bq25504 charging battery and supercapacitor

No answer yet? I thought I posted on the OFFICIAL TI forum....

If I understand correctly, you want to place a LiIon battery and supercap in parallel on the VBAT pin.  Does the LiIon battery have a protector that opens to protect the battery from over-discharge?  If so, then the super cap would slowly charge from the bq25504 in cold start until the protector threshold.  Then the protector would close, placing the LiIon battery in parallel with the super cap, which would cause the battery to most likely discharge below the protector threshold again as it sees the super cap load.  This would repeat for a while until the voltage at VBAT is above the protector voltage.  If the LiIon battery does not have a protector or if it is more fully charged then it will charge the super capacitor at assembly.  If the voltage at VSTOR is above 1.8V, the bq25504 operates in main boost charger mode.  If not, then it runs in the much less efficient cold start mode.

Let me place it another way... If I want to use two storage elements (battery and Supercapacitor) and a load where and in what way should I connect them?

Thanks in advance

The load is typically connected to VSTOR, unless there are large load transient which need to be pulled directly from the battery.  

Regarding two storage elements, do you want to switch between them, i.e., charge one and then the other and then discharge one and then the other, and repeat?  This would require external switches to switch between storage elements and comparators.

In theory, what I want to do is to have the two storage elements and charge them when energy is harvested from the solar panel and the load (wsn) is consuming less energy than the harvested. When there is no sunlight the elements will discharge and provide power to the wsn. That is the goal.

I can think of two options:

1.  two separate bq25504's with external back to back FETs to switch the storage elements of each to power the load.  The VBAT_OK signals of each could help you decide which element to use for powering the load.  Use this option if your solar cell can provide more power than one bq25504 can handle or you have two separate panels.

2.  one bq25570 with battery on VBAT and super cap on VOUT.  The buck converter of the bq25570 is powered from VSTOR so its output will have to be lower than the voltage on VBAT.  You will still need the back to back FETs to switch the load between the two elements.  The VBAT_OK signal tells you if the battery is above a certain level, not the super cap voltage so you will want to add a separate SVS to monitor the super cap voltage on VOUT.  Use this option if the solar cell can't provide more than the maximum input power of the bq25570 (=0.8 x solar cell VOC x 100mA).

After reading thru the datasheet for BQ25570, I was thinking for my system (solar energy harvesting for an animal tag that has a video camera)

0) Solar Cells that are approx. 5.05v open circuit, work quite well with 4.0v (80%) in sunlight

1) Supercap on Vstor (0.1 F, 6.6v rating)

2) Use Vout to power the ultra low power processor and sensors (< 1mA @ 2.7v)

3) Li-Ion battery (3400mA-hr, 3.6v nom, 4.20v full charge) for powering the camera

The issues with this design are:

When Vstor is depleted and the Supercap needs to be charged up to the battery undervoltage (3.3v) condition before Vout is turned on. This would likely be the case when the animal goes deep in the ocean for an extended period.  Ballparking a worse case estimate on charge time, I = C dv/dt = 0.3 mA = 0.1 * 3.3 / dt => 1100 seconds. (18 minutes) Ouch!

I may need to add some diodes/FETS to switch power to the processor from the Li-Ion battery for periods of very low light, but I'm trying to avoid the complexity power switching.

I am toying with the idea of a diode that goes from the Li-Ion battery to the supercap.   Does anyone know of an issue with that?   I'm going to go back to the datasheet and re-parse some details, but if someone has a recommendation I'd appreciate it.

Thanks,

Thom

What is the reason for the supercap on VSTOR?  Is it to allow continued operation down ~2.7V for VOUT after the LiIon battery's protector opens at 3.3V?

The IC has a PFET between VSTOR and VBAT with body diode allowing current flow from VSTOR to VBAT.  At battery attach (assuming both VSTOR and VBAT are depleted), that FET closes for 45ms in order to charge up the VSTOR capacitor > 1.8V if possible.  If the VSTOR voltage is above the VBAT_UV threshold of 2.1V approximately 64ms later, the FET will close open and remained closed until the VSTOR voltage drops below the VBAT_UV voltage.  If you battery has an internal protector that provides overdischarge protection, you do not need this feature.  Also, if your battery has an internal protector, you can add a diode from VBAT to VSTOR but that goes back to my question above.

