Battery fault (expanding battery) / replacement

[mbecroft 82]

I woke up this morning to find my pro1 has expanded (case come unclipped along front of keyboard).

I have not opened the shell to check but almost certainly it must be the battery. Has anyone else had an expanding battery?

Is there any source of replacement batteries?

Has anyone successfully replaced the battery?

[VaZso 1,998]

I haven't heard of any battery problems yet but the battery can be replaced relatively easily.

However, I can not provide you any replacement battery sources except what you may find at their Indiegogo site (and which claims delivery next February if that is true).

Anyway, currently I don't remember what an expanding battery may destruct inside (physically) but it may be a good idea to remove it before causing additional problems in the device...

[mbecroft 82]

Thanks for the advice & encouragement - I found the teardown guide thread and github (https://github.com/imax9000/fxtec-pro1-teardown) - as you say, does not look too bad.

It may be something else, but phone expanding for no reason my first thought is always the battery! So definitely will open it and check, and if it is indeed the battery, nurse it along with no fast-charging (I am actually annoyed that it fast-charges, I don't need this, and it wears out the battery). I see the spare battery on Indiegogo, do you think the Pro1X spare battery is definitely the same as for the Pro1? Knowing fx timelines it could be months before there are more batteries!!

Other option is to find any cell with close enough form and fit, and swap the electronics from the old to the new battery - I've done this many times.

There is also one fx for sale on ebay right now, the price is insane, but I am seriously thinking of buying it just so I have a spare, since I rely on this phone a lot for both home and work - so expense is justified! Anyway...will see once I open the phone!

Cheers -Mario

[mbecroft 82]

Well, damn. It most certainly *is* the battery...

[Slion 1,201]

2 hours ago, mbecroft said:

Well, damn. It most certainly *is* the battery...

You should really decommission that phone. Open it up, remove the battery, assess the damages and ask fxtec support for spare parts. Use another phone until you can get this fixed one way or another. Order a Pro1X if not already done. Then wait 2 years for it to be delivered 😭

[mbecroft 82]

@SlionSadly all too true in general...

However...  😃 !!! I FIXED IT !!! 😃

Disassembly was easy with guide referenced above on github. Actually impressed by how repairable this phone is.

Removed battery (cell), did the "prick a hole in the cell to release the gas" trick (don't try this at home - see below). Put it back together and my phone is working just fine after 4 hours' careful work. Will be ordering a new battery from Indiegogo.

Obviously, the battery is degraded but phone is working 100%. Yay!

* * *

The full story...

Root cause analysis: I now think it may have degraded gradually over time, but the sudden and severe outgassing event was triggered 1-2 days ago, when the phone was plugged into my car charger and fast charging from ~ 50%, while in direct midsummer sun. The worst possible scenario for a Li-Ion cell!

Li-Ion cell repair:

...for those who don't know, and for reference, especially for anyone else who has a battery problem on a phone where GETTING A NEW BATTERY TAKES 2+ MONTHS (are you listening F(x)tec) ...

Typical Li-Ion cells (of the type found in phones) are hardier than you might imagine. Even after some damage, the passivation/electrolyte interface is chemically self-healing to an extent. The problem is the ballooning meaning the cell no longer fits in the phone. In pouch cells, the outer soft metal shell is primarily a containment mechanism and does not play an active role in the functioning of the cell. So, it is quite safe to pin-prick the outer shell, squeeze out the gas, and reseal it. Obviously, there has been electrolyte loss (that's in part why it swelled - release of gaseous phase lithium metal compound) so for this, if no other reason, the cell will not perform as well as it used to...but in my case the out-gassing was relatively minor and the cell is holding up well enough for now.

Public Safety Announcement: we are talking about a small cell here, so it is not particularly dangerous, but: gaseous phase electrolyte (lithium/solvent) is toxic, and hydrogen may also be released which is a fire hazard. If attempting this, practice hazardous chemical handling safety rules: always use a fume hood or well ventilated area, safety glasses and gloves and protective clothing. Reseal the cell afterward with metal tape to (a) prevent further electrolyte evaporation and (b) preserve shell integrity to ensure proper containment in case of any future damage event.

