Inductive Charging Circuit Design Help

[DPG]

It does not define the min. input voltage for the 7805 specifically, it just gave a range for all 78-- models: 5v - 18v. My input voltage (when in optimal range of transmitter) is 5V at around 2 amps

Ah sorry, I think I've misunderstood. I thought you were regulating down to 5V from a higher voltage. I don't think I've read the thread properly.

 

[SparkyChick]

The LM7805 from Fairchild has a dropout voltage of 2v. What this means for you is that the minimum input voltage to get a stable 5v output is 7v. This rises as the load increases. This is why I specifically mentioned low dropout regulators... these have a much small dropout voltage and may be suitable, but it's entirely possible you would need to look at alternative solutions.

One possible alternative is using a zener diode to limit the voltage. Google something like "zener diode voltage regulator" or "zener diode voltage limiter".

 

[Lucien Nunes]

A lot of randomness here.

1. We need to understand more about the output of the inductive charger receiver module. You originally mentioned connecting the battery to the coil, but presumably meant to the output of the receiver module. That has its own voltage converter by the look of things, but the info is not very clear about what it outputs. 5V 2A to 12V 700mA, or something. What actually comes out... a regulated voltage? A voltage that is always 12V off-load but falls with increasing load? Before trying to connect any drones or batteries I would begin by trying various resistive loads and seeing what it delivers. Personally I would plot a curve.

2. Having discovered the nature of the output, then we can decide what additional converter, if any, is needed between its output and the drone charging input. I would definitely avoid using 7805 or similar linear regulator, as that could end up wasting over 50% of the limited available power as heat and decreasing the maximum charging rate as a result. A switching buck 3-terminal regulator such as a Traco TSR2-2450 might be a better choice, which is functionally equivalent to a 3-terminal 5V 2A linear but with very little power loss and heat dissipation. But as yet it is too early to say whether this type of regulator is suitable.

3. Alternatively we can look at the option of bypassing the drone's charging circuit and charging the battery directly.

But first things first, characterise the output of the inductive coupler output module...

 

[SparkyChick]

@ryanrocket ... see what I mean about someone with more capabilities than me

 

[Lucien Nunes]

I am not sure what you mean by:

Battery requirements: 3-12V at 2-3 amperes

The battery appears to be a 2-cell Li-Ion with a nominal 8.4V charging requirement. Less than 7.4V it won't charge at all, 10V or more it will probably overheat and be damaged. If you are referring to the Mavic charger input, that might be specced for USB-C flavoured voltage options and probably isn't designed to go as low as 3V.

Going back to my point about a linear regulator being wasteful, just take a look at the conversion stages that the power will be going through before it gets to spin the drone motors. For argument's sake I'll take switching stages as 90% efficient, battery 85%:
PSU: 230V AC to 12VDC, 90% out.
Transmitter: 12V DC to HF AC, 81% out.
Receiver: HF AC to 12V? DC, 73% out.
Linear regulator: 12V to 5V, 30% out.
Battery charger: 5V to ~8.4V, 27% out.
Battery charge/discharge efficiency: 23% out.
Motor driver: 21% out.

21% gets to the motors, 79% wasted as heat, of which the linear (e.g. 7805) regulator is responsible for 43%. In real power applications design we try at all costs to avoid this daisy-chain of conversion stages. Ideally you want the inductive receiver to charge the battery directly, but that requires custom electronics.

 

[ryanrocket]

@Lucien Nunes

Ok so I just conducted some testing to find the output of the receiver module to be 5 volts when the 2 coils are practically touching. Of course, the output of the receiver coil directly affects the voltage it will be able to provide. For the sake of this, lets just assume it's at an optimal distance and is outputting 5V.

As for the drone battery; the battery is labeled to charge at that specification, however after a buddy and I used a bench PSU to test this, we found that the drone can actually charge at anywhere within the range of 3 volts @ ~3 amps to 12 volts @ ~1.5 amps, where the required amperage is inversely related (as voltage decreased, amperage needs to increase).