With such a large battery, the charger (with maximum input power of 500mW) will take a long time to fully charge the battery when running in charger mode.  It will take even longer to get into high efficiency charger mode from low efficiency cold start mode if VSTOR even drops below 1.8V.

You might consider a two chip solution, using the bq25505 with integrated MUX to switch the system load power from your super cap to the LiIon battery when the super cap is discharged and then a TPS62736 or TPS62740 to provide the VOUT voltage. 

Yes, I was thinking supercap for continuous operation down to ~2.7v.      Processor is just under 1mA and after I finish the low power part of the firmware, should be around 30 to 50 uA average consumption (less if I lower the sensor sample rate).   Peak current draw is during flash memory writes to uSD. 

My battery is an NCR18650B (3400mA-hr) and I'll need the protection offered by the BQ. 

Here's what I'm thinking:

Charge the battery before connecting to the energy harvester.  

Monitor the battery voltage with the processor and turn off the load (camera) at around 3.4v.   

Look into monitor Vstor and deep sleep the processor if it gets to 1.9v

The camera is out in the ocean for months and the great white shark spends enough time near the surface so that over the course of a couple weeks (by my estimate, charge rate will average around 10 to 20mA based on light data) I'll have a full battery and be ready to shoot video again.

I'm looking to keep the processor running off solar with enough energy storage to get thru the night and a few worse case deep dives.   I like the efficiency of the boost converter and it's ability to later go to a 1.8v processor/sensor design (for now I'm at 2.7v).   I'd like to think I'm getting better efficiency using the buck converter than a simple linear regulator like the REG101 at low current.

Before jumping into the BQ25570, I kicked the tires on BQ25504 and BQ25505, but perhaps I didn't understand the BQ25570 well enough when I made the decision based on the integrated buck converter.   I initially thinking the BQ25504 with it's higher charge current would be the way to go, but the integrated buck converter is what swayed my to 25570 (I'm space limited to a 0.95" diameter circuit board).

I'll look at the footprint of the TPS designs you suggested.   After writing this, I'm thinking that it might be an option to ditch the supercap and more carefully manage the battery to keep operational headroom for the processor and use the TPS switcher if I can get it to fit for producing the processor voltage.   The faster charge of the supercap is appealing but it's adding design challenges as it looks like I'm pushing the BQ25570 into a space it was not intended to go.

Thanks,

Thom

The boost chargers of the 504 and 570/505 are essentially the same, with the 570 having tighter tolerance and wider input voltage range.  The 504 maximum input voltage is 3.0V so you couldn't use it with your solar panel.  One of difference between the 504 and the 505/570 is the VBAT_UV feature.  The 504 allows you to set the VBAT_UV threshold to prevent battery overdischarge using resistors.  The 505/570 threshold is interally set to 2.1V.

The 570 was designed to charge only one storage element.  All members of the bq255xx family have two modes of operation:  low efficiency cold start when VSTOR < 1.8V and highly efficient main boost charger when VSTOR>1.8V.  If you attach a partially charged battery to VBAT, the device will automatically enter main boost charger.  if not, and you enter cold start, you will have to wait a long time before the voltage at VSTOR rises above 1.8V. 

Funny thing about data sheets and maximum values.   I evaluated the 504 with 5v open circuit solar and with MPTT the Vin stayed at 4ish volts in bright sun (several hours each day).   I was not sure how to read the data sheet as the absolute max said 5.5v and then later in the data sheet it indicates Vin max is 3v.   The typical for my application when in the water is for the Vsol to be around 3.5v with MPTT that takes it under 3v - I had assumed (in some sort of mentally handicapped/gifted kind of way) that I could use the 5v open circuit cells with the 504!

The performance results were good and it ran for weeks in sun and dark and in the test tank under pressure. 

Thanks for letting me know that the 504 is NOT intended for the 5v solar cells.  It migh be good to drop a note in the datasheet explaining the difference between absolute maximums and the 'max' for VIN.

We did not test the 504 at voltages higher than 3.0V but I am not surprised that it worked at higher voltages.