BQ24075T: VBAT too low to power on, but if it is already on the device does not power off

Hi Jaume,

How much current is current is being pulled at the OUT pin? Can you share a schematic with the LEDs and any information on what is at the OUT pin?

Best Regards,

Anthony Pham

Hi Anthony,

At turning on, we have measured around 350mA.

I can't share you the entire schematic, but I can share you the charger circuit and its following converters:

Regards

Jaume

Hi Jaume,

Do you have information on the capacity of your battery? Can you share some waveforms of the BAT voltage, OUT voltage, and the /CHG, /PG voltages when you osay there is flashing?

Best Regards,

Anthony Pham

Hi Anthony,

I have probably not explained myself properly, the LEDs that are blinking are not /CHG and /PG. The LEDs that re blinking are those called LED1 and LED2 on the output of the buck-boost converters.

Concerning the capacity of our battery, we have tested two types: 800 mAh and 1050 mAh.

Moreover, we have tested connecting a power supply to the battery pins (to simulate the battery working) and we observe that the power supply is not limited by the current (we defined a limit of 800mA on the power supply) and what we observe is that the device is consuming around 30mA.

I have just checked VBAT (in yellow) and VOUT (in blue). Please find attached the voltages at startup:

This sequence is repeated approximately every 55ms

Regards

Jaume

Hi Jaume,

Ah if that's the case, can we zoom out a little bit and increase the V/div on this so that they take up more of the screen? It's a bit difficult to read. I can look at the charger waveforms and forward this to the converter team to take a look.

Best Regards,

Anthony Pham

Hi Anthony,

Please find attached two differents images. As before, yellow represents VBAT and blue, VOUT

The first one is a zoom to the initial transient when turning on the device.

The second one is a startup overview. In red, I marked the region which has been zoomed before. This transient is repeated every 56ms as it can be appreciated with the cursors measurements

Regards

Jaume

Hi Jaune,

Is this right after you're plugging in the battery? I'm looking at your waveforms and I don't normally expect VOUT to be low if there is a batteyry present. 

Best Regards,

Anthony PHam

Hi Anthony,

Yes the scope image is taken when plugging the battery (where the red box represents that exact instant)

We have been reading on forums and also the datasheet with detail, this may be caused by the short-circuit protection of the BQ24075T?

Regards

Jaume

HI Jaume,

Yes this looks to be the case here. The time that you're seeing for every transient happens just about tREC(SC2), the time the device checks for exiting a short circuit event

Best Regards,

Anthony Pham

Hi Anthony,

Is there anything we can do to avoid this behaviour?

Is there any other battery charger which could help us?

Regards

Jaume

Hi Jaume,

Any chance that this remains when you plug in your power input and then remove it?

Best Regards,

Anthony Pham

Hi Anthony,

I'm not sure if I understood correctly your question.

When we have an external power input and then we remove it, we don't see this effect, the short circuit protection is not a problem in this case. There is a problem when we try to turn on a device only with battery when the battery is discharged below 3.7V

For our design, it's a backup battery. In most of cases, we will have a proper power input however sometimes we may need to turn on the device only with the battery and we have to assure that the device will turn on.

Regards

Jaume

Hi Jaume,

Jaume Forner Escrig said:

When we have an external power input and then we remove it, we don't see this effect, the short circuit protection is not a problem in this case.

Is this when the battery is at < 3.7 V?

Jaume Forner Escrig said:

There is a problem when we try to turn on a device only with battery when the battery is discharged below 3.7V

My suspicion is that the when the battery is below 3.7 V,  like 3.6 V, the high inrush current that's needed for the two converters that need to be powered is too much for the battery. When it tries to power up the DC/DC converters, it isn't enough and so the voltage on the OUT pin doesn't exceed the short circuit threshold. 

Best Regards,

Anthony Pham

Hi Anthony

Anthony P Pham said:

Is this when the battery is at < 3.7 V?

Yes, in any case, if additionally to the battery we have an external power supply to the IN pin of the BQ24075T, the device works as expected.

Anthony P Pham said:

My suspicion is that the when the battery is below 3.7 V,  like 3.6 V, the high inrush current that's needed for the two converters that need to be powered is too much for the battery. When it tries to power up the DC/DC converters, it isn't enough and so the voltage on the OUT pin doesn't exceed the short circuit threshold. 

We think that it is not caused by the battery. Because instead of a battery we have connected a power supply to the BAT pins limiting it at 3.65V and 800mA and the device did NOT turn on properly. And the power supply was NOT in protection mode, the total current consumed was around 40mA.

We really think it is caused by the short-circuit protection of the IC, and we would like to know how we can do so as to assure that the device will power up every time, without depending on the voltage of the battery

Regards

Jaume

Hi Jaume,

Jaume Forner Escrig said:

Yes, in any case, if additionally to the battery we have an external power supply to the IN pin of the BQ24075T, the device works as expected.

Does it still work as expected when you remove the power supply after you see it working as expected?

Jaume Forner Escrig said:

We really think it is caused by the short-circuit protection of the IC, and we would like to know how we can do so as to assure that the device will power up every time, without depending on the voltage of the battery

Anthony P Pham said:

Yes this looks to be the case here. The time that you're seeing for every transient happens just about tREC(SC2), the time the device checks for exiting a short circuit event

We're in agreeance here, I'm just looking for what is causing the short circuit to trip and I'm starting with the startup of the converters since those are connected to the OUT pin. 

Are you able to try adding some capacitance (for a start, perhaps 47uF) on the battery pin? This may help provide some current and stability for the transient if it is the inrush startup of the converters. And if you can share the waveforms that will be greatly appreciated. 

