[Sunshine55]

I have a pocket amp just finished, which works well but has presented me with a problem i have no solution for. It's about charging the battery.
I am using a 3.3v Li-ion battery. Sch for chargers are on the net by the hundreds. However they all require taking the batt out of the cir to charge it. My cell phone also uses the same 3.3v batt but i don't have to take it out to charge it. IC's like the LTC4053 or LTC4054 tale care of the cc/cs charging management for the battery but how is the batt disconnected from the phone while the charger is plugged in. And and the phone still works when batt is being charged. I can figure out a complicated cir that would disconnect the batt and use a reg to run the amp from the charger while it is also charging the batt threw another path and the LTC4054, but that just seems a bit much. A switch could be used to select charging or run but that is if i had the room for the sw. lol.
Anyone out there with a solution? Endless surfing has provided nothing but a lot of charging cir's. What have the cell manufacturers got that they are not telling us?

 

[MisterX]

Here is one way to do it.
 

 

[MindsMirror]

The typical application note in the datasheet of the LTC4053 show the battery, charger, and load all paralleled. Can you just do that, and have the charger output supply power to the load and the battery at once?

 

[Sunshine55]

No. The data sheet shows the supply feeding the reg then the reg feeding the batt and cir in parallel. Electrically it would work but having the cir and batt in parallel would confuse the reg IC since it would not just see the batt and it's condition. Also the reg has a program pin which accepts a res value depending on what current one wants to charge the batt. This meens if the cir is turned on the IC would have to supply batt and cir. Then,, if cir is turned off while charging the batt and running the cir all perameters will change and the batt will be getting the currect intended for both cir and batt. Which would be far  too much and boom.

 

[Sunshine55]

The PTC4054  will declare the batt dead and begin charging when the batt is 3 v. So
that would mean the amp must work down to 2.4v. Soooooo close yet so far away.

 

[MisterX]

Charge pump?
 
Example:
http://www.linear.com/product/LTC1751

 

[Sunshine55]

The train gets another car.  I have done some measurements and Shottky diodes seam to drop the least amount internally
so using the diode bypass method is looking ,,,maybe. Tests with the 1n4004, drops .54v, (I'm sure varying a bit from diode to diode)
but the Shottky UF4004 version only drops .45v. So if i get UF4001s, which have only one diode in the package would drop the least.
At least less than the 4004 series since i don't need the higher voltage value.
Using isolation diodes however decouples the cir from the supply and the batt, thus, needing filtering again. So it is then how much
capacitance can the IC handle at startup before going into shutdown mode. I know i'm using a batt, but, Bluetooth is being used as
well and really likes to get into the audio so filtering is a must.
 I looked at the charge pump solution like the 1751 also Maxim has a 2 chip solution (MAX1555 and the MAX8881) but these both
decouple as well requiring filtering after the supply source, whichever that may be at the time. Perhaps boosting high then regulating
down to 3.3v would allow more filtering.
 I'm also not ruling out the switch idea,,,,, yet. OR perhaps a barrel connector with the extra sw in it to cut the cir pwr out when plugged in,
if i can find one with an isolated sw. I would like to use the amp while charging though.
(can't have everything) keeps whistling in my head.