BQ25710: VDPM stops regulator?

Hi Konstantin,

PROCHOT_VDPM triggers at a lower level than the regular VINDPM. please refer to datasheet 9.3.9. So I guess VDPM was active in your case. You can set up VDPM value very low to "disable" it.

Also keep in mind, at charge current as high as 3A in your case, voltage drop on cables starts to be significant. You may want to make sure you've minimized the loss between adapter and charger.

Best Regards,

Linhong

Hi Linhong,

I tried replying yesterday but it seems the post didn't  go through.

Thank you for responding quickly!

Mainly I am trying to set a max charge limit and charge as much as possible up to the voltage limit. Is that possible?

Rather than telling the charger "I want to charge at X rate", Can I say "Charge as fast as possible up to the X current limit, as long as voltage stays above Vmin"?

Thank you!

Hi Konstantin,

Yes! This is exactly what VDPM is designed for. Once you set up a proper VDPM(slightly lower than nominal input voltage but not enough to crash the input source), charger will adjust/reduce charge current, but keep it as high as possible without crashing the input source.

Best Regards,

Linhong

Hi Linhong,

Alright, that's good to hear. 

How can I reliably enter that state? 

I currently have a polling loop that monitors status bits and on ac detection, executes the following in this order:

1) Set input voltage (0x3D) to approx 4.1v,

2) Set charge current (0x14) to 448mA,

2) Set en_ico_mode to 1.

and now I'm in this state:

Dec 14 08:21:09 beaglebone kernel: bq257xx-charger 1-0009: REG0x12 : 0x70e
Dec 14 08:21:09 beaglebone kernel: bq257xx-charger 1-0009: REG0x14 : 0x12c0
Dec 14 08:21:09 beaglebone kernel: bq257xx-charger 1-0009: REG0x15 : 0x20d0
Dec 14 08:21:09 beaglebone kernel: bq257xx-charger 1-0009: REG0x30 : 0x210
Dec 14 08:21:09 beaglebone kernel: bq257xx-charger 1-0009: REG0x31 : 0x10b7
Dec 14 08:21:09 beaglebone kernel: bq257xx-charger 1-0009: REG0x32 : 0x830
Dec 14 08:21:09 beaglebone kernel: bq257xx-charger 1-0009: REG0x33 : 0x4a65
Dec 14 08:21:09 beaglebone kernel: bq257xx-charger 1-0009: REG0x34 : 0x8120
Dec 14 08:21:09 beaglebone kernel: bq257xx-charger 1-0009: REG0x35 : 0xe07f
Dec 14 08:21:09 beaglebone kernel: bq257xx-charger 1-0009: REG0x20 : 0x8400
Dec 14 08:21:09 beaglebone kernel: bq257xx-charger 1-0009: REG0x21 : 0xa880
Dec 14 08:21:09 beaglebone kernel: bq257xx-charger 1-0009: REG0x22 : 0xa00
Dec 14 08:21:09 beaglebone kernel: bq257xx-charger 1-0009: REG0x23 : 0x1d00
Dec 14 08:21:09 beaglebone kernel: bq257xx-charger 1-0009: REG0x24 : 0x0
Dec 14 08:21:09 beaglebone kernel: bq257xx-charger 1-0009: REG0x25 : 0x0
Dec 14 08:21:09 beaglebone kernel: bq257xx-charger 1-0009: REG0x26 : 0x3c3d
Dec 14 08:21:09 beaglebone kernel: bq257xx-charger 1-0009: REG0x3b : 0x744
Dec 14 08:21:09 beaglebone kernel: bq257xx-charger 1-0009: REG0x3c : 0x1400
Dec 14 08:21:09 beaglebone kernel: bq257xx-charger 1-0009: REG0x3d : 0x380
Dec 14 08:21:09 beaglebone kernel: bq257xx-charger 1-0009: REG0x3e : 0x1400
Dec 14 08:21:09 beaglebone kernel: bq257xx-charger 1-0009: REG0x3f : 0x4100
Dec 14 08:21:09 beaglebone kernel: bq257xx-charger 1-0009: REG0xfe : 0x40
Dec 14 08:21:09 beaglebone kernel: bq257xx-charger 1-0009: REG0xff : 0x89

Is there something wrong with this? Could Adc enabling during device initialization be interfering with charge regulation?

Thank you,

Konstantin

Hi Konstantin,

If I understand correctly, you were able to make 3A charging. I suggest you use this as a starting point, measure the actual input voltage seen by charger, then set VDMP slight lower than measured value, gradually increase ICHG to test if it triggers VDPM. repeat until you find your target VDMP value. 

Set up a VDMP higher than the "actual" input voltage will cause charging stop.

ADC doesn't matter in this case.

Best Regards,

Linhong

forgot to mention, according to your datalog, Reg 0x21, you were in PP_VDPM.

To avoid this, you want to start from a known working condition, then change incrementally, as described in my previous post.