Though I am not exactly sure what the PCB on the receiver module is doing (due to both my lack of knowledge and the lack of documentation), that's where I am getting my 5V DC from. I need to figure out what to put in between that receiver circuit and the drone's battery circuit to prevent it from the aforementioned effects.

 

[Lucien Nunes]

the receiver coil is able to pick up 5V at 2 amperes when in an optimal range.

the drone can actually charge at anywhere within the range of 3 volts @ ~3 amps to 12 volts @ ~1.5 amps, where the required amperage is inversely related (as voltage decreased, amperage needs to increase).

What is probably happening now is that the two are interacting like a relaxation oscillator. Receiver outputs 5V; drone tries to take >2A; voltage decreases; drone takes more current (this is positive feedback at work); voltage collapses to a point where the drone stops charging; voltage recovers; cycle repeats. Watching the voltage on a scope would show the behaviour clearly. The underlying cause is that the inductive coupler modules and coils are not powerful enough for the load of the drone charger.

Your two inital bullet points were actually pretty well on the money, but I would add a third:

• Reduce the drone's current demand at 5V so that it's within the receiver's ability to supply.
• step down the 5V with minimal loss so that there is enough increase in current to satisfy the drone's demand at a lower voltage.
• Get an inductive coupler system that is powerful enough for the job.

Let's suppose the last one is not possible for now. For my own purposes I would begin by looking at the first option, by reverse-engineering the drone's charging circuit. Depending on the design, it might be possible to increase the battery current sensing resistor or alter the associated feedback network to make it charge at a lower rate. This would mean the charger IC gets a false impression of what is going on, which can impact on the termination of the charge. If the current sense resistor is shared by the discharge monitoring circuit, then it must not be changed or it will also throttle the drone power. I can't honestly recommend going this route if you don't have a good knowledge of battery charging circuits and methodology.

That leaves us option 2: step down the receiver output to a voltage at which the drone input power is lower. The receiver is offering us a maximum of 5 x 2 = 10W. The drone wants anywhere between 3 x 3 = 9W and 12 x 1.5 = 18W. So there is a possibility of a precarious stable zone at 3V if we can do the step-down at an efficiency of 9 / 10 = 90%, which is just about possible with an off-the-shelf buck converter.

I have to go now but you might like to have a look for the most efficient stepdown converter or switching regulator with 3.3V output that will work on 5V input. Obviously a linear regulator won't work as that reduces the voltage without increasing the current. It's a moot point whether even a switching regulator is going to work. It's so very close to there being not enough power at all, and then you'll have three active circuits interacting and possibly fighting each other. I really can't predict what will happen although technically it could work.

You might just like to try a trick first, of putting say 470μF in parallel with the receiver output. If there is another metastable state where the load lines of the converter output and the drone input cross, it might settle there if the time constant is drastically changed. I accept no responsibility for the possible outcome of the receiver blowing up (if it goes into more violent oscillation.)

 

[Lucien Nunes]

I'll add that you might also get some joy from adding a small amount of series resistance. Because the drone input current increases with decreasing voltage, it exhibits negative dynamic resistance. It's not a true negative resistance and is probably far from linear, and we know it's not a constant power input, so there might be spots where the negative resistance is not very high. If you add enough real resistance in series to make the net resistance positive, you might find a stable operating point. A few hundred milliohms might be worth a try.

It's a total kludge and probably won't work, but apart from modifying the drone charger or using a larger inductive coupling setup, any answer is likely to be a kludge.

 

[ryanrocket]

I'll add that you might also get some joy from adding a small amount of series resistance. Because the drone input current increases with decreasing voltage, it exhibits negative dynamic resistance. It's not a true negative resistance and is probably far from linear, and we know it's not a constant power input, so there might be spots where the negative resistance is not very high. If you add enough real resistance in series to make the net resistance positive, you might find a stable operating point. A few hundred milliohms might be worth a try.

It's a total kludge and probably won't work, but apart from modifying the drone charger or using a larger inductive coupling setup, any answer is likely to be a kludge.

Alright. After some in-depth testing, we got the drone to charge using the buck converter, capacitor, and a new micro USB cable. It used 3.7v at like 2 amps. Thanks for the guidance and help! Now to print the PCB's.