Problem on Radio Energy Harvesting circuit- Is Cockcroft Walton circuit fail?

[allhksar]

Dear all,

I am building up a radio harvesting circuit on WIFI signal, but I fail. Actually I am very new to antenna and RF aspects.I have several questions about the circuit.

First of all, I follow this schematic diagram from this article - http://www.hindawi.com/journals/apec/2010/591640/

Equipment that I have used:
6X 1N5819 DO-41 1A/40V diodes
17X 1pf ceramic capacitors
5X 0.1uH 0307 Axial-Lead Inductor
1X 2DBI 2.4-2.5GHz wifi pcb internal ipex antenna , like this:

1X UNI-T UT33C DMM

C1 is made up of 11X 1pf capacitor in series.
L1 is made up of 5x 0.1uH inductor in parallel.
C2=C3=C4=C5=C6=C7=1pf capacitor


The result I have encountered:

When I place the antenna vary the distance between 1cm to 3meters in any direction from my home wifi router, zero volt across Test point 3 with ground point , even zero volt test point 1,2 with ground point measured by the DMM.


I doubt at some area that I may misunderstand:

1. What is MS in the schematic diagram that illustrated in the research ?
2. Do I use wrong type of capacitor or diode or inductors? Since I have found some capacitors, inductors and diodes that specifically used in RF.
3. Do I need earth the antenna ? Actually, the ground point of antenna I have connected to the earth poin that near to C7 which is also connected to DMM negative(COM) pin. And I just follow this IPEX instruction:

4. Is the antenna too weak ? Since I have only 2DBI gain of antenna, however, I wonder why I even can not get 0.1mV when I place the antenna very very close to the home wifi router.



Apart from this, a very strange phenomenon occur in this circuit:
When I change the diodes to 1N4004 and keep other equipment unchanged, and placing the antenna near to a charging mobile phone or wifi router (on top of the device), there is a reading, like 30-50 mV shown in DMM.
I wonder is this reading due to the AC charging current?

I hope someone can help me.
Thanks all

 

[Gryd3]

Let's stop you here before you go any further.
What you are doing is not going to give you anything useful. At all, even remotely.

Think of it this way. If your Wireless router can broadcast at an amazing 1 Watt (yeah right), then the amount of energy you can harvest from it will be a function of surface area and distance from the antenna. So... if you manage you put a large bowl over top of the antenna to capture half of the emitted signal, then you get 500mW ... Great! well... considering how close you have to be... not so great.
Also, as soon as you start to add distance, that number will get pretty damn tiny very very fast...
Double your distance and you can now only harvest 125mW. Double your distance again, and that falls to 31mW ... talk about sad, and keep in mind this distance is probably only 1 or 2 feet depending on the size of bowl you started with.
Also keep in mind that the area behind your harvester will now receive no emissions from the Wireless Access Point...
Anyway. Let's proceed again shall we?
A diode has a voltage drop... if the voltage on the antenna does not surpass this voltage, you get NO power. If the voltage does... then well, you just lost a bunch of power. Dang!
To make matters worse, there is more emissions than just that one WiFi access point you are trying to harvest from, and as the RF intermingle, you will find that very complex waveforms emerge which you can't 'harvest' .
You're out of luck, and this is not the first time it's been tried. If you want to wirelessly transmit power, please look into Tesla's work, or into 'induction'

 

[allhksar]

Let's stop you here before you go any further.
What you are doing is not going to give you anything useful. At all, even remotely.

Think of it this way. If your Wireless router can broadcast at an amazing 1 Watt (yeah right), then the amount of energy you can harvest from it will be a function of surface area and distance from the antenna. So... if you manage you put a large bowl over top of the antenna to capture half of the emitted signal, then you get 500mW ... Great! well... considering how close you have to be... not so great.
Also, as soon as you start to add distance, that number will get pretty damn tiny very very fast...
Double your distance and you can now only harvest 125mW. Double your distance again, and that falls to 31mW ... talk about sad, and keep in mind this distance is probably only 1 or 2 feet depending on the size of bowl you started with.
Also keep in mind that the area behind your harvester will now receive no emissions from the Wireless Access Point...
Anyway. Let's proceed again shall we?
A diode has a voltage drop... if the voltage on the antenna does not surpass this voltage, you get NO power. If the voltage does... then well, you just lost a bunch of power. Dang!
To make matters worse, there is more emissions than just that one WiFi access point you are trying to harvest from, and as the RF intermingle, you will find that very complex waveforms emerge which you can't 'harvest' .
You're out of luck, and this is not the first time it's been tried. If you want to wirelessly transmit power, please look into Tesla's work, or into 'induction'

