Asking for piezoelectric bulk energy harvesting 2

[bp786]

Hi all,

I' m new in this forum. Greeting to all.

I'm doing a project of energy harvesting from piezoelectric bulk type by force impact input. The output of the device when no load is a pulse like with amplitude 60-70 V.peak and pulse width about 500 us.
The standard technique of full bridge rectifier with parallel storage capacitor, the voltage across the cap. gets 20-30 mV(storage cap: 220-470 uF) at the first impact and rate decreases thereafter.

I would like to ask for any idea to get the storage energy more efficiently.

PS. The purpose is to store the harvested energy into a capacitor and give to a DC-DC converter for charging a battery. The energy sources brought from piezoelectric together with electromagnetic induction which now gets quite satisfactory energy output.

Can anyone help me on the piezoelectric energy harvesting issue? I also attach the photo and primary experiment.

Thank you.
Bundit

 

[OperaHouse]

Kinda rude asking us to create an account just to see your picture.

 

[LiteYear]

Thinking on the run here, I suspect the basic diode->cap arrangement is probably the most effective, direct way. That the rate decreases as voltage rises should be expected since energy is proportional to the square of the voltage on the cap. Now to make it more efficient, I think the best way would be to optimally load the piezo - that is, find the peak power curve of the piezo and use PWM into the cap to hold the current at that peak level. Similar to a maximum power point tracker on a solar panel. The piezo will hit maximum voltage with no load and minimum voltage with a short circuit. Somewhere between those two will be the maximum power point and it will change over the cycle of the piezo.

Have you done a literature survey? Surely there's some research data out there. There are already practical implementations of the concept.

 

[DRWeig]

When I click your link, I get my own DropBox folder. I suspect that others who already subscribe to DropBox will get the same. There's a specific method in DropBox to link a file... You might try uploading your pix to engineering.com as described in the "Attachment" area below the reply-to-thread box at the bottom.

Best to you,

Goober Dave

Haven't see the forum policies? Do so now: Forum Policies

 

[IRstuff]

I think you need to describe more fully how the piezoelectric component is stimulated. What are the characteristics of the "impact?" Knowing this may reveal some customizable approach that would work for this specific application.

TTFN
faq731-376

Need help writing a question or understanding a reply? forum1529

 

[bp786]

I have attached file to engineer.com.

 

[ijl]

My guess (educated or not) is that your capacitor is way too large. The capacitor should be so small, that the voltage over it is about half of the produced voltage, otherwise much energy gets lost. You could then charge a larger capacitor from this small capacitor with a (very) small current using a switching mode power supply.

 

[bp786]

#IRstuff - The 'impact' in this work uses a metal bar to hit the piezo periodically as an impact force. The output waveform of the hit force is a pulse-like(with positive and negative amplitude).

 

[IRstuff]

Schematics? pulse durations, repetitions, repeatability? People shouldn't have to be dragging this information out of you.

TTFN
faq731-376

Need help writing a question or understanding a reply? forum1529

 

[VEBill]

How many microwatts (averaged over your choice of minutes) does this banging piezo system produce? This calculation will need to include the repetition rate (the numbers of mechanical strikes per unit time).

Depending on this answer, it may be completely impractical to use a switch mode power supply to increase the voltage. In fact, the entire system may be completely impractical.

 

[bp786]

#VE1BLL - The tested is conducted by connecting full wave bridge rectifier with parallel storage capacitor, then measure DC voltage across the capacitor each time the piezo is hit. The calculated storage energy (W=0.5*C*sqr(dV)) gets 3 uJ at the first hit for capacitor 10 uF which is the most available DC level as I tested varying the capacitance. The accumulated energy for this cap. reaches about 5 uJ for 10 hits(3 second/hit interval), the almost energy gets from the first 3-4 hits for all capacitors variation.
Do you have the practical idea to transfer this charges to accumulate with another energy generator,magnetic induction which produces energy 8000-9000 uJ for a sequential excitation of the system movement?

Thank you.

 

[ijl]

The energy of a capacitor is W = 0.5*C*V^2 (square, not square root). If the energy in a 10 uF capacitor is 3 uJ, then the voltage over the capacitor is about 0.77V. If you put the same electrical charge in a capacitor of 0.2uF, for example, then the voltage is about 39V (charge = voltage times capacitance). The energy in the 0.2uF capacitor is 0.15mJ, 50 times the original energy. Something like this should be possible, because your piezoelectic element produces 60-70V.

 

[VEBill]

"...about 5 uJ for 10 hits (3 second/hit interval)... ...to accumulate...8000-9000uJ..."

For 5 uJ (10 hits), but you need 9000 uJ (x1800), so you'd need 18,000 hits. At 3 seconds between hits, that's at least 15 hours of banging away. And that's assuming 100% efficiency and zero losses referenced to your 5 uJ reference point.

9000 uJ is .009 watt-seconds. Once every 15+ hours (if we make false assumptions).

I'd use a battery.