How else could I have fixed it:

The other option was to simply find any Li-Ion cell (say, from another phone) that fits within the available dimensions, and transplant the charge control electronics/flex cable from the stock cell to the donor cell. Just using cells I have lying around, I could have put together a 1000 mAh pack in this way (1/3 original capacity), and that was my plan B. If I researched cell dimensions in other phones, I could probably have bought a cheap donor phone with a suitable battery, today, and swapped it in. This remains an option if the current cell further degrades before I get a new one.

Other observations:

What's interesting is that in the closed configuration, the display has more play in it around the hinge axis than before. Which makes me think the phone had already been expanding for some time, just not to a noticeable degree - but enough to warp/bend something slightly. Next time I open it up I'll see if I can adjust anything to put it back in trim. (To be fair to F(x)tec, they actually did a fantastic job on this phone...the chassis and opening mechanism is well designed and robust, so maybe nothing warped at all and the slight play is normal.)

Another small point of interest: *all* the screws in my phone were not done up over-tightly, which is good. HOWEVER...upon removing the back cover...TWO internal screws fell out onto the table. They had been literally loose inside the phone for I don't know how long. One black screw on the central metal support member, and 2nd on the right-hand PCA. As a result, the central metal support member was bent up where the screw was missing. I'm lucky that neither screw ended up causing an electrical short and permanent damage. I don't fault F(x)tec for this as the torsion on the screws was mainly about right.

A curiosity: in the course of working on it I booted the phone with the back cover removed (i.e. no antennas at all) and I was still seeing excellent UMTS, wifi and BT performance - mind you, I am next door to the tower, and the wifi AP is a few meters, so...

I will definitely be only ever plugging the phone into a slow-charging USB outlet and install a PMIC control tool so I can actually disable or control charge current at a more granular level.

Finally, I need a better set-up to protect the phone when used in the car! I have some ideas...

Once again, impressed with F(x)tec's design in all respects. It is one of the most repairable phones I've seen in a while, and very robustly designed & constructed. Internally, test points are labelled for things like vBAT, serial console, button-press simulation, and numerous others. It is actually feasible to do board-level diagnostics/repair of this phone. The separate USB-C PCA is genius - every phone should be designed this way. It is a trivial swap-out if the (notoriously easily damaged) USB-C connector fails - that is, if F(x)tec would support their impressive phone with available spare parts - HELLO!!

Am simply glad to have a working phone without waiting for several months and/or buying one for $1000+ off eBay - though I remain sorely tempted to buy it anyway just so I have a spare phone in case of any eventuality! Probably should instead wait for the Pro1 X though, for the newer chipset and corresponding newer kernel/Android 12 & longer support. From the latest updates on Indiegogo, it appears the Pro1-X is well on its way, and I'm confident they will deliver. At least, everything they say makes sense - and having fully disassembled my phone and seen the build quality for myself, I can tell you that these guys are not messing around and this is a better designed phone than just about any other Android phone on the market.

I hope this post helps anyone else who finds themselves in a similar predicament - please feel free to ask questions.

Phew!

-Mario

Edited December 26, 2021 by mbecroft

[mbecroft 82]

Oh... and the ugly: I lost the rubber microphone damper (for the main mic) somewhere on the floor while removing the back cover. Phone calls are still loud and clear, but now markedly more sensitive to slight hand movements or brushing of the phone against the ear etc. while talking 😞 Moral of the story: microphone dampers are important - don't lose them!!

[VaZso 1,998]

10 hours ago, mbecroft said:

and if it is indeed the battery, nurse it along with no fast-charging (I am actually annoyed that it fast-charges, I don't need this, and it wears out the battery).

I have picked a USB charger which has a 5V 1A-capable port along with a QC 12V compatible second port.
I use the former to charge my Pro1 overnight which is well enough for me for daily use and it does not cause any excessive heat of the phone at all.

10 hours ago, mbecroft said:

I see the spare battery on Indiegogo, do you think the Pro1X spare battery is definitely the same as for the Pro1?

Yes, I think.
They wanted to manufacture the very same phone which was not possible, but they still wanted to keep as much of the Pro1 as possible, so there is high chance for the battery remained the same.

I am glad you could solve the most urgent problem in your phone.
I am agree there are two possibilities of long-term solution:
- Buying a spare battery directly from F(x)tec which may be a slow thing
- Find a similar-sized battery which has similar parameters but made for another phone and install that one

I hope you will have a replacement battery soon.

