BQ24650: BQ24650 precharge disable and fast charger timer

Hi Khoi,

There is no fast charge timer on this device. I think if you place the system before the sense resistor (L - System - RSNS - BAT), you wont have this issue, because the 250mA doesn't pass through the sense resistor and the IC doesn't think it is already in precharge regulation.

Regards,
Steven

I thought about that option, but the load when active, could pulse 10A from the battery for a short duration to drive DC motor.

If I tie system before RSsense, would the system draw 10A thru the FET and inductor, which are sized for 2A, instead of the battery. The battery is sized to handle the pulse current.

Thanks again,
Khoi

Khoi,

What are the specs for your solar panel? What FETs are you using? If the solar panel Peak Power is at 3A, then your FETs should be able to handle that much current. The rest of the current would come from the battery.

The bq24650 will regulate input voltage so that solar panel is at the peak power point. So it will in fact limit your input current when input voltage hits MPPSET regulation point.

Regards,
Steven

we are using SI7288DP-T1-GE3 and diode similar to TI EVM design. I thought the Rsense is used for current limit 2A detection as well.

So what you say should be ok to connect load before Rsense. we will get small power loss over Rsense during normal operation but it should not draw from solar input over 3A.

Khoi

I think this is fine. What is the Maximum Power Point of your solar panel? What is the voltage and power at this point?

As for RSENSE, this will only regulate when the current is going into the battery. If input voltage drops, MPPT is higher priority than ICHG regulation.

Regards,
Steven

The solar panel is 17V, 2A peak (30W). Solar panel will limit the load current. However, in some case we have 17V DC source to power the unit instead of panel so it will be an issue as the DC source voltage won't drop.

Reading the Application Report SLUA376–June 2006 (http://www.ti.com/lit/an/slua376/slua376.pdf):

I'm seeing similar behavior in Topology 1 issue 1, 2, 3.

If we use Topology 2, Design considerations and potential issues:

1. The total current is only limited by the IC peak current protection and the thermal protection thresholds but not the charge current setting pin ISET1 (Rsense in our case) . The charge current does not drop when the system current load increases, as shown in Figure 9. This solution is not applicable if the system requires a high current. 

Considering all the input options, I think our best way is still topology 1 and we need to put part of system to sleep to make sure it draws less than precharge. Also maybe we can reduce the precharge time as shown in appnote Application Report SLVA473–August 2011

Do you agree? 

Khoi

Khoi,

I think it would be better to reduce precharge phase like in the app note SLVA473 so that you don't have any trouble with precharge timer. Hope this works for you.

Regards,
Steven

Also following the appnote SLVA473 below, I only come up with R1 = 4.2K. The appnote show 5K for a reason I couldn't follow.

Design Example:
The bq24650 evaluation board is used to test the proposed circuitry. The feedback voltage regulation is given in the data sheet (SLUSA75) to be VFB REG: 2.1 V and VLOWV: 1.55 V. The reference voltage is VREF: 3.3 V. In the EVM board, the R3 and R4 values have been given to be R3 = 499 kΩ and R4 = 100 kΩ. They are designed for VBAT (Full Charge): 12.57 V. The diode is chosen to have approximately 0.5-V forward voltage. Using Equation 1 of the bq24650 data sheet, R1 and R2 can be designed. Resistor R2 = 10 kΩ was chosen to be 10 times smaller than R4; solving Equation 1 for R1, the value was found to be 5 kΩ.

Thanks,
Khoi

Khoi,

Where are you reading the ICHG current? Into the battery or across the SRP-SRN resistor? I think the load is taking too much current and the solar panel is powering the load instead of charging the battery. Solar panel should be capable of providing the following: MPP = (MaxVBAT*(ICHG+ISYSaverage))/0.9 (assuming 90% efficiency).

For the precharge region modification, the R1 and R2 resistor divider should be around 2.2, so that VFB is clamped to 2.2-Vdiode. Pretty much you don't want to go below 1.55V on VFB.

Regards,
Steven

Steven,

Thanks I think I got all the issues on the charger figured out. I was referring to this section from the datasheet.
We were checking the status pins in the field unit to see if battery is charging or not, but if the solar panel voltage is not high enough, Status pins are similar to the case when it's charging.

8.3.2 Input Voltage Regulation:

If the solar panel or other input source cannot provide the total power of the system and bq24650 charger, the input voltage drops. When the voltage sensed on the MPPSET pin drops below 1.2 V, the charger maintains the input voltage by reducing the charge current. If the MPPSET pin voltage is forced below 1.2 V, the bq24650 stays in the input voltage regulation loop while the output current is zero. The STAT1 pin is LOW and STAT2 pin is HIGH.

Thanks,
Khoi

Hi Khoi,

Yes, if the solar voltage is not high enough(Vin regulation), BQ24650 thinks it is in fast charging mode, So stat1 will be low stat2 will be high.

Alen Chen

We finally had the board and tested R1/R2 divider. Since the diode forward current is so low, the forward diode is only around 2V to 3v (instead of 0.3 to 0.7V per 41N4148 gen purpose diode). To let charger VFB tracks battery voltage above 2.0V, I had to tune R1/R2 divider to around 2V instead of 2.2V so if VFB < 1.7V, diode will become forward.

Also another question comes up:

R sense = 0.02ohm => I charge = 2A so the battery full when current < 0.2A. We need to keep idle load below 0.2A or else charge will never terminate in constant voltage region.

solar--charger---battery --ina226 ---load

If we select R sense = 0.01 ohm => I charge = 4A, battery full when current < 0.4A? This would anywa if our solar panel won't source more than 2A charge anyway. Do you see an issue with this option?

Thanks,
Khoi

Hi Khoi,

If there is no misunderstanding, your question is about the operation when solar panel cannot support system load(including charging and SYS loading), right?

Or could you please verify your question? Thank you!

Best regards,

Ann Lien

iIf we set r_sense to 0.01 and since the chip amd FET are sized for 2A, we don't want to exceed it with larger than 2A solar panel (18W) but we will gain 0.4A I_charge full range.

We don't expect the solar to power yhe load aince the battery in series with thr load will suplly larger than 2A current for motor.

Khoi

Hi Khoi,

If we use RSR=0.01, then the I_charge will be 4A. If FETs and output component are set for 2A, it couldn't meet the charge current requirement.

Does "Gain 0.4A full range" mean the charge current will be 2.4A in your case? 

Best regards,

Ann Lien