Glo Glo Battery Problems

[Peakcrew]

Oh, well - worth a try!

I'll be interested to see what David's test for screen-touching comes back with.

[rcjoop]

My kobo glo is less than 2 m. from the wireless router with poor battery performance (< 3 days)

[Peakcrew]

Originally Posted by rcjoop

My kobo glo is less than 2 m. from the wireless router with poor battery performance (< 3 days)

It looks like this is a dead end - well, not even that, really. It was just a thought generated by moving to a newer place where the wifi signal degrades dramatically across the building, and having left my phone's wifi on one night - almost half the battery gone in eight hours!

It is frustrating, this. Since Kobo have eliminated the battery issue (and we have no choice but to trust them on that), there seems to be an environmental issue that affects some but not others. Two or three broken units shipped to the same people is an unlikely chance - I had a similar problem with an electric toothbrush a couple of years ago: to all intents and purposes I had three units with faulty chargers. Only when I was sent an upgrade to one with a charge indicator did I realise that there was an environmental issue - the pins on the mains plug didn't properly connect in the socket, though the two testers I'd used claimed all was well. (I did apologise to the very helpful folks at the service department!)

[ghudod]

I think the wifi thing is a red herring. The further your Kobo is from a good wifi signal, the more power you will use as your Kobo will have to transmit back across the same distance/interference. This will make your battery go down much faster, and I don't think it's related to the sleep issue.

I generally only turn my Glo off rather than sending it to sleep, and I've had it last for 10 days reading 3-4 hours a day on average. I have seen occasions where I've allowed the Glo to sleep, and then turn itself off, and then a few hours later, I turn it back on to see a chunk taken off the battery ie, more than if the Glo had been on and being used to read books.

So I think there is an issue with the Glo going automatically from "sleep" to "off". I think it's not only not switching off properly, but that it's getting itself stuck into some kind of loop or race condition where it's actually burning more power than normal, possibly running the CPU at 100 percent instead of going off. I think it's a bug somewhere in the firmware that can be fixed by an update.

[Mrs_Often]

Originally Posted by ghudod

So I think there is an issue with the Glo going automatically from "sleep" to "off". I think it's not only not switching off properly, but that it's getting itself stuck into some kind of loop or race condition where it's actually burning more power than normal, possibly running the CPU at 100 percent instead of going off.

That seemed to be the suspicion of Shai-Hulud as well... but, it doesn't explain the battery drain for readers that are in sleep mode where auto turning off is set to never.

Originally Posted by ghudod

I think it's a bug somewhere in the firmware that can be fixed by an update.

I hope so. But, if it is, why does it only happen to some people and not others?

[Mrs_Often]

Originally Posted by rcjoop

My kobo glo is less than 2 m. from the wireless router with poor battery performance (< 3 days)

Hi rcjoop, welcome to the forum and sorry to hear you're also experiencing the battery issue. I'm glad the problem front is growing though (sorry). Have you let Kobo (or Sameer) know you're also one of the unlucky?

[davidfor]

Originally Posted by Peakcrew

Oh, well - worth a try!

I'll be interested to see what David's test for screen-touching comes back with.

Would you believe no battery use?

After posting last night, I changed the sleep and light off settings to "Never". Then I left the Glo sitting on the table near me. At that point the battery was showing 100%. An hour or so later I realised I had not turned the light on, so I did. It sat there for a couple of hours with occasional pickups to look at something (I think the sort order of sideloaded books). When I finished at the computer, I read for about 30 minutes, checked the battery lever and put the Glo back on the table. The battery level was 97%. The Glo was left leaning against something, so unless a moth was attracted to the light or the cat decided to sleep on the mouspad, it wasn't going to do anything.

This morning, about 8 hours and 15 minutes later, I checked the battery level. Still 97%. But, I pressed the screen a few times, closing and reopening the book and the battery level dropped to 95% and then 94%.

I'm going to take the 94% level. It feels like the device needs some activity to update the battery level.

As I won't be reading this morning, I am leaving the Glo on with something touching the screen. It should get two or three hours like that, so I can see if it appears to reproduce the results from the other night.

[Peakcrew]

Originally Posted by davidfor

Would you believe no battery use?

After posting last night, I changed the sleep and light off settings to "Never". Then I left the Glo sitting on the table near me. At that point the battery was showing 100%. An hour or so later I realised I had not turned the light on, so I did. ... It should get two or three hours like that, so I can see if it appears to reproduce the results from the other night.

Actually, you have just reminded me of something I thought of this afternoon - I haven't seen a report from someone whose Glo works properly regarding how long the battery lasts with a) the light left on at different levels and b) with wifi left on. Also, has anyone reported whether the Glo gets warm anywhere - if there is something running constantly to the extent of running a battery flat in about a tenth of the time expected from a healthy one, some component should be be generating heat - perhaps enough to feel through the case? I can tell when I've left my phone's GPS on by temperature alone, for instance.

I'll be back at about 5am UK time - F1 qualifying delayed from today!

[davidfor]

Originally Posted by Rev. Bob

To chip in my tuppence, as I've just been dealing with moving sets of EPUBs onto my new Glo/2.4.0 - after a fresh load, that section of the Recent Reads order is in straight Z-A order by raw book title, regardless of directory, filename, or opf:file-as metadata. (I say Z-A order because, of course, the Most Recent files are the ones last processed, so the Zs appear at the top of that list.)

Caveat: This is manually loading books onto the internal memory, not using Calibre or other software.

You're right. It looks like there has been a change since the last time I did this. Or at least the last time I paid attention.

