I’m copying this battery charger schematic over from the sparkfun shield into a project I’m working on. They added a schottky diode (D3) that isn’t in the reference schematic for the MCP73831. I have some ideas why they added this diode, but I don’t know enough to be certain. Could someone help clarify this mystery for me?
P.S. While we’re at it, D1 is just there for reverse voltage protection? If so, did they choose a schottky for the low voltage drop at 1A since it’s a charge input?
Ah, that’s drawn oddly.
I believe the reason they’re doing this is actually because they want to “broker” the setup. Let’s look at 3 situations:
In the first case, there is a small drop from the schottkey (.2V or so) and Vin “sees” 4.8V
In the second case, the same thing happens and the battery is charging. The lipo battery with a lower voltage “loses” in because the voltage on the anode of D3 is lower than the 4.8V of Vin
In the 3rd case, the battery now has a path back around the charger, via D3 to power Vin.
At least that’s how I read it. Does that make sense? I don’t like how they drew it, but D1 and D3 are there to broker what is powering Vin and it should work fine.
What is a better way to draw it?
Thanks for that!
I looked at it a bit more after reading what you said, and it makes sense.
Also, I was puzzling whether or not the LiOn would accidentally waste a bunch of power charging itself off itself or idling the charge chip, but the datasheet says “the input supply must rise to a level 150 mV above the battery voltage before the MCP73831/2 become operational.” Since the input in the battery only case will always be .2v lower we’d never hit a weird edge case like that:)
Wow, a lot of logic with three diodes!
Cool! Now If I could only get a good grasp on when to use a decoupling capacitor why of the values I think I might even fully understand this circuit 
Thanks!
I’d probably separate it out and show Vin as a very discrete element. Maybe just have the two diodes connected to their respective nets and have it going into Vin. Something like this:
Maybe even labeling it as “power brokering”
And of course I’d attach that top diode 
I was actually looking at a similar schematic today for the Adafruit Trinket, and wondering about D2 and D3 here:
It looks like they connected them similarly to how @ChrisGammell drew it above. I’m guessing they serve the same function possibly.
Both diodes D1 and D3 (D2 and D3 in the second example) are used for the same purpose, to allow current flow in one direction but block current flow in the other direction. The reason they are needed in these examples is pretty much as @ChrisGammell describes, although I think he has coined the term “power brokering” 
Diodes used this way are typically referred to as blocking diodes
While I disagree that the first example is drawn “oddly” perhaps the following is slightly more intuitive.
Eventually you will learn to recognize the purpose of circuits regardless of how they are drawn.
When choosing a blocking diode the four main parameters you need to consider are:
Items 2 and 4 are generally why schottky diodes are used but it depends on the circuit. If the circuit can tolerate it regular rectifier diodes can be used instead.
Item 4 is especially important when the diode is used to supply power from a coin cell to a circuit such as a real-time clock (RTC). In such circuits a high Ir can dramatically shorten battery life.
Of course, if you are using the diode in a high current application you will also need to consider the maximum power dissipation.
All royalty checks should be made out to “Christopher Gammell, word inventor™”
Agree, this smaller change keeps things simpler and localized to the parts.
Those 4 points you make at the end are spot on! Thanks!
