Push buttton as a on/off Switch.

[Ian.M]

The quiescent current of the regulator drains your battery.  The total quiescent current of the whole circuit should be under 30uA, as that is comparable to a good 18650 cell's self-discharge (1%/month).  If you can get it under 5uA it will be entirely negligible.

[Ian.M]

Most ultra-low quiescent current fixed regulators have a max Vout  of 5V.  It would be advisable to swap out the CD4000 series CMOS chip for a 74HC one or possibly a  74LVC2G04 SMD dual inverter, for the toggle latch, then use a N-MOSFET with a max Vgs threshold <2.5V, with grounded source as a level shifter to get the latch output up to the P-MOSFET gate, which is referenced to the +Batt rail which may be up to 21V above ground.  If your P-MOSFET's max Vgs is under +/-25V, you'll need a potential divider between the N-MOSFET drain and +Batt to drive the gate with a lower voltage swing.  The choice of total resistance is a compromise between additional current when on and speed of switching on/off.  I wouldn't go below 100uA through the divider when on.

Alternatively, provided the charger circuit doesn't share a ground with the audio side of things or any other load on the same battery with a common ground, switch the negative side of the battery with a low Vgs threshold N-MOSFET as a low side switch.

[Michaelaudio]

For simplicity id obv use a uC but i cant program anything so i have to do it that way.
I guess i should be able to find a NMos with Vgs <2,5V and Pmos Vgs above 25V.

Ultra Low qsc converters r always switching regularors, i assume.
I have to use a bosst converter for the amps, to get it at 24V fixed. Do U think this could couse any Trouble? If not, then another one wont bei a Problem i think. Back in the das i build my First BT speaker and used Lead batteries with a boost converter and an amp Board, didnt couse any problems.

[Ian.M]

No.  Many ultra-low quiescent current regulators are linear.  It wouldn't make sense to use a switching regulator just for the toggle latch, which only needs a couple of uA quiescent, and few mA peak when it switches.

I should note that if you can program an Arduino, you can program an ATtiny AVR MCU, using an Arduino running the ArduinoISP sketch as a programmer, and some ATtiny AVRs have low enough quiescent current to be viable for this application.   However if you have negative aptitude for C programming, you may well be wise to avoid MCUs.

[Michaelaudio]

I guess the MAX1615EUK+T will do the Job.

I learned the verry basics of arduino programming 3 years ago but never did anything since then so i'll stay with the current solution

[Michaelaudio]

Its Not easy to find a P Mos with AT LEAST +/-25Vgs. I didnt quite understand it, If U wanna switch a fet on, u want Vgs as low AS possible?..

[Ian.M]

Vgs(max) is a completely different parameter to Vgs_threshold(max). 

• Vgs(max) is the limit above which the gate oxide may fail, destroying the MOSFET.
• Vgs_threshold is the gate voltage that is just enough to start turning the MOSFET on.
• Vgs_threshold(max) is the *LOWER* voltage limit at which *ALL* MOSFETs of the specified type will have started to turn on.

Vgs(max) and Vgs_threshold(max) are not closely related except that a thicker gate oxide is more robust and also reduces the effect of gate voltage on the channel.

As a rough rule of thumb you want gate drive of at least  double Vgs_threshold(max) to get the Rds_on low enough to handle significant  currents, but obviously less than Vgs(max) by an adequate safety margin.

Power MOSFETs with Vgs(max) > +/-20V aren't particularly common.  I did suggest using a potential divider to reduce the gate drive: A 3:1 resistor ratio (larger one as the gate pullup) would give you slightly over 15V of gate drive with fully charged batteries, dropping to about 11V at their safe discharge limit, which would be suitable for 20V Vgs(max) MOSFETs.

[Michaelaudio]

And i thought U mean Vgs(th), idk why. Sometimes a Lil knowledge refresh is all it takes, i got u know. Guess i wont need the devider, the FQD11P06TM hast a Vgs of +/-30V

[Michaelaudio]

It Looks Like this now.

[ledtester]

What's your battery voltage?

For the same price of a MAX1615 you could use a MAX16150 which will also handle all of the pushbutton logic:

https://datasheet.octopart.com/MAX16150BWT%2BT-Maxim-Integrated-datasheet-140091574.pdf

or have a look at the TI TPS3421 which is about 50% cheaper:

https://www.ti.com/product/TPS3421?HQS=TI-null-null-octopart-df-pf-null-wwe

[Michaelaudio]

Oh thats nice, guess i'll use the 16150 then, so i can safe space.
Its about 21V (5x 18650)

[ledtester]

Note that the max Vcc for the MAX16150 is 6V.

[Michaelaudio]

Im using a 5V Regulator, should be fine.

[Michaelaudio]

Any improvements i have to do?