BQ25180: BQ25180 Voltage Ripple on SYS

Hi Christian,

Thanks for reaching out.

It would be helpful to see scope captures of this behavior. I'll follow up with you over E2E private message about this.

In the meantime, I have a few questions for initial debugging of this issue:

Could you please share the register configuration of the device?

Just to confirm, the ripple behavior only occurs after removing the device from the wireless charging pad?

What is your expected system load current? Are you testing with resistors or the actual system load on your PCB?

Best regards,

Alec

Hi Alec,

Thanks for your reply.

 The configuration of the I2C registers is as follows:

So far, I could only reproduce this issue by removing the charging pad, which means when the power source for SYS changes from IN to BAT. Initially this issue occurred with our electronic as the load on SYS, but in the meantime, I have isolated the charger chip part from the electronics. On SYS is now only an ohmic resistor decade as the load.

I also uploaded some oscilloscope pictures. If you see the SYS ripple, the channel is on AC coupling. On the picture BQ25180_Issue.png, the yellow VSYS Signal is on DC coupling.

Thanks and best regards

Christian

1k load:

10k load:

100k load:

100R load:

BQ25180_Issue:

Hi Christian,

Thank you for sharing all of this information. This is very helpful. I'll review this and follow up with you tomorrow, 12/10.

Best regards,

Alec

Hi Christian,

Could you zoom in on VIN, VBAT, and VSYS at the moment the device is removed from the wireless pad?

In the BQ25180_Issue capture, it looks like VSYS on the DC-coupled channel stays at the same voltage the entire time. From the register settings you shared, VSYS is regulated to 4.6V when a 5V input source is connected. When the device is removed from the wireless pad, I would expect VSYS to fall from 4.6V to the battery voltage. A zoomed in capture would help confirm whether this is the case.

It looks like the ripple increases a few seconds after the device is removed from the wireless pad. Is anything happening during this time? Is the load on the decade box being increased, or does this ripple always just appear after a few seconds?

Best regards,

Alec

Hi Alec,

Sorry I was not writing it correctly, in the picture BQ25180_Issue the load is still our electronics, not an ohmic load. After removing the charging pad, the oscillation starts, but the electronics is still running and the current is high, therefor the oscillation frequency is higher. After around 2.5 seconds our device goes into a sleep mode, the load current decreases and the oscillation frequency changes. The Amplitude of the oscillation looks small in the first 2.5 seconds and much bigger after that, but this is just a resolution problem of the saved .png file.

VBAT was quite high in the picture and you just don't see the jump from 4.6V to VBAT of the VSYS voltage, also a resolution problem.

I made new pictures where you can see it in a better way. You can see, the ripple starts right away and the VSYS (Yellow) jumps from 4.6V to VBAT.
To me it really looks like the power source switch from IN to BAT starts the oscillation. I also added our Schematic part of the BQ25180.

Thanks and best regards

Christian

Hi Alec

To clarify your question, the issue does still occur whether I replace my downstream load with a resistor decade box. The load on SYS cannot be pulsed because the only load then is the resistor decade box.

The Ripple shape with only the resistor decade box as the load looks like this:

best regards

Christian

Hi Christian,

Just to confirm, is the MCU still connected to SYS in these waveforms? If the ripple increases when the MCU goes into sleep mode around 2.5 seconds after removing the input source, it seems the ripple could be related to the MCU.

I did some testing on a BQ25180 EVM using a resistor decade box connected to SYS and your register settings. I did not see any ripple after disconnecting the input source. How repeatable is this issue on your end?

Best regards,

Alec

Hi Alec

I could also reproduce this issue with the BQ25180EVM Eval Board.

I did not change any registers, I just took the EVALBoard out of the box.

I just did the following:
I connected a LiPo Battery to the VBAT and GND pins.
I connected a 1k resistor between the VSYS and GND pins.
I connected 3x1.5V AA Alkaline Batteries in series to the VIN and GND pins.

I disconnected the Alkaline Batteries from the VIN pin. I had to do this several times and suddenly the ripple was there!

I measured VSYS with AC Coupling and 20mV/DIV and 50us/DIV:

VIN Voltage = green
VSYS AC Coupling = red
VBAT Voltage = blue
VSYS DC Coupling = yellow

This is the ripple after disconnecting the VIN voltage (zoomed in):

and a little bit later:

Please try it again to reproduce the issue with your EVALBoard.