[mbecroft 82]

57 minutes ago, VaZso said:

I have picked a USB charger which has a 5V 1A-capable port along with a QC 12V compatible second port.
I use the former to charge my Pro1 overnight which is well enough for me for daily use and it does not cause any excessive heat of the phone at all.

[...]

Yep very good idea simply using a 1A-only port. I am exactly with you here, I don't need fast charge, the phone already lasts all day and 1A or even 0.5A is enough to recharge overnight. I was just lazy and did not do this, and now paid the price. I am a little happy though because I learned a lot, and I am now more certain of the quality and my ability to maintain/repair the phone. The Pro-1X will only be as good or better.

Did more digging, and agree the Pro-1X is as compatible as it could possibly be including the battery. So I'll be ordering one on Indiegogo - along with the extra screen protectors.

Current battery has only minor damage, it is fine. It has performed perfectly today including a one-hour phone call, a long drive with Waze navigation (GPS, CPU, GPU, display all drawing power) and some standby time. Voltage and temperature are nominal under load.

* * *

It really does look like they will hit the Feb production run, from what I am seeing. There are many updates from F(x)tec on Indiegogo (over many months). They have already been through 3 board spins and de-risked nearly everything, as well as delivered prototype working phones. Plus it is very similar to Pro1, which is already proven. I doubt there will be further delays, provided they build enough to meet demand.

* * *

Also Interesting is that I found date code and serial no. on the battery. Mine is manufactured 2019-06-16 and is S/N (I forget) approx 350. It is part of the same batch as the one in the teardown photo below. Same date and the one pictured is S/N 1798 - so looks like they did a run of perhaps 2000 pieces - I have one of the earliest :) I don't know how many runs were done or total examples built - would be interesting if anyone else has a different batch. It is frustrating to wait for Pro-1X, but I see now that they are methodical, experienced engineers, and are doing it the right way.

Picture from teardown:

[mbecroft 82]

Minor update: run-down tested repaired cell - using CPU/GPU-intensive benchmarks for several hours. Got down to 3% and 3.45V while pulling ~1.3A on OpenCL benchmark - terminated test at this point. Definitely was close to rated 3100mAh.

I tried the trick of using a charger with 1A and 2.1A outputs, but even when connected to the 1A output, it charges at 1.5A (so obviously my charger does not signal the 1A charging rate, or current-limit the supposed 1A output). In fact, I tried 2 different chargers and on all of them it charges at 1.5A. (More specifically the "Max USB current" parameter is reported as 1500 mA in all cases - see screenshot below). I tested this for a short time and there is some heating (so far < 5 degrees above ambient).

Is now charging at 500 mA from laptop USB port, with no heating at all.

(Measured charging rate a bit under 1.5A because some power is going to screen, CPU etc. with phone on for screenshot.)

 

Edited December 28, 2021 by mbecroft

[EskeRahn 5,460]

26 minutes ago, mbecroft said:

I tried the trick of using a charger with 1A and 2.1A outputs, but even when connected to the 1A output, it charges at 1.5A

Yes, never trust the rated max output on a charger! Especially if it is an 'unknown' brand. I always use a meter to test, and have seen both way below and quite a bit above the rating.
As many in here will know, I also recommend charging at no more than 5W for daily charges. I only use fast charge when I really need it, and not for full charges.

[mbecroft 82]

I just need to find the right way to poke the PMIC to limit charge rate (and charge only up to 80 or 90% for longer life).

[mbecroft 82]

Question: just out of curiosity, what is the highest charging current supported by the Pro1, either at 5V or with high-voltage fast charger? Could not see it in the specs.

Edited December 28, 2021 by mbecroft

[EskeRahn 5,460]

I'm not sure, but think the max is 15W (QC at higher voltage)

[Rob. S. 1,662]

8 minutes ago, mbecroft said:

I just need to find the right way to poke the PMIC to limit charge rate (and charge only up to 80 or 90% for longer life).

For limiting the maximum charge level, I use the Magisk module 'Advanced Charging Controller' (ACC) which also has a corresponding app (AccA). In normal use, I let it stop charging at 75% and start again when the level is down to 70%. 

ACC also has options to limit charging current and voltage, but these don't work with all phones and it seems they specifically don't work with ours. Neither do we have a 'battery idle mode' which could be prioritsed by ACC (completely bypass the battery when charger is connected and battery does not need charging).

[mbecroft 82]

Hmm, interesting, thanks @Rob. S.. Looks like I may need to poke around a bit more...