Best Regards,

Linhong

Hi Linhong,

After leaving the charger connected overnight, it seems to have started charging at some point. Still have the pp_vdpm state set, but at least it's charging. It also starts charging relatively quickly after disconnecting and reconnecting:

Dec 15 01:26:39 beaglebone kernel: bq257xx-charger 1-0009: REG0xff : 0x89
Dec 15 01:26:39 beaglebone kernel: bq257xx-charger 1-0009: REG0xfe : 0x40
Dec 15 01:26:39 beaglebone kernel: bq257xx-charger 1-0009: REG0x3f : 0x4100
Dec 15 01:26:39 beaglebone kernel: bq257xx-charger 1-0009: REG0x3e : 0x1400
Dec 15 01:26:39 beaglebone kernel: bq257xx-charger 1-0009: REG0x3d : 0x380
Dec 15 01:26:39 beaglebone kernel: bq257xx-charger 1-0009: REG0x3c : 0x1400
Dec 15 01:26:39 beaglebone kernel: bq257xx-charger 1-0009: REG0x3b : 0x744
Dec 15 01:26:39 beaglebone kernel: bq257xx-charger 1-0009: REG0x26 : 0x5353
Dec 15 01:26:39 beaglebone kernel: bq257xx-charger 1-0009: REG0x25 : 0x1c00
Dec 15 01:26:39 beaglebone kernel: bq257xx-charger 1-0009: REG0x24 : 0x500
Dec 15 01:26:39 beaglebone kernel: bq257xx-charger 1-0009: REG0x23 : 0xd00
Dec 15 01:26:39 beaglebone kernel: bq257xx-charger 1-0009: REG0x22 : 0x4100
Dec 15 01:26:39 beaglebone kernel: bq257xx-charger 1-0009: REG0x21 : 0xa880
Dec 15 01:26:39 beaglebone kernel: bq257xx-charger 1-0009: REG0x20 : 0x9400
Dec 15 01:26:39 beaglebone kernel: bq257xx-charger 1-0009: REG0x35 : 0xc07f
Dec 15 01:26:39 beaglebone kernel: bq257xx-charger 1-0009: REG0x34 : 0x8120
Dec 15 01:26:39 beaglebone kernel: bq257xx-charger 1-0009: REG0x33 : 0x4a65
Dec 15 01:26:39 beaglebone kernel: bq257xx-charger 1-0009: REG0x32 : 0x30
Dec 15 01:26:39 beaglebone kernel: bq257xx-charger 1-0009: REG0x31 : 0x10b7
Dec 15 01:26:39 beaglebone kernel: bq257xx-charger 1-0009: REG0x30 : 0x210
Dec 15 01:26:39 beaglebone kernel: bq257xx-charger 1-0009: REG0x15 : 0x20d0
Dec 15 01:26:39 beaglebone kernel: bq257xx-charger 1-0009: REG0x14 : 0x12c0
Dec 15 01:26:39 beaglebone kernel: bq257xx-charger 1-0009: REG0x12 : 0x70e

Alright, my current understanding for what was happening is: 

Upon vbus connection, the charger resets it's vindpm to 1.28v below vbus, Later I detect vbis I set a higher vindpm. Then when I write in a charge rate, it shoots past the limit and hits the vdpm limit initializing some sort of slow charge detection. 

Is this a likely scenario? 

Also, If the pp_vdpm state is set, what is a good way to reset it?

Best,

Konstantin

Hi Konstantin,

According to Reg0x20, now you are officially in VINDPM. :-)

What you described makes sense, which means you are too close to your input source(adapter etc) limit(voltage drops too much with heavy load), or you have too much line resistance in cable. You may look if anything can be improved. On the other hand, you need to lower your VINDPM. You should be able to back to charging once condition is removed. no need to reset.

Best Regards,

Linhong

Yea, but it seems the stat_vdpm bit is still set there. 

It does seem that the charge very slowly starts ramping up over a period of at least a couple of hours.

At some point the In_vindpm bit start flickering on-off and eventually stays on.

But so far I have not found a way to clear that stat_vdpm bit.

So really I am curious as to what are the ways that Stat_vdpm is cleared? Also, Is it safe to ignore that bit? It seems like it does not really affect much, especially since I have not seen it set to 0,even at startup.

Also, Is it expected for it to take hours to achieve charge regulation? If not, what could be causing that?

Thank you!

Hi Konstantin,

No. Charge doesn't need hours to recover. Something else must have kicked in. Maybe fusb302 ??

To better debug this, I recommend remove all other components in the circuit, to isolate the problems. Then add back in piece by piece.

Best Regards,

Linhong 

Hmm, Fusb302 is actually not really part of the system. it is connected to the vbus line, but it has no control over it. It is also not enabled in my tests.

In  the design the Vin from a usb port is connected directly through the reference design with the source current limiting jumpers selected. 

Would you be able to say in what situations Stat_vdpm is cleared?

Best,

Konstantin

Hi Konstantin,

To make VINDPM out of your way, you can set VDPM to minimum and charge current to a smaller value(<< adapter capacity). Then you will never trigger VINDPM.

Best Regards,

Linhong

Hi Linhong,

So there is no way to reset the stat_vdpm bit without power-cycling? 

Best,

Konstantin

Hi Konstantin,

You can set FORCE_LATCHOFF =1 and reset the bit. 

Best Regards,

Linhong