Inductive NiMH charger

[Doktor Jones]

Hi... I've seen a few threads about this sort of thing before, but inductance/coils/etc is all new to me (other than using pre-made transformers), as well as battery charging. Anyways, I want to make a charging "pad" for a device with no wires/connectors. I know the Palm Touchstone and SoniCare toothbrush use inductive coils to transmit power, and I'm figuring I should be able to rig up something similar.

The device is fairly large, so I can have up to about a 3" coil in it (maybe 3.5"), and has two AA batteries in it; currently the batteries are just alkalines, but for this retrofit I'd be putting NiMHs in. It's very low drain (<1mA) and would be able to sit on its charging pad for ~6 hours a day, so I wouldn't need much current for the charging -- even 10mA would be plenty. I'd also like to just maintain a float charge if possible, to minimize charging circuitry in the device.

The only restriction I'm placing on the base is no PICs or other programmable microcontrollers, because I don't have (or want to devote) the time and money into it (since I don't have a programmer, and don't have any desire to learn to program them right now).

From what I've read around the forums here, it seems I have the option of just using the 60Hz line frequency for the charging coil, or applying any variety of oscillators to obtain a more efficient charging circuit via higher frequencies. Would a 555 timer be suitable for this? If my calculations don't fail me, with C1 = 1000pF, R1 = 1kΩ, R2 = 22kΩ I would get a ~32KHz square wave; would this output drive the coil sufficiently via a 2N3904 to provide the charging current I desire.

Once I have the charging pad down, it's on to the battery charging circuitry! Once again, a new area to me, and as I stated before, I want to keep it to a minimum of components.

Thanks for any input!

 

[ronv]

I've been thinking about the same thing for my wireless mouse. Here is a nice write up.

http://www.discovercircuits.com/dc-mag/Issue-1/issue-1-receiver-construction.htm

 

[Doktor Jones]

That site is very interesting and rather informative, but unfortunately a lot of the formulas he was referencing were over my head; also, will I need to measure inductance when tuning? I don't think my meter does inductance.

As for the charger circuit, would a LM317L or LM117H work as suggested here? (I'm looking at TO-92 or TO-39 packages, because I don't plan on exceeding 100mA, and the smaller/lighter-weight I can make the charging circuit, the better)

 

[Doktor Jones]

For ease of reference, here is the 50mA constant current charger in the LM117/317 datasheet. This should work for my needs, no?

 

[Dragon Tamer]

I've been thinking about the same thing for my wireless mouse. Here is a nice write up.

A wireless mouse that gets it's power from the mouse pad... BRILIANT!

I like the link that ronv posted but if you are trying to run the circuit at 60Hz, then you are going to run into some problems. The schematic in the link has a resonent frequency of about 125.8 kHz, your 60Hz signal would do nothing to it. That's not to say it can't be done, you will just need to use a larger cap or inductor to change the resonent frequency to 60Hz.

If you want to try to do this with a 555 then you are going to need some way to boost the current output of the timer. Wireless charging is a lot less efficeint than using a regular transformer. The current that the source needs to provide will be a good bit higher than what the load is actually consuming. A fair bit of power is going to be lost in the transfer. I can't remember what the max current output on the 555 is but I don't think that it is more than 50mA. When you use the 555 as the source of the oscillator then you want to try to ensure that the duty cycle is 50% for the best result.

I've never heard of measuring the inductance of a coil while it is being wound. But it is very common practice for people to make their own inductors. There is a formula that you can use to figure out how many turns you need to get to a certain inductance. Besides most multimeters don't do inductance, that takes an inductance meter .

For your charger circuit, I myself try to stay away from simple chargers but to maintain the simplicity of the circuit, I'm going to say go for it. I would acutall try the 50mA constant curren source first on a breadboard with a normal power supply, and simply check that it is working properly. The only thing that worries me is that it is a 50mA current source, it has no way of controlling it's output to maintain a steady voltage. You may need to drop in a zener diode with a value slightly higher than the voltage of the battery to prevent over charging.

 

[alec_t]

The only thing that worries me is that it is a 50mA current source, it has no way of controlling it's output to maintain a steady voltage. You may need to drop in a zener diode with a value slightly higher than the voltage of the battery to prevent over charging.

Constant-current charging is finefor a NiMH battery and over-charging won't be a problem if the current is well-below the recommended trickle-charge limit for the particular battery used. IMO a zener diode is of dubious value, as the battery voltage when fully charged remains almost constant with further charging.

 

[Doktor Jones]

If you want to try to do this with a 555 then you are going to need some way to boost the current output of the timer. Wireless charging is a lot less efficeint than using a regular transformer. The current that the source needs to provide will be a good bit higher than what the load is actually consuming. A fair bit of power is going to be lost in the transfer. I can't remember what the max current output on the 555 is but I don't think that it is more than 50mA. When you use the 555 as the source of the oscillator then you want to try to ensure that the duty cycle is 50% for the best result.

Well that's why I was talking about driving it via 2N3904; I don't really have any size constraints on the base unit, so I could use a heatsinked 2N3055 if need be. With R2 significantly greater than R1, the duty cycle should be ~50.5-53%.

What formula would I need for determining the capacitor and inductor values for a desired resonant frequency (i.e. if I'm driving the exciter coil @ 75kHz, how do I figure out what value cap and inductor I'll need)?

Also, is there any functional difference between making the coil round or making it square (as shown on that site)?

 

[ronv]

Here are some links to a resonance calculator and inductor calculator. You will need to "tune" them once you get close. You can probably pick values for the transmitter and change the frequency of your timer to get resonance. Then you will need some smaller caps for the receiver to tune it in. Do you have a scope by chance?

https://www.daycounter.com/Calculators/LC-Resonance-Calculator.phtml

https://www.daycounter.com/Calculators/Air-Core-Inductor-Calculator.phtml

 

[Dragon Tamer]

If you want to get the most voltage out of the circuit as possible, you need to consider the resonant frequency of the cap and inductor in your receiving unit. The resonant frequency is exactly the same as the formula for a tank circuit since that is basically what we have here.

The formula for a tank circuit is:
f=1/(2*pi*√L*C)

Using a bit of logic, we can reverse this to solve for the cap or inductor given a certain frequency. That equation is:
L^2=(1/f)/2/pi/√C
or for solving for the cap:
C^2=(1/f)/2/pi/√L

So let's say you have a 60Hz signal and an inductor of 27.8mH, the cap that you would need for that circuit would be C^2=(1/60)/2/pi/0.167 (0.167=√0.0278)
the result of this is 0.01588 once we square this we get 0.0000252F or 25.2uF.

After you solve for your cap or inductor, run the numbers back through the equation for resonant frequency and if your frequency is within 10% to 20% of your actual frequency then you have the right value.