Possible Charging Bug? Just Had This Happen.

[mali]

^ I'm not sure if charging works via the OTG port. If you got an adaptor you could try it on the full sized port.

 

[peca]

I'm pretty sure, that USB OTG can be used for Pandora's charging. MWeston mentioned charging through "USB input", which I guess is the USB OTG:

...Currently, Notaz and I have chosen 1A on the AC adapter and 600mA from the USB input...

As MWeston says 600 mA through USB, then my 850 mA adapter should be enough.

 

[tsh]

Charging through the USB-B port appears to push about 200mA into the battery in basic 'on but idle' state. Whilst trying the sd read | zip, I reached breakeven - not charging or dis-charging. The full size port is not the one to use for charging!

It may depend on the USB connection being correctly configured though (there is a pin required to be tied if high current can be taken without a software handshake).

Edit: Does anyone have the data sheet for the PMIC? I thought the part might be twl4030, but I've drawn a blank finding data on a part with that number.

 

[urjaman]

Edit: Does anyone have the data sheet for the PMIC? I thought the part might be twl4030, but I've drawn a blank finding data on a part with that number.

IIRC It's TWL5030's (TWL4030's successor) general availability version called TPS65950. Look here.

 

[Prometheus]

Pro, I never had any such issue with either of my units when using a mains charger. I too have WiFi disabled on boot, and, on my old unit, had a background loading from an SD card.

Alas, I'm not going to be able to try to get to the bottom of this one now. The left nub on my second unit has failed in the same way as the first, unfortunately (again, after being used only for mousing), so I need to return it again.

 

[mali]

^ That's bad It might be best to wait a few days before returning it. The new boards with the factory-tested nubs might be more reliable, eventually.

 

[tsh]

Alas, I'm not going to be able to try to get to the bottom of this one now. The left nub on my second unit has failed in the same way as the first, unfortunately (again, after being used only for mousing), so I need to return it again.

Make sure you give details of the charging problem, it is probably worth geting someone to investigate this in case it is a more general problem.

 

[Prometheus]

^ That's bad It might be best to wait a few days before returning it. The new boards with the factory-tested nubs might be more reliable, eventually.

Jacquelyn said she'd get back to me with what the turnaround time is, so I haven't been asked to return it just yet.

Alas, I'm not going to be able to try to get to the bottom of this one now. The left nub on my second unit has failed in the same way as the first, unfortunately (again, after being used only for mousing), so I need to return it again.

Make sure you give details of the charging problem, it is probably worth geting someone to investigate this in case it is a more general problem.

Good call - I never thought of that. It will be added to my covering note.

 

[MWeston]

^ I'm not sure if charging works via the OTG port. If you got an adaptor you could try it on the full sized port.

I see many people trusting your words as things to try when they have problems so you should definitely choose your words more carefully. The OTG port has a peripheral mode which allows power input, hence the ability to charge. A host port is designed by USB spec to be OUTPUT ONLY, therefore trying to rig up some cable to try and stuff power into the host port has the potential to damage all devices. Don't try to charge Pandora from the host port for the love of all that's holy!

Pandora mine charging good - from bundled charger. But today I try charge it by mobile phone charger (mini USB B plug), but have no response on plug/unplug charger.
...
When I plug it to Pandora, I can't see any response from OS

Whenever I see these topics, I always try to repeat the problem myself and I can't simulate the same problems although I have found charging won't start for me if I have a big load in the host port. If I start charging and then plug in the load, it works fine so some logic in the PMIC is deciding not to charge under certain circumstances.

The thermistor for the battery is a standard 10Kohm, NTC, B=3435K which is pretty close to the default calibration value in the TPS65950 if I remember correctly but I don't know if it is being interpreted right. It's been a while since I used the kernel output to see what temp was but it was way out to lunch the last time I checked many months ago. Maybe it spat out some sort of raw numbers and I didn't know how to convert them too. I tend to look at the fuel gauge temp since it is more accurate but the kernel and charging circuit do not use this reading for anything (only the gauge uses it). I would think the PMIC hardware itself would respond properly to the 10Kohm resistance in the battery pack regardless of OS settings, but it's something to check.

As for the USB port, since it doesn't pull much current (600mA right now), if you were streaming audio over wifi with the screen on bright or something like that, there might not be any current left to charge the battery. Under that circumstance, the battery could still drop in percentage. I think the charge LED would still light up though. The light always comes on for me whenever I tried it but haven't tried it lately with all these hotfixes (although ED says it shouldn't matter).

