One of the keys to a fully wireless J-Box is establishing the efficacy of harvesting enough RF from the antenna transmit signal to power everything – the wireless command/control circuitry and the RF relay.
With parts on order, I decided to do some design work and simulations as preparation last night. It’s been difficult finding a simulation tool capable of doing exactly what I need with RF, without spending thousands of dollars. I’m getting by with the free TINA-TI version for now.
Here’s the initial design and simulation. I used this excellent thesis paper entitled “WIRELESS BATTERY CHARGING SYSTEM USING RADIO FREQUENCY ENERGY HARVESTING” as the basis of the voltage doubler for the overall RF energy harvesting design below.
(click image to zoom in)
In the schematic above, we see a signal generator that’s attempting to simulate the incoming RF signal on the antenna, along with an inductive RF coupler with about 50 microhenries of coupling. The harvested signal is then passed through a voltage doubler circuit consisting of Schottky diodes and capacitors, along with C3 which serves as the “battery” and stores the rectified current. A couple of zener diodes are used to regulate the voltage down to usable levels for 12 and 3 volt circuits. Schotkky diodes were chosen due to their relatively low forward voltage rating of around 0.35 volts (just 350 millivolts), so even with very low signal strengths it should be possible to harvest RF power.
The 470uF capacitor was chosen to hold more than enough power to drive the relay and circuitry for a brief period of time. Tripping the 12 V latching relay should take no more than 5 ms. It will be interesting to see if the relay can be operated at this low power level. I’m thinking a separate capacitor will actually be needed to power the circuitry – one that’s not competing for power with the relay.
The low-power circuits being evaluated operate at less than 10 ma. during full transmit and consume 100’s of microamps otherwise, so they should run fine on much less power. It’s likely the relay’s power consumption that’s going to make or break this design.
In the simulation run shown above, I used 60 cycle AC as the “RF” input signal as the TINA simulation program isn’t efficient when dealing with high-frequency RF signal processing in the Mhz range (it’s VERY slow). In other simulation runs, the signal generator was set to 14 Mhz to simulate a 20 meter transmit signal and the same level of energy is harvested in a matter of milliseconds (the simulation takes minutes to run). Since I was able to prove the simulated results were similar in both cases, I reverted back to using 60 Hz. with TINA to make working with the tool faster.
So what level of input voltages are being assumed? For this simulation, I’m assuming 10 volts of RF. The J-Box will be located near the end of the element, where the resistance is highest (est. 2K – 5K). With a given level of transmitter output power, assuming a reasonably well matched antenna, it should be possible to calculate the voltage levels near the ends of the antenna elements.
So Voltage is equal to square root of Power times Resistance. Let’s assume R is 2K for now.
At 1 watt, E ~= 44 volts
100 watts, E ~= 447 volts
1500 watts, E ~= 1,732 volts
So even at QRP levels of just one watt, it should be possible to acquire enough voltage near the ends of the antenna elements where the J-Box sits.
With parts coming in soon, should be able to test this theory and find out for certain. You can download the TINA circuit design below.