Thanks and best regards,

Christian 

Hi Christian,

That's great that you were able to reproduce the ripple on the EVM. That should really help narrow this down. I'll try to replicate it on my side and follow up with you tomorrow.

Best regards,

Alec

Hi Christian,

Thanks for your patience. I'm still investigating and will follow up early next week with an update.

Best regards,

Alec

Hi Alec,

Thanks for you reply, sounds "good", looks like this BQ25180 chip has a general issue.
I already checked the new chip version BQ25188 and I could not reproduce the issue with this version.
But a change to the newer chip version would delay our project by a few months.

The thing is, we are in the final design phase and market launch is planed very soon. 
changing to the new chip version would mean to start our validation all over again.
And beside that, we already have many thousands of the BQ25180 chip on stock.

Is there any workaround or solution with the older chip?
Best solution would be a pure Software change, like changing the chip registers.

On the other hand, can you confirm that the new chip version BQ25188 does not have this issue?

Thanks and best regards

Christian

Hello,

We are currently out of office for the Holidays, we will respond as soon as possible. Sorry for the delay.

Sincerely,

Wyatt Keller

Hi Alec,

Happy new year to you too.

I already tried your suggestion but unfortunately it did not help, the ripple still occurs.
I removed the 10uF cap C38, this did not help.
Then I even replaced C38 with a "pulldown" 10k resistor, this also did not help.

I also could see the issue when VBAT was not close to VIN, VBAT was around 3.7V and VIN 5V.

Even on the EVAL Board BQ25180EVM there is 4.7uF and 0.1uF at the IN pin.

I hope there are more things I can try.
Maybe we cannot make the issue disappear but make it very rarely...

Thanks and best regards
Christian

Hi Christian,

Thanks for trying that. I'm investigating workarounds for this specific behavior and will follow up with you by the end of the week.

In the meantime, does the ripple ever go away on its own after some time? And just to confirm, you can still see the ripple with C38 removed and VIN=5V?

Best regards,

Alec

Hi Alec,

Removing C38 has no impact. As soon as VIN=5V the Ripple disappears.
The ripple affects the functionality of our product. We can observe negative effects for days after we noticed the issue. Also, answer to your question - the ripple does not go away by itself.

By the way - another question on differences between BQ25180 vs. BQ25188.

Table 8-6 in BQ25180 states that pushbutton input is available in battery mode (if enabled).
Table 7-6 in BQ25188 says the pushbutton is not available in battery mode.
Description of bits 6-5 and 4-3 of SHIP_RST is different for both products. BQ25188.
This makes us not sure...

Does BQ25188 wake up from ship mode just with a pushbutton in battery mode (VIN = 0V)?

Best regards
Christian

Hi Christian,

Thanks for your patience. At this point, I haven't been able to find a reliable software workaround for this behavior on the BQ25180. I can continue debugging, but just to set expectations, it may take some time to fully understand the conditions and find a solution.

Would a BOM change be possible? I'll do more investigating and follow up with you early next week. Any workaround would like likely be hardware-related.

If possible, I do recommend moving towards the BQ25188.

And yes, the BQ25188 can wake up from ship mode with just a pushbutton in battery mode as long as the EN_PUSH bit is 1.

Best regards,

Alec

Hi Alec

Thanks for your investigations and answers.
At the moment we also consider to switch to the new chip BQ25188, if there is no workaround with the older chip.

Just to let you know, it would also be enough for us to just detect the ripple failure mode of the chip, maybe some status registers change in the failure mode... in that case we can ask our customers to reset the device to remove the ripple, I hope you understand what I mean.

About the Button Wakeup procedure, it seems to work with the new Chip BQ25188, but not the same way as with the older chip. Is there something different like status register or similar?

Best regards
Christian

Hi Christian,

We have a better understanding now of what's going on and are still confirming a few details.

The behavior is related to the SYS load current during the transition from VIN to battery power. We're checking whether this can point us to a workaround.

Could you share the expected system load of your application?

I can also confirm that this behavior is not present on the BQ25188.

Could you clarify the pushbutton behavior you're seeing? Are you referring to waking the device from ship mode?

Best regards,

Alec

Hi Alec

Sounds good.
The expected system load is 20-40mA.
The push button wakeup seems to work, I just need to test more, I will get back to you if we run into a problem there.

Best regards
Christian

Hi Christian,

Thanks for sharing your expected system load. I'm checking on this and will follow up once I have an update.

Sounds good on the pushbutton. Let me know if you run into any issues there.

Best regards,

Alec