That PMIC still has some secrets to unlock or maybe it's just a buggy, errata full piece of shit! I don't know but it does work and the hardware design is the standard reference design so no future board revisions will change anything. I wish we had more people interested in looking at the charging driver and seeing what they could contribute. Notaz is the main/only guy on that and he says the driver looks to be full of half finished features so this is one area of concern that could use some more eyes.

 

[mali]

^ Sorry, I will only give advice when I'm sure about it from now on. As I'm no USB expert I shouldn't have posted dangerous guesswork.

 

[Evil Star]

Alas, I'm not going to be able to try to get to the bottom of this one now. The left nub on my second unit has failed in the same way as the first, unfortunately (again, after being used only for mousing), so I need to return it again.

The nub factory better get their shit together before my Pandora is shipped :blink:

 

[tsh]

That PMIC still has some secrets to unlock or maybe it's just a buggy, errata full piece of shit! I don't know but it does work and the hardware design is the standard reference design so no future board revisions will change anything. I wish we had more people interested in looking at the charging driver and seeing what they could contribute. Notaz is the main/only guy on that and he says the driver looks to be full of half finished features so this is one area of concern that could use some more eyes.

Yes, I looked at the driver and wondered if it's all like that! If there is anything specific, maybe I can have a poke around, but the kernel stuff is all pretty new to me.

 

[hdonk]

I've found what I believe to be a workaround for the lack of charging issue I've been having. If no charging light comes on, or the power bar remains green after plugging in the external PSU, simply put the andora2ut4: in to low power mode by flicking the power switch, close the lid, wait until the power light comes on (unplugging it first if it was already on), then open up & flick back to full speed.

All these steps may not be necessary, but they appear to sort out the issue I've been having.

 

[EvilDragon]

Alas, I'm not going to be able to try to get to the bottom of this one now. The left nub on my second unit has failed in the same way as the first, unfortunately (again, after being used only for mousing), so I need to return it again.

The nub factory better get their shit together before my Pandora is shipped :blink:

That's what they're currently doing.
Testing EVERY single one of them... fixing or destroying the bad ones (depending on what's cheaper for them ) and producing new ones for free.
They admitted something has gone wrong with mass production (geeez) and too many of them are bad. They do this for free (thank god) and they're pretty fast testing them

Some nubs failed (about 20 Pandoras from 500 delivered units we confirmed to be working good got nub problems after a while... and there are TWO nubs on each unit, so it's nothing to panic about), but those will be replaced.
Shouldn't have happened - and the prototypes were fine... but we will replace those failing nubs and in the meantime the manufacturer is taking care of increasing the quality.

We're annoyed by those nub issues, too... but it's no catastrophe, and the problem is already solved. Adds a lot of work and some shipping costs for us... but we'll manage that

Now just think about what happened if we assembled all 4000 units at once... and not ourselves, but in Texas or China... where nobody would've cared about if stuff was working or not... THAT would've been a catastrophe

 

[Prometheus]

ED, it makes me very happy to hear that. I'm glad that the nub factory is doing the right thing about the problem, too.

 

[alucart1984]

Alas, I'm not going to be able to try to get to the bottom of this one now. The left nub on my second unit has failed in the same way as the first, unfortunately (again, after being used only for mousing), so I need to return it again.

The nub factory better get their shit together before my Pandora is shipped :blink:

That's what they're currently doing.
Testing EVERY single one of them... fixing or destroying the bad ones (depending on what's cheaper for them ) and producing new ones for free.
They admitted something has gone wrong with mass production (geeez) and too many of them are bad. They do this for free (thank god) and they're pretty fast testing them

Some nubs failed (about 20 Pandoras from 500 delivered units we confirmed to be working good got nub problems after a while... and there are TWO nubs on each unit, so it's nothing to panic about), but those will be replaced.
Shouldn't have happened - and the prototypes were fine... but we will replace those failing nubs and in the meantime the manufacturer is taking care of increasing the quality.

We're annoyed by those nub issues, too... but it's no catastrophe, and the problem is already solved. Adds a lot of work and some shipping costs for us... but we'll manage that

Now just think about what happened if we assembled all 4000 units at once... and not ourselves, but in Texas or China... where nobody would've cared about if stuff was working or not... THAT would've been a catastrophe

This is very good news. One day I will get a Pandora so I want to make sure when I do decide to get one that it will be in working order for sure.

 

[joseluisjazz]

I've found what I believe to be a workaround for the lack of charging issue I've been having. If no charging light comes on, or the power bar remains green after plugging in the external PSU, simply put the andora2ut4: in to low power mode by flicking the power switch, close the lid, wait until the power light comes on (unplugging it first if it was already on), then open up & flick back to full speed.

Thankyou thankyou thankyou hdonk. I couldn't charge my pandora any more by any means and your instructions fixed it!!