Liion charger IC selection

[gaminn]

Hi, I spent nearly two hours searching for charger IC without no luck. Can you please help with selecting the IC:

I'm designing standalone single cell liion charger. I want:

Vin = USB
Ichgmax >= 1.5 A, I prefer switch mode charger IC as linear would get hot
integrated solution (fet, diode, current sense resistor) - I don't want tens of other components around
settable current limit (by resistor)
no timer or RC settable timer
dynamic power management (charger automatically lowers charging IC if Vin drops under certain limit due to vin overloading)
can be purchased on Mouser

I don't need stuff like power path, USB OTG, I2C control....

http://www.linear.com/product/LTC4001 is almost ideal if it can be purchased on mouser and if it has dynamic power management.

Thanks!

[Ice-Tea]

Might wanna check with Ti. They have a gazillion devices, all top notch.

[gaminn]

I think I checked all manufacturers. TI was the first one I checked... I must be blind... going to check it once more.

[gaminn]

I found MP2625 from Monolithic Power. Never heard of this manufacturer but the IC seems to be exactly what I need. Do you have any good/bad experience with Monolithic Power ICs?

[nctnico]

Look at the LTC4053 and don't be too overly set on a switching solution because it is likely not extremely more efficient than a linear solution. Also for USB charging I'd set the current to less than 500mA if you want to charge from standard USB ports (if they aren't current limited).

[gaminn]

Primarily, it will charge from USB 2A wall adapter.

2 A charge current linear solution would generate 2 x (5 - 3.7) = 2.6 W of heat. 2 A charge current switch mode solution with 90 % efficiency will generate 3.7 * 2 / 0.9 * 0.1 = 0.82 W.

[nctnico]

If you look at the LTC4001 solution you'll see it produces nearly 1.25W of heat (and probably much more using an affordable inductor). Reaching over 90% efficiency at these low power levels is very hard. You'll also need to consider EMC when going for the switching solution which may need extra components / board spins. Besides that there will be some cable losses at 2A so at the charge point you may be left with 4.8V.

[gaminn]

I guess the inductor will add 0.2 W power losses. According to up to 95% efficiency of MP2625 I think I can still achieve much better efficiency compared to linear solution.

I hope there will be no EMC issue - I used many low power DC/DC ICs and never had any...

[Ice-Tea]

Monolitic Power is a chinese manufacturer, but decent. You'll find their stuff in a lot of customer electronics. Bit like A&O.

[NiHaoMike]

If your application allows it, charge to 4.1V/cell at 1C or less. Your batteries will last a lot longer.

[KL27x]

Primarily, it will charge from USB 2A wall adapter.

2 A charge current linear solution would generate 2 x (5 - 3.7) = 2.6 W of heat. 2 A charge current switch mode solution with 90 % efficiency will generate 3.7 * 2 / 0.9 * 0.1 = 0.82 W.

Not necessarily quite that bad. USB specification allows voltage sag to something like 4.4V... Which most of your 2A chargers are going to do at 2 amps.

Charging li ion cells is practically why 2A USB chargers exist. And it doesn't take 5V to do that.

[Mr.B]

I use CONSONANCE CN3083/CN3063.
A "China Special" and unfortunately not available at Mouser or Digikey.
600mA max charge current.
One package solution - with one resistor to set max charge current.
I have used them in 'only' hundreds of devices, but they have not failed me yet.
An easy low component count and low cost solution for me.
Datasheet attached.

[HKJ]

The TP5000 chip works fine, but I do not know how easy it is to get.
I tested a charger with it here: http://lygte-info.dk/review/Review%20Charger%201.5A%203.6-4.2V%20UK.html