Thanks for your reply, but why this research works ? They have success to harvest energy to drive a calculator!!!
FYI - http://www.hindawi.com/journals/apec/2010/591640/

As you mention diode forwarding voltage, but to my knowledge, this kind of voltage is varying to the current...In my case, the current should be very small, which means the forwarding voltage is small.. Am I right ?

In the article, they claim they can obtain a 0.015V to 0.4 V , but I never get this value... only zero reading..

 

[Gryd3]

Thanks for your reply, but why this research works ? They have success to harvest energy to drive a calculator!!!
FYI - http://www.hindawi.com/journals/apec/2010/591640/

As you mention diode forwarding voltage, but to my knowledge, this kind of voltage is varying to the current...In my case, the current should be very small, which means the forwarding voltage is small.. Am I right ?

In the article, they claim they can obtain a 0.015V to 0.4 V , but I never get this value... only zero reading..

This article did not disclose the output 'power' of the base station or Access Point they were using. Granted. I didn't read the whole thing.
Voltage, current, and power are all directly proportional to each other.
So if you collected 5mW , that means that your device would only provide 1mA @ 5V or 1.6mA @ 3V

 

[allhksar]

This article did not disclose the output 'power' of the base station or Access Point they were using. Granted. I didn't read the whole thing.
Voltage, current, and power are all directly proportional to each other.
So if you collected 5mW , that means that your device would only provide 1mA @ 5V or 1.6mA @ 3V

They mention about the power...
They have plotted a graph about received power against output voltage. In my case, I use mobile phone to measure received power about -23dBm, according to this graph, I should get around 0.1-0.2 V, however I get 0V..

 

[AnalogKid]

You are using power diodes, which have a relatively large capacitance, slow response time, and large Vf. Change to Shottkey signal diodes. To rectify an ac signal at 2.4 GHz frequency, you need exceptional diodes, ones with very low junction capacitance and fast response times.

ak

 

[allhksar]

You are using power diodes, which have a relatively large capacitance, slow response time, and large Vf. Change to Shottkey signal diodes. To rectify an ac signal at 2.4 GHz frequency, you need exceptional diodes, ones with very low junction capacitance and fast response times.

ak

I have changed the diode to BAT85 , even BAP64-02 (rf diode), I get the following strange result:
when I move the antenna near the charging mobile phone , I get the voltage up to 80-100mV-(even i turn off wifi), when the mobile is not charging, the reading is 0.1mV. I wonder where does the energy come from ? My antenna is marked as 2.4GHz wifi 2DBI antenna

 

[Alec_t]

If your phone has a switched-mode charging circuit your antenna is probably picking up RF hash from that. It may well be marked 2.4GHz, but it will pick up all sorts of frequencies (with varying efficiency).

 

[davenn]

I really think you don't understand how tiny a power level this is
... Gryd3 summed it up nicely in his opening comment in his first post

Let's stop you here before you go any further.
What you are doing is not going to give you anything useful. At all, even remotely.

do you really understand how small a power level -23 dBm is ?

that is 0.005 mW

They mention about the power...
They have plotted a graph about received power against output voltage. In my case, I use mobile phone to measure received power about -23dBm, according to this graph, I should get around 0.1-0.2 V, however I get 0V..

