer to step the 100mv up to 5volts with only small losses... this would take some calculating....at the very least you could use it for a battery charger, then use the battery periodically to power the electronics...

This message has been edited. Last edited by: Ed Hudson,

On the theoretical end I was thinking that whatever generator element you use could be modeled as a damping device attached to your mass spring system.

Start out by modeling motor as a lightly-damped mass spring with forcing function provided by an assumed unbalance. If you continue to add damping to the system the vibration will lower but there will be a limit how much power you can extract from the system. If that much energy is less than you need your out of luck. If that much energy is more than you need you still have the challenge of coverting that energy into electricity.

I have another thought. Aren't there devices that can covert heat to electricity. Certainly a thermocouple is one although I'm vaguely remembering Peltier junction might be better suited and is used in several applications for larger scale transfer between heat and electrical energy. Can the heat energy or temperature difference from motor skin to ambient be used to provide electrical output?

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