BQ24780S: BQ24780S ACDRV don't turn on

Martin Andersson70

Part Number: BQ24780S

When applying a powercube with 5V at the AC input the ACOK goes high but ACDRV does not. At AC the voltage is 3,6 V and VCC to BQ24780S is 5V at ACDET we have 2.6V. When supplying Vbat it works as design, BATDRV goes high and we have supply to the circuit. We tried applying an external supply at CMSRC to see how the ACDRV charge pump would react, it followed the voltage of CMSRC and not 6 V as the datasheet states when ACOK is high. We also tested to sweep ACDET and did not manage to get the ACDRV charge pump to start.

over 7 years ago

0 Wang5577 over 7 years ago

TI__Guru 56650 points

Do you check it on an EVM or your board? Do you have 4.02k series resistor on CMSRC and ACDRV pin? If you do, please short or reduce that resistor to see any help.

0 Martin Andersson70 over 7 years ago in reply to Wang5577

Prodigy 140 points

Hello

We are using our own board. We are measuring on the BQ side of the resistor (4.02k) so shorting/lower value would not make a difference.

Regards

Martin

0 Wang5577 over 7 years ago in reply to Martin Andersson70

TI__Guru 56650 points

I am not saying the measurement issue. The 4.02k limits the CMSRC charge pump circuit current for ACDRV. If the input voltage is very low, ACDRV charge pump voltage could be very low and causes this issue. So, please short out 4.02k on CMSRC to verify this issue.

0 Martin Andersson70 over 7 years ago in reply to Wang5577

Prodigy 140 points

Hello

We have tried to short the 4k resistor (R5803) on CMSRC but this did not solve the problem. Both CMSRC and ACDRV are close to zero when 5V is applied to USB_VBUS_IN:

Regards Martin

0 Wang5577 over 7 years ago in reply to Martin Andersson70

TI__Guru 56650 points

If the input is 5V, how does the VCC voltage reach 5V? It should have a diode drop from the 5V input, right? What is VCC voltage when 5V is applied to USB_VBUS_IN? Please increase the USB_VBUS_IN to 10V to see the circuit work or not?

0 Martin Andersson70 over 7 years ago in reply to Wang5577

Prodigy 140 points

We have tried to apply 10V at VCC line (on C5804) but nothing changes. We can not apply 10V at USB_VBUS_IN since we will then damage the first buck converter (N5880).

0 Wang5577 over 7 years ago in reply to Martin Andersson70

TI__Guru 56650 points

datasheet list ACOK high condition:
1.VVCC_UVLOZ < VVCC< ACOVP
2. VACDET> 2.4 V
3. VVCC– VSRN> VVCC_SRN_FALL+ VVCC_SRN_HYST

Those conditions are also for ACDRV turn-on. Plus, we cannot overload the ACDRV pin: too much load (low resistance or high capacitance) between ACDRV and CMSRC.

Please check those conditions. Please order an EVM and check the component one by one until find the root cause.

0 Martin Andersson70 over 7 years ago in reply to Wang5577

Prodigy 140 points

Hello
We have measured that all these conditions (1-3) are met. Also since ACOK always goes high when we apply 5 V at Vbus and hence they should be met.

Regarding overloading we have 47nf between ACDRV and CMSRC, as you could see in previous added schematic . we believe we tried to remove this capacitor without any change

0 Martin Andersson70 over 7 years ago in reply to Martin Andersson70

Prodigy 140 points

Hello
when trying to get I2C (SMB) communication with the device we noticed that BQ24780S pulls SDA line low.
Regards Martin

0 Wang5577 over 7 years ago in reply to Martin Andersson70

TI__Guru 56650 points

I cannot point out the root cause right now. Please order an EVM and check the component one by one until find the root cause.

0 Martin Andersson70 over 7 years ago in reply to Wang5577

Prodigy 140 points

We have tried to remove the Mosfets N5800.

CMSRC were feed manually with a power box. When applying 3.9V, we got 3.9V out on ACDRV.
When applying 4.0V (or more) we got 10V out on ACDRV.

R5803 were shorted.

ACOK goes high when applying 5V at USB_VBUS_IN.

Before we removed N5800, there where no voltage at CMSRC (N5803) when we applied 5V at USB_VBUS_IN.

0 Martin Andersson70 over 7 years ago in reply to Martin Andersson70

Prodigy 140 points

We have contacted or TI contact to order a EVM

0 Wang5577 over 7 years ago in reply to Martin Andersson70

TI__Guru 56650 points

Great! You find the front end circuit issue. Yes. Without input source, CMSRC charge pump cannot provide power to ACDRV circuit.

0 Martin Andersson70 over 7 years ago in reply to Wang5577

Prodigy 140 points

Hello

How come neither application schematic in datasheet nor EVM has a connection for source to CMSRC, just connection to Source of ACDRV transistor, which does not goes high until ACDRV goes high?

We tested to modify the board according to:

Remove R5803 and connect C5804 to CMSRC pin. V5801 were also removed.

When 5V is applied to USB_VBUS_IN, we have 3.6V at L5880, 10.8V out from ACDRV. N5800 opens and we have 3.55V at R5804.

So this is our planed connection for our next build, please see attached schematic snippet with marked changes and let us know if you see a problem with this:

0 Martin Andersson70 over 7 years ago in reply to Martin Andersson70

Prodigy 140 points

Please be aware that it is very critical that we get input on this before tomorrow lunch (22/1) since we are releasing CAD then. especially important since this solution is outside of your reference schematic

0 Wang5577 over 7 years ago in reply to Martin Andersson70

TI__Guru 56650 points

In the beginning, ACDRV uses REGN voltage to pass power CMSRC. And then, CMSRC charge pump circuit supports ACDRV. I am not sure the potential issue with the connection, such as: input hot plug-in, CMSRC faces the input voltage directly; if "buck_out_3.6V" is low and "USB_VBUS_IN" is too high, will the ACDRV damage the FET or other circuit?

0 Martin Andersson70 over 7 years ago in reply to Wang5577

Prodigy 140 points

Do I understand you correctly that you consern is that ACDRV goes high while "buck_out_3.6V" is low causing at 10 V gate-source voltage on the FET.? But my understanding is that ACDRV vwill not go high until ADET goes high which it will only do if "buck_out_3.6V" is high

0 Wang5577 over 7 years ago in reply to Martin Andersson70

TI__Guru 56650 points

Yes. ACDRV will not go high until ACDET goes high. But, if the input is 20V, the ACDRV could be 26V.

Power management

Power management forum

BQ24780S: BQ24780S ACDRV don't turn on