SOLAR LITHIUM ION/POLYMER CHARGER Application Help

[adafruit_support_mike]

Got it. So given my design how would you recommend I give the XBee a "clean" on/off transition?

There are chips that monitor voltage and emit a signal if it falls below a threshold value: http://www.digikey.com/product-detail/e ... -ND/482250

I've been thinking of making use of the error and shutdown pins on the LP2989 but perhaps there is something better you can recommend?

That's the standard solution. Shutting off the regulator is an easy way to make sure you got everything. Most voltage monitors (like the one above) produce active-low output, which would control the /SHUTDOWN pin correctly.

Also, is there no way to regulate off the MCP73871 (i.e. I really do need the linear regulator)?

If you're taking power from the MCP73831's LOAD output, no. That's a simple pass-through with no regulation.

[amahpour]

So the TC54VN3002EZB drops out at 3.0V. Don't I want my turn off at 3.7V?

Also, it doesn't resolve the issue of cloudy days in which the device power will still oscillate (granted on the order of seconds versus milliseconds).

Are there no regulator solutions on there for this type of application? I've spoken to Linear Tech about this and they don't really have anything on the micro scale with solar harvesting. Perhaps something with TI?

[adafruit_support_mike]

So the TC54VN3002EZB drops out at 3.0V. Don't I want my turn off at 3.7V?

That would shut the system off just about where a LiPo hits its normal operating voltage.

You want the shutoff voltage to be near the low end of the load's normal operating range. For a 3.3v device, a 3v cutoff is reasonable.

Also, it doesn't resolve the issue of cloudy days in which the device power will still oscillate (granted on the order of seconds versus milliseconds).

The voltage sensor linked above has hysteresis of about 50mV. It will signal low-voltage when the input falls below 3v, but won't stop until the input rises above 3.05v. If you're measuring the voltage of a LiPo, that can only happen after the battery has absorbed some more energy.

You can increase the hysteresis by adding a voltage divider:

hyteresis.jpg (11.4 KiB) Viewed 640 times

When the voltage monitor's output is high, it only sees Vin as the supply voltage. When Vin falls to 3v, the output goes low, pulling one end of the 10k resistor with it. That pulls the voltage at the monitor's supply pin down to about 2.6v.

With the monitor in that condition, Vin has to rise to about 3.5v before the center of the voltage divider crosses the monitor's 3.05v turn-on threshold. For that to happen, the battery will have to absorb enough energy to keep the system from shutting off within a few seconds.

[amahpour]

Hi,

Thanks for the circuit diagram. I'm just trying to figure out what exact pins this would go to on the IC. Perhaps you can correct me if I get this wrong?
Vin = IN, pins 18 and 19 on the MCP73871 IC
Monitor output = ?

Just to clarify, this does or does not include the precision voltage reference?

Also, if I am shutting off at 3.05V what will the regulator's output be at 3.06V? Meaning, I would need at least 3.5V to regulate a 3.3V output, no? That's why I asked if I should be shutting off at 3.7V (unless I change my regulator to a buck-boost).

[adafruit_support_mike]

The voltage monitor I linked a few posts above has a built-in voltage reference and three pins: VCC, GND, and OUT. GND goes to GND and VCC would connect to the LiPo's positive terminal.

As I remember the conversation, you wanted to send the XBee a "voltage is getting low" signal so it could shut down gracefully. To do that, you'd connect the OUT pin to one of the XBee's inputs. You wouldn't want to shut off the MCP73871 on the Solar Charger because then it wouldn't charge the LiPo. If you just wanted to cut off power to the XBee, you could use the voltage monitor to control an N-mosfet between the XBee's GND pin and the LiPo's GND.

[amahpour]

Hi. That's kind of what I'd like to do except no need to transmit a signal.

As I mentioned before. On a series of cloudy days my battery dies. When it becomes partially cloudy the battery charges up, turns on the regulator (thus the XBee), and turns back off after the battery drains. This keeps happening time after time. After a few power cycles the XBee goes crazy and won't respond anymore unless I power cycle it manually (lame).

My objective is to create a solid hysteresis circuit that controls the regulator to power off at close to not working and only turns back on at a solid voltage. The initial thought was to turn off at 3.5V since the regulator needs that to turn it into 3.3V and a turn on of at least 3.6V. I thought I could make use of the shutdown pin on the regulator.

Additionally, I've tried buck-boost converters with my solar charging circuit that have UVLOs which will up convert the LiPo battery voltage but I got far worse results. The converter itself starts to go haywire. I'm okay with sacrificing some power if I can get a working design but for now I'm at a standstill - the solar charging circuit simply does not work with the XBee in my current configuration. It's really hard to believe, honestly, so I MUST be doing something wrong. I just can't figure it out.

A penny for your thoughts.

[adafruit_support_mike]

Use the voltage monitor to control the XBee's connection to GND with an N-mosfet.

The XBee can run from a 3v supply, so that's a safe operating voltage. Cutting off at a higher level won't improve stability and will make the whole system less useful.

The voltage monitor has built-in hysteresis, and adding a couple of resistors as shown above will force the battery to charge to a higher voltage before the XBee turns back on again.

[amahpour]

So an N-MOSFET instead of the shutdown pin on the linear regulator? I thought that was the initial suggestion...

[adafruit_support_mike]

Could be.. the details on exactly how you want to shut down the XBee are kind of fuzzy.

You'll probably want some kind of power control external to the XBee because, IIRC, having it toggle on and off at low voltage was the original problem. You can do that with an N-mosfet, you can use a voltage regulator with an enable pin, or you can check to see if there's a reset pin on the XBee that will prevent it from booting.

A voltage monitor will tell you when the supply voltage falls below 3v, and when it rises above a voltage you can set as shown above. Those are your control signals. What you do with them depends on what you want to achieve.

[amahpour]

So was the circuit shown above based on one of the application notes in their datasheet?

Specifically Application 4.2:

App4.2.PNG (31.16 KiB) Viewed 555 times

[adafruit_support_mike]

No, it's just a standard positive-feedback hysteresis circuit.

The circuit above does use the voltage monitor to shut off a mosfet that controls power to the load. It doesn't have any hysteresis beyond what's built into the voltage monitor though.

[amahpour]

Cool. So now I'm just trying to figure out what kind of N-MOSFET to pick out. There are so many. Any recommendations?

Any one of these?
http://www.digikey.com/product-search/e ... geSize=500

[adafruit_support_mike]

The 2n7000 is a good general-purpose N-mosfet for low power applications: http://www.digikey.com/product-detail/e ... ND/4902350

[amahpour]

Hi,

So I just got the TC54VN3002EZB and applied 4V to VIN, tied VSS to GND but I only see 300 mV on the output. From reading the datasheet it seemed like I should see VIN as my VOUT when VIN > 3V. Am I misunderstanding this?
Note: I used a voltage divider to drive down my input from 5V to 4V.

Also, I bought the 2N7000-G as well. In order to drive the input voltage to my XBee I would connect the drain to VCC (3.3V), Gate to VOUT on the TC54VN3002EZB, and Source to GND. Do I need any Rgs resistance? Does the input of the XBee VCC connect to the source of the FET?

[adafruit_support_mike]

Do you have a pull-up resistor on the voltage monitor's output?