BQ24075 short-circuit issue with TPS63020

Hi Tom,

What are the values of ISET and ILM pins (Resistors)?

Is it possible to have scope captures of the Ibat, I output and the I input during start up?

 Also check your output voltage, because if it falls by 0.25V below Vbat, the output will be turned off.  

Thanks

Tahar

Thanks for getting back to me.  ISET is 890 ohm, and ILIM is 1.1 kohm.  Attached is a schematic of my power supply.  The bq27500-v130 and bq24075 are connected as per app note SLUA490. Note that the node labeled "Vbat" in the schematic is not the battery itself; "Batt+" and "Batt-" are the battery terminals.  Also, there is a jumper connected between TP1 and TP2 in the upper right.

My issue occurs when the 5v boost converter (TPS63020) is enabled while the battery voltage is either low or has a high source impedance.  The TPS63020 has a high inrush current because of the 1000uF capacitor at its output, but otherwise has no load on it for my test.  In the attached oscope capture "li-po_at3_8v.png", I show a successful startup of the 5v rail, with a lithium-polymer battery at 3.8v as the input.  Channel 1 is the total system current, channel 2 is the 5v rail, channel 3 is the positive battery terminal voltage ("Batt+" on the schematic), channel 4 is the OUT pin of the bq24075, and the light pink trace at the bottom is a math trace, taking the difference between channel 3 and 4 (what the bq24075 uses for short circuit detection).  As you can see in this screen cap, the difference rises to right around 250mV, but isn't quite enough to trip the short circuit protection.  This makes sense, because the current only peaks at about 4.25A, and the bq24075 datasheet says it can handle 4.5A through the OUT pin.

On the next screen cap, "powersupply_at_3_8v.png", the input to the system is a bench power supply at 3.8V, connected through 1m leads.  Between the source impedance of the power supply itself and the inductance of the leads, the source impedance appears to be greater than that of the Li-Po battery at the same voltage.  However, even though the current draw is roughly the same (this peaks at over 4.5A, but is only there for 20us or so), you can see that the short circuit protection gets activated.  The difference between channel 3 and channel 4 appears to be greater for a given output current.

Finally, please take a look at "cr123_at_3_0_v.png".  In this case, I used a CR123A primary lithium battery as the system input.  This type of battery has a much higher source impedance and has trouble pushing more than about 2A of current.  In this case, when the 5v rail is enabled, the system current only reaches about 2.25A.  However, even this low current causes enough of a voltage drop across the bq24075 to trigger the short circuit protection.

For the record, I appear to be able to prevent this problem by adding a soft-start extender to the TPS63020 as per app note SLVA307.  However I want to be sure that I know why this is happening, because the 5v rail will be subjected to high current spikes in normal operation, and I don't want to trip the short-circuit protection when I have a low battery if I'm not drawing more than 4.5A.

powersupply_at_3_8v;

li_at_3_8v

cr123_at_3_0v;

Hi Tom,

It looks like the charger is doing what it supposes to do; it is just protecting the battery during heavy load at the start up. If the difference between Vbat and Vout is greater than 250mV it shut down.   

Adding the soft start on the TPS63020 will help. Also, adding a capacitor at the input side of the TPS63020 will help absorb the rush demand of the output of the boost. You need to size the cap appropriately with your out put cap(~2x).      

Thanks

Tahar

Yes, I understand why the short-circuit protection is getting triggered.  And putting a larger cap at the input of the TPS63020 would probably be a matter of good practice as well.  However, that's not what I'm concerned about.

What I'm concerned about is that I'm getting 250mV across the charger in certain situations when I'm not even approaching the current limit of the chip.  For example, in the screencap I sent of the CR123 battery (which was just for testing, I know those can't be charged) - only 2.25A was being drawn through the bq24075, which is HALF of the stated current limit for that part, however it was still enough to cause a 250mV drop.  This is a concern for me, because I don't want to be triggering the short-circuit protection every time my product draws a couple of amps, when the data sheet clearly states that 4.5A is the maximum recommended current.  My question is: why whould such a low current be causing that high of a voltage drop?

Tom,

That is because of the nature of your battery and your load. It does not matter how much load you are drawing as long as there is a 250mV across Vout and Vbat, the charger should shut down Vout.

Can you try avoiding using the CR123 battery in your application?

Hope that helps!

Thanks

Tahar

Issue:

-TPS63031 provides constant 3.3V 50-200mA, NO issue.

-Using a battery at 4V, when the MODEM is turned on, the circuit functions as expected.

-When the battery drops to about 3.5V, and the MODEM is turned on, the bq24075 short circuit protection kicks in and bq24075 current supply halted.

-Replace the battery with a power supply, when the power supply voltage drops to approx. 3.5V SAME issue as above, when MODEM is turned on, the bq24075 short circuit protection kicks in and bq24075 current supply halted.

Current work-around is to use an external ‘Soft Start’ (App. Note from TI) circuit on TPS63020 and issue goes away.

Basically the facts listed above are all correct - I might also mention that even when using a source which is current limited to 2A (like a 3.0V CR123A battery), the short circuit protection in the bq24075 kicks on.  And the real question probably isn't best phrased as "why is the external soft start necessary"; but rather "Why does the short circuit protection get triggered in certain situations even when the current through the bq24075 is less than its specified maximum?"  Or, to rephrase, "What about the bq24075 causes a 250mV drop from BAT to OUT, even when it is only passing about half of its maximum specified current?"

-Tom-