search

Tactile-electronics / windmill-electronics

Creative Commons Zero v1.0 Universal
0 stars 0 forks source link

Design circuit for CN3791 charging chip #2

Closed amcewen closed 2 years ago

amcewen commented 2 years ago

Lay out the schematic for the charging part of the circuit

amcewen commented 2 years ago

Notes from reading through the datasheet...

the charge current is set by the external sense resistor RCS and an internal 120mV reference, the charge current equals to 120mV/RCS

We need to decide what current we should (at a maximum) charge the battery at. There are recommendations based on the size of the battery, so it'll probably be a component that gets tuned based on the size of battery it's likely to use. Once we've chosen the likely charge current we can work out the inductor to use — need to check the datasheet again for details of that.

Not quite sure how to choose the MPPT voltage (the Maximum Power Point Tracking, which tries to optimise the power from the solar panel). However, once we work it out...

The maximum power point voltage is decided by the following equation: VMPPT ＝1.205×(1＋R3／R4)

We'll need to add...

P-channel MOSFET with Rds(on) in the range of 30mΩ

Diodes should handle the expected current, but shouldn't be lots over that, as that'll lose some efficiency.

We could (see Fig 3 in the datasheet) switch out D1 and use another MOSfet for a bit more efficiency, but that'll be a bit more expensive (MOSfet, capacitor and resistor instead of a single diode)

amcewen commented 2 years ago

From this note on charging Li-Ion and Li-Po batteries:

Constant current charge: In the first stage of charging a li-ion battery or cell, the charge current is controlled. Typically this will be between 0.5 and 1.0 C. (NB: for a 2 000 mAh battery the charge rate would be 2 000 mA for a charge rate of 1C).

For consumer based LCO cells and batteries, a charge rate of a maximum of 0.8C is recommended.

A single 18650 Li-Ion battery (at least the one I've got here on my desk) is 2000mAh. For the bike lights we've done for Peloton we're using 4x18650 packs that are 8800mAh.

I doubt we'll get anything like 2A of charging from a solar panel or windmill, so it's likely a bit academic, but if we go for a 0.8C of a 2000mAh battery that would be a maximum charge rate of 1.6A.

So Rcs would be calculated as 120mV /Ich. Assuming the current would also be measured in mA for that, we'd get 0.066 ohms for Rcs. 0.068 ohms seems a more common value on Farnell, which would give us a maximum of ~1.76A charge current.

Now that we've got a value for Ich we can look at the inductor. The datasheet says that 0.3 x Ich is a reasonable starting point for the inductors ripple current, so that would be ~0.528A. Combined with the expected charge current of 1.76A we'd want a peak current rating of ~2.3A for the inductor.

The inductance should be greater than 5 x Vcc — Vbat. With the windmill power we might get Vcc of 20V, and a Vbat could be as low as 3.5V, so an inductance of 100uH would seem reasonable.

amcewen commented 2 years ago

Apart from setting the Vmppt value, which I've spun out as #3, the schematic is done with the components all chosen. Next step is to start arranging the actual PCB.