Best Regards,

Anthony Pham

Hi Anthony,

Anthony P Pham said:

Does it still work as expected when you remove the power supply after you see it working as expected?

Yes, when removing the power supply it continues to work as expected

Anthony P Pham said:

And if you can share the waveforms that will be greatly appreciated.

I have tried to obtain the inrush current waveform by means of a shunt resistor (0.91Ω), but the device turns on properly when I connect any resistance.

I have no other means of measuring the current. Is there any other information that could help you? Or any other way to measure the current I could do?

Regards

Jaume

Hi Jaume,

The ideal measurement would be a current measurement through something like a hall effect based current probe. Your method of using a shunt resistor would work to measure current but it reduces the amount of current in AC scenarios and so this affects the inrush I believe. 

Without the shunt resistor, did you try adding capacitance? If you aren't able to get current waveforms, similar waveforms to the ones you've provided previously will suffice.

Best Regards,

Anthony Pham

Hi Anthony,

Please find attached the scope. As before, in yellow VBAT and in blue VOUT

The first image corresponds to signal in general:

And the second one, it is a zoom to the start up:

I cannot really understand why increasing capacitance could help solve this problem. I thought that a capacitive line would increase more the inrush current, how can a larger capacitor help us solve this issue?

Regards

Jaume

Hi Jaume,

I was hoping it would be charged fast enough by the battery prior to the startup of the DC/DC converters but looks like this might not be the case. Something you could try is adding an RC filter on the EN pin of the TPS63802.Or are you able to pull the EN pin low for a brief period and then pull it high on the TPS device?

It may be good to create a thread related to the TPS63802 to verify that an RC filter would help delay the inrush current.

Best Regards

Hi Anthony,

I'll try to place a RC filter to the EN pin of the TPS63802. However, I'll create a new thread related to the TPS63802 and the effects that the RC filter could have on the IC.

Anyway, is there no method to help the BQ24075T not enter into short circuit protection mode? I mean by adding an external MOSFET, IC, which could reduce the inrush current

Regards

Jaume

Hi Anthony,

In any case, the RC filter in the EN pin did not help avoid this issue. Still the same behaviour of the charger

Regards

Jaume

Hi Jaume,

Apologies for the delay. There is no way to bypass the short circuit protection mode as of present. Ideally, we'd need to current limit the inrush from the TPS starting up, similar to how your shunt resistor worked. I do agree, an external FET would most likely work. 

Best Regards,

Anthony Pham

Hi Anthony,

Where the external FET should be placed? Between OUT and BAT?

If we place an external FET, SYSOFF signals would also be functional to turn off the device?

Regards

Jaume

Hi Jaume,

Anthony is currently out and will be returning on Wednesday. In the mean time I have read through the thread 

Jaume Forner Escrig said:

Where the external FET should be placed? Between OUT and BAT?

This appears to be what the solution is getting, allowing the FET to handle initial flow of current and then allow it to later be disabled so that the charger FETs can allow the IC to behave as intended.

Jaume Forner Escrig said:

If we place an external FET, SYSOFF signals would also be functional to turn off the device?

A PMOS FET that is tied to the SYSOFF signals might ensure that it is disabled when the SYSOFF signal is pulled high, however this is a solution I believe will require a bit further analysis to understand other implications on the system function if this FET is added in parallel.

Jaume Forner Escrig said:

When we have an external power input and then we remove it, we don't see this effect, the short circuit protection is not a problem in this case.

So when a power input is placed in parallel with the battery on assertion the device is able to turn on and continue functioning even when the power supply is removed. Is there a current limiter placed on the battery or some protection IC between the battery and the charger that might affect the batteries ability to source current?

Anthony previously mentioned, increasing the Battery capacitance may allow the SYS rail to draw enough current from the capacitor quickly enough that SYS does not fall Vo(sc) below BAT to trigger the short protection mechanism. Have you been able to try this solution?

Best Regards,

Juan Ospina

Hi Juan!!

Noted the solution with the MOSFET but as you say we may need to study the implications that it has.

Yes, I have tried increasing battery capacitance. I have tested with a 800mAh, 1050mAh and 2000mAh. Finally, with all of them we found the same behaviour. It's true that with the battery of 2000mAh stands up to 3.5V, but we are not still able to use the battery in its full range

Regards

Jaume

Hi Jaume,

Jaume Forner Escrig said:

I have tried to obtain the inrush current waveform by means of a shunt resistor (0.91Ω), but the device turns on properly when I connect any resistance.

I think it'd be very similar to what you did here but the discussed path from BAT to OUT also works! It may actually be more ideal depending on what you're ok with. 

With the solution, I think using the /PGOOD signal and/or the SYSOFF to control the gate of the FET. 

I can help explore this once I get a more ideal setup of what I have in mind.

Just so that I get a better understanding of your setup while I test a working solution, I understand that the behavior you're seeing happens when you plug in a battery. 

Does this also happen when the device is already on and the battery discharges down to 3.5 V?

Hi Anthony,

Yes, that's it. This behaviour occurs when you plug in a battery

Anthony P Pham said:

I understand that the behavior you're seeing happens when you plug in a battery.

No, if the device is already on we do not experience this behaviour

Anthony P Pham said:

Does this also happen when the device is already on and the battery discharges down to 3.5 V?

Regards

Jaume

Hey Jaume,

Thanks, just wanted to confirm. I'll try our a few things in the lab and try and get you a possible working solution soon. 

Best Regards,

Anthony Pham