I have dealt a bit with Qualcom PMIC before, but this one may be completely different, who knows!

Will report if I find a solution 🙂

Edited December 28, 2021 by mbecroft

[mbecroft 82]

Battery idle mode would be awesome. E.g. in the car, which is already a bad high-temp environment, this would mean you could run the phone with high-current power supply to run navigation apps etc., without touching the battery. I don't think I've ever seen that in a phone. Does ACC have that feature on some devices?

Edited December 29, 2021 by mbecroft

[mbecroft 82]

Thanks for advice @Rob. S., ACC seems to work really well for what it does. When I tried "test battery idle mode" it just hung saying "this will take a few seconds." After waiting a minute or two I gave up. Is this the same experience you have?

Just tested charge rate and voltage limits and both work perfectly. This is with latest stable LineageOS 11.

So I don't even need to dig into PMIC settings. ACC is doing everything.

Will see how it goes - at one point the ACC daemon died. But as of right now it is working! Which ROM were you testing on when this did not work?

Edited December 29, 2021 by mbecroft

[Rob. S. 1,662]

1 minute ago, mbecroft said:

When I tried "test battery idle mode" it just hung saying "this will take a few seconds." After waiting a minute or two I gave up. Is this the same experience you have?

Yes, it is.

Where ACC also does not help is the mysterious issue of phones suddenly stopping to charge (with the status bar still showing the arrow) which some of us have been experiencing for some months now every once in a while. AccA then reports the state being "not charging" instead of either "charging" or "discharging", and the only remedy so far is to reboot the phone. (There was already a lot of discussion about this here, and If I remember correctly, the same or at least a very similar issue has been noticed on other phones, too, with fixes being rolled out for their specific builds of LineageOS, but as of now we still have the issue in current LineageOS and probably also AICP.)

[mbecroft 82]

See my edited post - ACC is working really well including charge current and voltage limits.

One time only I encountered the "not charging" state, and this corresponded with when my battery nearly self-destructed. Not sure if this was coincidence or if it sensed too high battery temp or what. The charge control on the Pro1 is more advanced than on previous Qualcomm MSM chipsets I've worked with.

For one thing, it has the "Max USB current" setting (see in my screenshot further up in thread) and this gets set based on a number of parameters. It is this parameter which gets set by ACC to limit charging current.

Secondly, if the charger keeps pouring in power and the battery voltage is not increasing, it starts automatically derating the "max capacity" value.

In other words, the MSM8998 has some really advanced charge control logic compared with earlier chipsets. I can easily imagine that it may have something like if it keeps pouring energy in without expected voltage increase, possibly also monitoring temperature, then at a certain point it decides the battery is failing to charge and, (as a protection mechanism) leading to the "not charging" state. This is only speculation but is quite possible, and it may be your battery is starting to get worn out and is sometimes triggering such a mechanism.

Just food for thought. For now it looks like everything is working really well for me,. but I will update the thread if I dig deeper.

[EskeRahn 5,460]

24 minutes ago, mbecroft said:

One time only I encountered the "not charging" state, and ...

I have ONLY seen/heard of this phenomena on AICP and LineageOS, never on the Android stock. So I'm pretty sure that it is the common kernel of the two that got the bug. Lineage (and thus AICP) has added (or uses?) a smart feature for protecting the battery from 'overcharge' (could well be an underlying chipset feature they use?)

I'm almost certain that it is a bug in this, that puts the phone in this mode too early under some unknown conditions. The conditions are cleared by a boot, for some time. See my comments on the issue in the AICP thread.

[mbecroft 82]

Update FWIW, after running ACC for a few days. This is the best thing I've ever done to my phone (especially considering my current battery is not in top shape so I need to take care of it).

The screenshot says it all. ACC does what it says on the tin and is perfect for keeping the battery charged between whatever levels you want, and at whatever current and voltage you want.

Currently I've set it to start charging below 75% and stop above 85%, and at 650mA and 3.3V max (instead of 3.4V). Both of these dramatically reduce stress on the battery. As you can see, it is over 3 days since the battery was fully charged. The result is the ripples you see as the phone periodically gradually discharges while asleep to 75%, then recharges up to 85% (at the set voltage and current limits). The two sudden dips are when I used the phone for a serious lengths of time - appears a sharp drop and a sharp recharge in the picture, but remember this is a 3+ day chart. The recharges are all capped at 650mA. No need to mess about with using a special 1A-limited charger etc. I am plugged into a 2.1A fast charger, but charging rate is still as set.

