Lipo battery. From 4.2 to 3.7 in four hours

[danshtr]

Hi All,


I am powering my addressable LED strip using 328 and a lipo battery at ~130ma.

It takes 4 hours for the Lipo voltage to drop to 3.7.

I plan to add sensors, one of them is L3GD20H which is rated for 3.6V. Should I be worried?

I know that I can use a diode to drop the voltage, but this will shorten the amount of hours I can get from the lipo.

Any ideas?

-Dan

[suicidaleggroll]

3.6 is probably alright since you have a slow discharge, most of the battery capacity will be gone by that point.

eg:

You're around 0.25 C, so you'd only be wasting a small fraction of the capacity if 3.6 is your cutoff.

Do you have an over/under voltage protection circuit for the battery?  Any other way of shutting off the circuit when you hit 3.6-3.7 V?

[linux-works]

a few things.

I've powered a 328 arduino chip from a bare lipo, unregulated, just taking its output direct, and it runs until it gets to 2.8v or so.  this is a 'normal' 328p chip using a 16mhz resonator, no special bootloader setting fuses to non-5v/16mhz states.  I was surprised it went down that low.  note, it was a bare chip and resonator on perf board, not a prepared green board from a vendor.

also, make sure you have a lvc (low voltage cutoff) SOMEWHERE.  MANY batteries do not have it anymore!  all aircraft batteries for toy quads, for example, seem to omit the cutoff circuit and require the copter to have the lvc in its logic, not the battery.  for 'land batteries' I've heard that the circuit IS in the battery.  so, YMMV and you need to know this.  its BAD to undervolt a lipo, as I'm sure you know.

[danshtr]

Thank you all for your answers.

I am not worried about the 328. I think it and the roaches will survive a nuke.

I am worried about the  L3GD20H. It datasheet says the max V is 3.6v. Other than using a diode (see below), what other options do I have?

Regarding a diod. I am using 4001. Also, the current is going from 50am to 190 amp and back to 50am in a cycle (strobe effect).  I notice the drop to be 0.9 to 0.7V, and that the voltage after the diode flactuates with the amp usage. About +-0.1v.

So here are my questions:
1. Should I be worried baout the L3GD20H getting 4.2v?
2. If I choose to use a diode because of L3GD20H, which diode should I use.
3. If I choose to use a diode, are you suggesting that at 3.6V of the lipo, it will be already be 90% used?

Thanks again for your answers. Be easy on me, I am a noob. The only "formal" education I have is  the "Make: Electronics" book.

-Dan

[linux-works]

if you need a regulated voltage,then you need a regulated voltage.

take the lipo, step it down with regulation to 3.3 (are you sure your circuit really wants 3.6v?) and that will keep it safe.

in short, use regulation and a diode will just drop a constant bit, but that's not a good substitute for a regulator.

dc/dc converters will go both up and down and can get you where you want.  many are adjustable.  there are some tiny ones and I've used the '360' (I think that was its ebay name) to take lipo and create a nice 5v from it, and they are under a dollar each, in small qty.  turn the onboard pot to get the voltage-out you want and you're done.

[suicidaleggroll]

The MAX voltage is 3.6?  Yes, that's a problem.  Use a regulator to drop it to 3.3 or whatever that sensor actually needs.  A linear reg would be easier, but can't handle stepping the voltage up and will waste some power.  A switching reg would be more complicated but can step the voltage up or down and can get very high efficiency if that's your goal (which it usually is for battery powered devices).

[danshtr]

Yes, the MAX is 3.6v. It can operate 2.5v to 3.6v

I guess i'll have to find step down that operates between 3.6v and 4.2v as input.

BTW, what is wrong with using a diode (not a suggestive question).

[LaurentR]

I know that I can use a diode to drop the voltage, but this will shorten the amount of hours I can get from the lipo.

Logic circuits typically have their current draw drop with voltage, so the overall power including the regulator/diode will likely be lower than without.

[Audioguru]

You do not need a voltage regulator, the series diode will work perfectly until the low voltage cutoff circuit disconnects the battery when its voltage drops to 3.2V. Then the sensor gets 3.6V with a charged battery and still works at 2.6V when the battery is disconnected.

[danshtr]

Is there an simple IC to do the voltage cut- off? Or can you point me a to a simple circuit diagram?

[linux-works]

http://www.amazon.com/Lithium-Battery-Charging-Charger-Protection/dp/B00QGVQNA8

maybe that is a place to start?

that will take in 5v and charge the lipo.  you do need to have a charging solution, right?

this new version of the board (I'm seeing it come out now; was not around a few months ago) has an undervolt cutoff part, and so I -think- (not 100% sure) that there is a disconnect circuit that you can use, where your battery goes to one set of terms and the 'out' goes to another.  I need to verify, but I think this is correct.  (if I'm wrong, I'm sure someone will chime in soon enough, lol).

here's the 'mini 360' I was talking about:

http://www.amazon.com/4-75-23V-1-17V-DC-DC-Converter-Module/dp/B00NJCAI7G/

it will go from 5v down to a stable 3.3v for you.   I use this module and it works fine for when I'm internally 5v based.

there are still lots of things that are 5v based and so converting the battery TO 5v does make some sense.

usb is also 5v and the charger module runs from usb or 5v.  and with the mini-360, you can go from 5v down to 3.3v or even 1.8, I think.  use a few to derive any lower voltage rails you need from the 'standard' 5v.

so, something to think about.  yes, its extra up and down stages.  so what.  they are very efficient and easy to just by and glue together.  you can get things working quickly this way.

[danshtr]

Isn't that 360 requires 4.5 Vin min to operate?

Lipo won't cut it.

[linux-works]

Isn't that 360 requires 4.5 Vin min to operate?

Lipo won't cut it.

right, I suggested using 5v as the common base and getting the lipo up to 5v via an upverter.

from there, you can run native 5v things (there tends to still be a few good sensors around that are 5v) and for things that are 3.3v, you regulate DOWN from 5 to 3.3

going from unreg lipo range to 5v will use all of the lipo's useful power.  and going from 5v down to 3.3v will keep 3.3v things in perfect regulation even if the 5v rail drops a bit.

its just one way.  given how small and cheap up and down dc/dc verters are, its quite a do-able way.

[mikerj]

Yes, the MAX is 3.6v. It can operate 2.5v to 3.6v

I guess i'll have to find step down that operates between 3.6v and 4.2v as input.

BTW, what is wrong with using a diode (not a suggestive question).

Just use an LDO regulator if the current draw is low.