I just put my test SD card into the Touch. This card has about 550 fairly random books. They are all either epubs or PDFs. It was populated some time in the past while it was in the Touch using calibre. With a few more added occasionally since. So the path for the books is "{author_sort}/{title} - {authors}". The timestamps on the books and folders are nearly all in five minute range.

My investigations into what shows on the home screen before was that it is based on reading status (unread or being read) and two timestamp columns. These are "DateLastRead" and "___SyncTime". The "DateLastRead" is used if it is set, otherwise it is "___SyncTime". And for a sideloaded book, "___SyncTime" is when the book was processed. As the books have all just been added, only "___SyncTime" is used.

My memory from when I did this before was that the books were listed in the order returned with "ORDER BY ___SyncTime DESC". And because of the way the rows are added to the table, this effectively meant the last five rows in the table that matched the criteria. It looks like this has changed to "ORDER BY ___SyncTime DESC, Title DESC".

The other thing that goes with all this is the order the books are processed in. This isn't important unless there is a book that is breaking the processing and preventing other books being added. I am sure that with firmware 2.0.0, the order was the fully qualified file name, which made finding the bad book fairly easy. But, looking at it now, it definitely isn't. Or at least, that isn't the order that they are put into the database. It is also not file timestamp or any other obvious value.

[TechniSol]

Probably doesn't amount to a hill of beans, but one obvious difference between my Glo and Mrs. Often's is that mine is set for EST and US date format(mmddyyyy)...

As for the batteries being ruled out, I'm kind of curious how one rules out a negative unless they already had the batteries in the original units in question to test? Even testing existing stock wouldn't rule out a few bad apples in a barrel... (I really hadn't thought about it before, just accepted it at face value.) Of course, it does beg the question of why the replacements had the same problem...

If I were one of the boys at the big K I'd be really curious to have a look at the rest of the stock from where Mrs.O's and other's malfunctioning replacements were pulled... unless, of course, I already had a clue.

[rcjoop]

Originally Posted by Mrs_Often

Hi rcjoop, welcome to the forum and sorry to hear you're also experiencing the battery issue. I'm glad the problem front is growing though (sorry). Have you let Kobo (or Sameer) know you're also one of the unlucky?

Just sent an email, had not done it yet.

The idea that there might be a problem with going from sleep to power off seems unlikely in my case as this option was never turned on.
I only manually powered down twice to check if that solved the problem, like in the old windows days: just reboot and all is well again for a while.
No such luck of course.

[davidfor]

Originally Posted by davidfor

As I won't be reading this morning, I am leaving the Glo on with something touching the screen. It should get two or three hours like that, so I can see if it appears to reproduce the results from the other night.

After about four and half hours of this, with a 15 minute read in the middle, the battery charge is at 87%. That is about a 6% drop in that time and roughly agrees with what I saw the night before last with the accidental all night session.

[Peakcrew]

Originally Posted by TechniSol

Probably doesn't amount to a hill of beans, but one obvious difference between my Glo and Mrs. Often's is that mine is set for EST and US date format(mmddyyyy)...

If I recall correctly from reading the whole thread yesterday, the battery problem was being seen before the date-format change was seen.

As for the batteries being ruled out, I'm kind of curious how one rules out a negative unless they already had the batteries in the original units in question to test? Even testing existing stock wouldn't rule out a few bad apples in a barrel... (I really hadn't thought about it before, just accepted it at face value.) Of course, it does beg the question of why the replacements had the same problem...

That was what I was getting at, though I didn't word it that way. On this one, we have to trust Kobo unless people are willing to pull the back off their Glos to look at battery serial numbers etc. However, manufacturers have been known to deny that there is a problem with batteries until after several have burst into flames, leading me back to my point about components heating - at the discharge rate being seen, the battery should be getting warm. (Hmmmm - just had a thought: I doubt many people have infra-red thermometers at home, but would the liquid crystal strip thermometers show the temperature of the back-plate? If so, my thinking is that, if the issue is over-active electronics, there will be a higher temperature shown; on the other hand, if it is, say, incomplete charging for some reason, then there won't be such an elevated temp. Of course, some standardisation would be needed (location of strip, ambient temp etc). On the other hand, I might just be barking up the wrong tree, or in the wrong forest altogether )

If I were one of the boys at the big K I'd be really curious to have a look at the rest of the stock from where Mrs.O's and other's malfunctioning replacements were pulled... unless, of course, I already had a clue.

That's the first place I'd look too - even if just at the parts database to see if there are any common batch numbers. I'll have to go through the thread to check, but don't we have one person with three defective units located somewhere in the USA, and Mrs_Often with two in the Netherlands, and some other reports of single defective units from Europe? Of course, that doesn't mean they were bought in those places ... didn't Thalinor say s/he can't buy one directly in her/his part of the USA?

[Mrs_Often]

Originally Posted by Peakcrew

It is frustrating, this. Since Kobo have eliminated the battery issue (and we have no choice but to trust them on that),

Actually, the battery (nor the hardware) was ever ruled out at all. What Sameer said was it is unlikely to be the battery.

Originally Posted by TechniSol

Probably doesn't amount to a hill of beans, but one obvious difference between my Glo and Mrs. Often's is that mine is set for EST and US date format(mmddyyyy)...

Correct

[Mrs_Often]

Originally Posted by davidfor

After about four and half hours of this, with a 15 minute read in the middle, the battery charge is at 87%. That is about a 6% drop in that time and roughly agrees with what I saw the night before last with the accidental all night session.

Also sounds very much like the way my Glo sleeps... Maybe I should sleep it outside of its sleeve for a while.