BQ24618: BQ24618

Hi Christian,

I'm following up on the thread here:

I increased input voltage (VCC) of the charger IC to 15.5V.

Now, it should be more than sufficient to charge a battery pack to a maximum of 12.6V (three Li-ion cells/pack)
However, even now, charging terminates at the preset voltage (through Vfb divider at the output) of 12.3V
but then the charger still keeps switching between "suspended" and "charging" modes. 
The following scope capture shows the charger in suspend mode for long time, dropping into "charging" mode
for about 150mS, then going back to "suspend" mode, in an infinite loop. 
Initially the battery voltage is about 11.25V and the charger is in suspend state.
Then it starts charging and the battery voltage jumps to 12.275V, then charging ends and it goes back into 
suspend mode (and battery voltage drops to 11.25V).
It stays in suspend mode for about 1.6 seconds and starts charging again (because battery voltage drops
below RECH threshold)

It appears that the internal resistance of the battery has increased quite a bit (looking at the bump of 1V 
on battery voltage while charging).
But the charger should be toggling between "charging" and "charge complete" modes and
NOT between "charging" and "suspend" modes.
It is a mystery to us why the charger goes into suspend mode at all. 
There are some conditions listed in the datasheet for the IC to go into suspend mode
and none of those conditions are occurring in our system:

1. charge-en (CE) is pulled high (to disable charging)(not happening)

2. Power adapter is disconnected (UVLO, VCCLOWV or sleep mode happens)(not happening)
3. Power input has over-voltage condition (not happening)
4. Vref or REGN regulators (internal) are overloaded (not possible because there is no external loading on REGN and Vref regulators)
5. IC overtemp condition (145C) happens (not possible)
6. TS voltage (thermistor) goes out of range (i.e. battery is too hot or cold)(not possible, Vts=1.96V fixed, chrg ok)
7. timer times out (10 hours)(but suspend happens within 
   8. Charger goes into sleep mode (VCC falls below SRN  (not happening) or UVLO condition happens (not happening)
   
   So, basically there is no reason for the IC to go into suspend mode, yet it keeps going in and out of suspend mode.
   
   With a different battery pack (with much lower internal resistance), the charger transitions from "charging" state to
   "charge complete" and never between "charging" and "suspend" states. 
   
   So, the higher internal resistance of the battery pack is playing some role in the charger IC going into suspend mode.
   However, the datasheet does not explain why it can go into suspend mode. 

Thank you,

Owen

Hello Owen,

Are you using your own board or an EVM?

If you increase the battery charge voltage limit, do you see any difference. I think the device might be entering BATOVP.

Best Regards,

Christian.

We have the charger IC on our own board, but the circuit is very similar to the one in datasheet page-27 (Figure 19).
The battery pack (of 3x 4.2V cells) never charges to more than 11.9V even though the battery charge voltage limit is set to 12.35V
(for a battery pack of 12.6V max), so it is not likely to go into BATOVP condition. 
Also, when the external DC power (15V to 15.5V in this case)  is on, our system draws power from this power rail and the battery
just gets charged if charge level is lower than what is set.
When this condition (STAT1 constantly toggling high/low), the total duration is about 1.8 seconds, of which the charger is on for 150mS 
and off for the remaining time. Does this 150mS have any significance to the charger IC ? 
Looks like fast charge mode turns on, lasts for 150mS, goes off and the charger is in suspend mode, then turns on again, etc.
I have attached another scope capture which shows that there are a few level shifts in VBAT (output of the charger going to battery).
The red waveform is VBAT. 
When STAT1 transitions high-to-low (charging), VBAT jumps to about 12.25V, then drops to 11.3V when STAT1 transitions low-to-high (suspend).
But about 2/3rd way through suspend duration (about 1 second), VBAT jumps up by about 200mV, then at the end of suspend duration,
STAT1 goes low again and charging starts again. I am not sure whether that is because it goes into precharge mode first
(when STAT1 is still indicating suspend mode) and then goes into fast charge mode when STAT1 transitions low.
I am not sure what determines the duration 1.85sec for the entire cycle,with the charger in suspend mode for about 1.7sec and in fast charge mode 
for 150mSec. If VBAT reaches the set-point (through Vfb pin) it should indicate "charge complete" (STAT2=0, STAT1=1) and not "suspend"

Hello Anoop,

Owen Mears said:

1. charge-en (CE) is pulled high (to disable charging)(not happening)

This should be pulled low to disable charging.

Do you have a load on your system?

Best Regards,

Christian.

Sorry, I made a mistake with polarity of CE signal.
What I meant is that charger can go into suspend mode if CE is driven to disable charging but that is not happening in our systems.
In our systems, the CE pin is always high, so charging is always enabled when external PSU is providing power.
Our system draws between 20-25 Watts, so the load is always present. But when external PSU is providing power,
no power is drawn from battery and the battery is charged until charging is completed (or gets suspended for any reason).
Is there any requirement as to when charging can be enabled by the system through CE pin ?
Or can we keep it permanently in enabled state ?

My concern is that on battery packs with higher internal resistance, the charger keeps going between charge and suspend states
about every 1.8 seconds. It should be going to "charge complete" state and not keep toggling like this. 
With battery packs with lower internal resistance, charging completes and the charger stays in that state. 
I measured internal resistance of four battery packs, one known good pack vs three used packs (that were in our RMA units)
and the failing ones have between 30% to 70%  higher internal resistance (170 mOhms vs 220-290 mOhms)
(also, when charging gets suspended, these battery packs have open circuit voltage between 11.3V and 11.9V v/s good pack
  that charges all the way to 12.25V)
The RMA products work fine when I insert a battery pack with lower internal resistance, indicating that there is no inherent
manufacturing or other defect in these units, it is only the battery pack. 

How does internal resistance of a battery pack affect the charger going into suspend state instead of charge-complete state ?

Hello Anoop,

Anoop Hegde said:

Is there any requirement as to when charging can be enabled by the system through CE pin ?
Or can we keep it permanently in enabled state ?

This can be permanently enabled.

Anoop Hegde said:

My concern is that on battery packs with higher internal resistance, the charger keeps going between charge and suspend states
about every 1.8 seconds. It should be going to "charge complete" state and not keep toggling like this. 
With battery packs with lower internal resistance, charging completes and the charger stays in that state. 
I measured internal resistance of four battery packs, one known good pack vs three used packs (that were in our RMA units)
and the failing ones have between 30% to 70%  higher internal resistance (170 mOhms vs 220-290 mOhms)
(also, when charging gets suspended, these battery packs have open circuit voltage between 11.3V and 11.9V v/s good pack
  that charges all the way to 12.25V)
The RMA products work fine when I insert a battery pack with lower internal resistance, indicating that there is no inherent
manufacturing or other defect in these units, it is only the battery pack. 

I have to debug this futher, I'm gathering more information and will provide and update after.

Hello Anoop,

Can you provide your schematic?

Best Regards,

Christian.

Please see attached.

Gigamon_BQ-charger-schematic.pdf

Hello Anoop,

Do you see this issue if you increase the input voltage?

Can you provide a scope capture of VCC, SRN, Sysstem voltage, ACDRV, measure close to the pins.

Best Regards,

Christian