Why using short traces with the MCP73831 LiPo charger?

[yamadanao]

   

I'd like to use an MCP73831 to charge my lipoly battery. In section 6.2 I see they recommend short wires to the battery.

Here is the date sheet of MCP73831 :http://www.componentschip.com/details/Microchip-Technology/MCP73831EV.html


I bought my battery with wires, 20 gauge and approximately 30cm long. The maximum charge current will be 500mA. I know that there will be a voltage drop over the wire and I therefore will lose power. This power loss will be very small: 0.60?0.03331?0.52=0.0049965W

(length, resistance / meter, current). So I don't see how this is a good reason to keep the traces as short as possible.

So besides reducing power loss, is there any reason to keep the traces as short as possible?

[alm]

Quote from the datasheet:

For optimum voltage regulation, place the battery pack
as close as possible to the device’s VBAT and VSS
pins. This is recommended to minimize voltage drops
along the high current-carrying PCB traces.

So they are worried about voltage drops, not power loss. Any voltage dropped over those traces will make the battery voltage appear higher, leading to premature charge termination. Based on your numbers the voltage drop over the wires would be about 10 mV, or about a 0.25% error (for V_BAT = 4V). This could screw up the 0.05% typical / 0.3% max load regulation. If this is a problem in your application is for you to decide.

My guess is that you can get away with it, since fast charging is in constant current mode, and currents will be much lower in constant voltage mode. Still, it is clearly not something that Microchip endorses.

[Bruce Abbott]

I don't see how this is a good reason to keep the traces as short as possible.

You are right, there is no good reason. The only thing you have to worry about with this IC is burning it out. Charge at the lowest current you can put up with.

[mikerj]

I don't see how this is a good reason to keep the traces as short as possible.

You are right, there is no good reason. The only thing you have to worry about with this IC is burning it out. Charge at the lowest current you can put up with.

There is a perfectly good reason for keeping traces/wires short.  Power dissipation is not the only concern when designing in this device, the datasheet should always be read thoroughly.

For optimum voltage regulation, place the battery pack as close as possible to the device’s VBATand VSSpins.
This is recommended to minimize voltage drops along the high current-carrying PCB traces.

Long wires between the battery and the charge control IC will introduce introduce additional inductance (adding to the inductance of the cell itself) that, in the extreme, could reduce the loop stability.  A bypass cap across the output of the charger controller is recommended.

[Bruce Abbott]

the datasheet should always be read thoroughly.

For optimum voltage regulation, place the battery pack as close as possible to the device’s VBATand VSSpins.
This is recommended to minimize voltage drops along the high current-carrying PCB traces.

The calculated voltage drop is only 10mV. This is less than possible inaccuracy of the chip itself (64mV), and should have minimal effect on how much charge the cell gets. 

Long wires between the battery and the charge control IC will introduce introduce additional inductance (adding to the inductance of the cell itself) that, in the extreme, could reduce the loop stability.  A bypass cap across the output of the charger controller is recommended.

Yes, which is why a 4.7uF capacitor is shown connected directly across the output. But the datasheet says nothing about minimizing inductance.

[KL27x]

I believe the bypass cap is mostly necessary if the battery is removable. This will prevent unnecessary power consumption and the status pin flickering while a battery is not connected.

As for inductance of the leads... I'm sure you can make a theoretical case where the exact inductance of the wires causes oscillation/fail of the current and/or voltage regulation. But I would be surprised if you could do that by accident, with normal wires that are not intentionally wound up into coils and/or crossing a room.

In fact, the main concern I would have for extra length of the wires is the voltage drop, as others have stated. But my concern is not that the battery charge will terminate early. It will get close enough, in this example, anyway. You can also select the cutoff current, of which there is a range with different part numbers - at least between 2.5% to 10% (of max charge current?), from memory. I think the main practical concern is that the last few percent of the battery will take significantly longer to charge. For w/e the reasons, though, it is trivial to put this device right by the battery with a little daughter board, if necessary. But of all the things that I can get wrong when making a board, <2 foot wires to the battery, here, I'd be totally fine to give that a try on a prototype.

This is all based on my own personal use of this device and my limited knowledge of electronics, though. I can't wait to hear how many ways I am wrong.   

[mikerj]

But the datasheet says nothing about minimizing inductance.

No, but it does say that inductance will affect loop stability, which amounts to the same thing.