Lipo w/UVLO and down to EOD powering solution?

Mag748

Hello,

I am trying to come up with the best way to run a new propeller project on a single cell LiPo. I know this has been done countless times, but this is the first attempt I am making and after much researching, there are still some concerns I have.

Initial data
The prop should always be powered by 3.3v. When in "sleep" mode, it will run at RCSlow, etc.
There are peripherals that all run at 5V and should not have any power while the prop is "sleeping".

Problem
My original idea was to boost the LiPo to 5V and then LDO to 3.3v. The problem with this is that the booster consumes large amounts of current, so I would need to turn it off in sleep mode. But that would turn off the prop as well.

My second idea is to LDO the lipo to 3.3v for the prop and separately boost the lipo to 5V for the peripherals. The booster/charge pump would have a shutdown feature that can shut off the booster and disable the peripherals while in sleep mode.

The second option seems the best to me, but it introduces 2 problems that the first option did not have.

• When boosting to 5V before the LDO, I can use the Lipo down to its 2.7V End of discharge (EOD) voltage and still have full 3.3V to the prop.
• The 3.3v LDO would be connected directly to the lipo, and have the potential of running the battery down past its EOD voltage. With my first idea, the 5V booster connected to the lipo had a built in UVLO (Under Voltage Lockout) feature that would shut everything down in the event the battery reached EOD.

So, my question is, how can I combine these ideas into one perfect solution:

1. No 5V peripherals should receive power while the prop is "sleeping"
2. All 5V peripherals should receive regulated 5V when prop is "not sleeping"
3. Give a constant regulated 3.3V to the Prop within the full voltage range of the LiPo (Vin: 2.7V - 4.2V)
4. Protect the LiPo against undervoltage (Vin < 2.7V)

An additional bonus feature is to minimize all quiescent currents. But I am fine with small quiescent currents I.E. the ~2uA the LDO consumes while powering the prop.

Any suggestions would be greatly appreciated.

PS. I know there are some LIPOs that have built in under voltage protection, which is great, but I want my circuit to be safe in the event a lipo without this built in protection is used.
PPS. Am I completely wrong assuming there is no way to boost the Lipo to 5V without wasting huge amounts of power? Huge amounts would be >100 uA

Thanks,
Marcus

kwinn

Use a buck/boost regulator with UVLO for the 3.3V and a 5V boost regulator with shutdown for the +5v. An alternative would be to have the prop monitor battery voltage and control the shutdown.

SRLM

I can't get you a "perfect" solution, but I'll suggest a good one. I doubt that you'll find something that hits 2uA for an LDO regulator (since it effectively disappates the difference in heat).

But anyway, how about something like this for the 5v rail: http://www.pololu.com/catalog/product/2115
For the 3.3v rail, you could use something like this: http://www.dimensionengineering.com/products/anyvolt-micro

If you really wanted to do deep with this, you should browse the ST, TI, and the others for "Power Management IC". Those do exactly what you want, at the cost of significantly increased design time.

WBA Consulting

This could easily be done with a setup like the one on my PropBSC module. It uses a buck-boost IC (TPS63060DSCT) to generate the 5 volt rail from a 2.1-12V input and that is fed to a step down simple switcher (LMR10520YSDE). You could use two buck-boost circuits (3.3 and 5) and simply control the enable pin of the one driving the 5V rail with the prop.

Don M

I have used these with great success. https://www.sparkfun.com/products/11231

5v or 3.3 volt out. Built-in charger.

Tracy Allen

Power supply, always an issue with tradeoffs, no? How much current do you need at 5V and 3.3V?

I use the LT1615 for boost, 20µA quiescent current, 0.5µA in shutdown, up to 350mA output. There are many chips of that sort that vary in output current and voltage handling. The LTC3525-5 in a 6-pin SC70 takes 3V in a 5V out with Iq=7µA in burst mode. Even so, for lowest current in RCSLOW mode, it is best to shut down the boost and operate directly from battery/LDO.

I use the LT4071 for the embedded LIPO battery charger. It has the UVLO function that comes between the battery and the regulator. It can be set for either 3.2V UVLO (for long battery life and more charge cycles) or at 2.7V UVLO (to extract the last smidgen of capacity at the expense of battery life). However, it is strictly micropower, 50 to 100 mA max. Good when powering off of harvested energy. Can you operate your Prop at 3.0V?

Mag748

After quite a bit more thinking about this, I have an updated idea.

I should only need about 200mA at 3.3V and 50mA at 5V. The 3.3V rail will be LDO’d directly from the battery with an MCP1700.

The battery will be boosted to 5V with an MCP1640 which has a true output disconnect shutdown that will be controlled by the prop. I noticed that a lot of boosters simply connected the input to the output when they shutdown, which would not have been good.

As for the UVLO battery disconnect device, I was considering the Seiko S-8241 series, but it requires external MOS switches and seems like it’s more trouble than it’s worth. The LTC4071 does look like it would be something that could work, but at almost $5 it seems a little steep. I was going to use a MCP73831 as the battery charger. I think I will just do away with the protection.

I realized that the brownout voltage for the prop is 2.7V which is the same as the EOD of the battery. So without any additional circuitry, I can simply enable the brown out detector, and when the prop goes into reset, it will disable the 5V booster.

The only last issue is that I have a display on the 3.3V rail that will need to be told to shutdown before the prop goes into reset, or it will remain on. So, I really should monitor the battery level with the prop so that right before brown out, it shuts the display down. How would that best be accomplished?

I am now going to create the schematic for this power system and see how it all goes together. I will post again with that and any other notes I come up with.

Thanks,
Marcus