I suggest everyone who notes setting the charging rate and voltage with ACC is not working, try again on latest LineageOS. It is working perfectly for me on 20th December LoS.

Caveats:

- What is being limited is not in fact the cell charge current, but rather the USB input current. So with my limit set to 650mA, if the phone is working hard and pulling 1A, the battery will be discharging. If phone is idle and asleep, it will be charging at near to the 650mA set point. So it is a balancing act to choose a good current limit that does not stress the battery, but at the same does not leave the phone discharging when working hard. This is precisely what "battery idle mode" is intended for, but so far does not work on my phone (in line with other reports). It may be possible to get it working, or it may be the MSM8998 PMIC as implemented in the Pro1 simply does not have this capability - have not yet investigated further to find out.

- ACC is the tool that does the low-level interaction with sysfs and whatever other APIs it uses to monitor and configure the PMIC. AccA is an app that helps make it easy to configure ACC, but is not essential. You can also configure ACC from the command line, for example. ACC is not doing magic, just using existing sysfs or other APIs for controlling the PMIC. If "battery idle mode" is simply not possible, an alternative would be a closed-loop control system that monitors battery charge current and adjusts the USB input current limit to constantly maintain battery charge/discharge current at or close to zero even during varying loads. This is speculative, but straight-forward in principle. AccA already reports charge current, voltage etc. at pretty high frequency (> 1 Hz) so a tight control loop is possible in theory. This would be ideal when, say, in the car where the phone is hot, so you do not want to be charging the battery at all, even at low current - but also don't want to be discharging it.

So - even if your battery has not started to explode like mine nearly did, disabling fast charging, and reducing the charge voltage from 4.4V down to something a little less stressful on the battery will go a long way to extending the useful life of your battery, and removes the need to have a special current limited charger just to avoid unwanted fast-charging. If I had done this some time ago, my battery would probably not have nearly died.

 

[VaZso 1,998]

@mbecroft, I have installed ACC when you wrote about it but I had to temporary disable it.

Should it work with default state or should I start configuring it after enabling?

My USB connector is worn and I am waiting F(x)tec's reply about a replacement board for a while so my charging does not work well in every positions...

After installing ACC, I could not really get it to charging but it has started charging after reboot, then it stopped... the state of my USB connector would also make me similar jokes.

Also, I have tried to run ACC's test which told me to connect charger, then failed to set charging current and other tests... does it pass for you or it works despite of the results?

My problem may has been simply caused by my USB connector which has its contacts worn out (too short) at one side (where it does not charges) and starting to worn at the other side (which is starting to fail) - so currently it is somewhat working only in one direction.

[mbecroft 82]

44 minutes ago, VaZso said:

@mbecroft, I have installed ACC when you wrote about it but I had to temporary disable it.

...

Which ROM are you running? I can confirm it works perfectly on LineageOS 11 December 20th release. (everything except "idle mode").

When I installed ACC (which I did via Magisk, but in theory this is not required), it defaulted to something like max charge up to 75%. Once I had the AccA app I could configure min and max charge levels, charge rate, charge voltage to anything - tried several settings and it always operated exactly as set. I did not even have to reboot after installing ACC - after all it does not modify system files, it merely talks to the PMIC via sysfs and/or other APIs. For me it "just works."

So yes, it should "just work" in its default state, but you will almost certainly want to configure it according to your needs and preferences. It even has a one-button press to bypass charge limits and fast-charge the phone one-off in the event you really do need max battery. Honestly, although I could written shell scripts and Tasker interactions to do all of this, ACC + AccA 90R% "just works" for my needs.

Again I can confirm 100% it works and with zero weirdness or glitches on LOS dated 20 December 2021. So...?

[Rob. S. 1,662]

1 hour ago, mbecroft said:

I suggest everyone who notes setting the charging rate and voltage with ACC is not working, try again on latest LineageOS. It is working perfectly for me on 20th December LoS.

I confirm that it works with my phone, too! (It's been a while that I tried and found it to not work, there were a few updates for both LineageOS and ACC in the meantime.) 

It's does not follow my voltage and current limits exactly, but it generally does what it says it does.

Edited January 2, 2022 by Rob. S.