 

[Gryd3]

I really think you don't understand how tiny a power level this is
... Gryd3 summed it up nicely in his opening comment in his first post




do you really understand how small a power level -23 dBm is ?

that is 0.005 mW

As a comparison. The 0.005 mW Dave posted would mean that you would need to be able to store this energy. The tricky part here is that the 0.005 mW will get smaller as it deals with components in your device 'before' the storage.
To put that scale to something useful... a 3.7V 20mA white LED is going to require 74mW to light.
That will require 14,800 seconds of charging at 100% efficiency to light the LED for 1 second.
The other problem is that the storage device you are using will suffer from something called 'self-discharge', and if you need to pass this energy through a converter or charge controller of any sort, you will loose the energy.
At some point, the energy is just too small to harvest to be able to do any work with.

 

[d r jones]

Thanks for your reply, but why this research works ? They have success to harvest energy to drive a calculator!!!
FYI - http://www.hindawi.com/journals/apec/2010/591640/

As you mention diode forwarding voltage, but to my knowledge, this kind of voltage is varying to the current...In my case, the current should be very small, which means the forwarding voltage is small.. Am I right ?

In the article, they claim they can obtain a 0.015V to 0.4 V , but I never get this value... only zero reading..

hai - looked up the spec of the 1n5189 - it has self capacitance of 110 pF - redraw circuit replacing all diodes with capacitance and from that estimate what happens , the Cockkroft-Walton power generator for high voltage low current from high current low voltage

Its a Silicon diode with forwards drop of 0v6 - no matter if you use RF signal diodes those still have 0v6 . Original circit fro UAE does not specify which diodes . Get some Germanium diodes forward volts drop around 0v2 - even so the power is going to be very very small . Its attractive to consider the wasted power of radios - but impracticle . Not knowing how many radio stations in UAE - there might be only a few - whilst Hong Kong - suspect many which will exacabate the problem .

But its fun trying - you will have other ideas that will work

 

[hevans1944]

@allhksar You are correct that the forward voltage drop of a diode is a function of current. If you look at the V-I curve for any diode the forward voltage goes through zero at the same time the current goes through zero. The reverse current is small (but never zero) as the reverse voltage starts at zero and becomes increasingly larger, becoming a constant reverse leakage current until either zener or (more likely) avalanche breakdown occurs. Given these characteristics any diode should certainly act as a rectifier, even for microvolt level signals. If this were not true, it would be impossible to build a crystal-radio receiver, something every budding electronics enthusiast should do at least once in their life!

However, the V-I curves reflect the diode behavior under steady-state DC conditions. Add in a gigahertz or so of AC signal and other considerations enter the picture, the most important being junction capacitance (which acts as a reactance in parallel with the diode) and reverse leakage current (which acts as a constant-current discharge path in parallel with the diode). Together, these two factors limit how much DC "power" you can extract from the diode-rectified AC signal. It just gets worse when you add voltage multiplier stages and you reach a point of diminishing returns after three or four well-designed stages.

Another consideration is your antenna. The authors of the paper you cited constructed a double-sided microstrip dipole antenna for GSM band without giving much detail. Okay, they didn't give ANY detail, but a later image (Figure 14) shows a larger dipole antenna (perhaps for a lower frequency band?) connected to what appears to be a strip transmission line. The authors also mention using series-tuned LC filters, and one such is shown on their schematic in Figure 8, but they were similarly lacking in details, although the left side of Figure 13 appears to be involved with this. Note also that the voltage multiplier appears to be constructed with SMD components, probably to help control parasitic capacitance. When working at gigahertz frequencies, circuit layout and construction are critical to obtaining expected results.

I don't see an antenna in any of the images you have posted so far. Are you using a monopole (rubber ducky) type antenna? Sounds like it, if the "gain" is just 2 dbi. You should try using a dipole antenna cut to resonate with the transmission frequency of your WiFi wireless router.

Personally, as @Gryd3 stated in his original post #2, I would give up on this and go find something more useful to spend time on. AFAIK, the ONLY practical application of wireless power harvesting is for powering remote sensors. Someone correct me if I am wrong about this and there is another "killer app" out there waiting to suck energy from the ether. If you can't master this "simple" power harvesting project, there is no way you are prepared to create remote sensors that live on picowatts of power and spend most of their time consuming almost no power at all.

But have fun trying. Maybe you are this generation's next Tesla!

Edit: BTW, MS on the schematic stands for Micro Strip.