Related
I have seen a lot of posts about people not being able to use their USB port to flash back to stock ROM in order to claim warranty.
I never paid any particular attention until I saw @work two bricked devices with the same issues in the morning. then a fried called me about the same issue. Pretty angry about his warranty.
Please post here the following:
1. Last kernel flashed. Root or not.
2. WHEN the USB stopped working? Ex: app installed, wipe, recovery, etc.
3. Charging or not?
4. Steps tried so far, helped or not.
Please keep it as concise as possible. Also, if you read this and remember a post about these issues, please direct the OP onto this. We need to see if the SGS2 is failing on USB solder or something is messing the flashing process.
Thank you.
well i keept flashing like mad and a few flashes ago my usb port didnt work..
but i tried on another computer it worked so i think the ports are bad. :O some times
USB port seems to be very sensitive. If you bend it by a fraction of millimeter it might stop working or start to behave weird.
http://forum.xda-developers.com/showthread.php?t=1132588&page=2
But if you get stable connection to PC and no charging problems, then this is probably different issue.
I think it's a hardware problem. I'm fully stock but swapped in an insecure kernel to get root. Back with the stock kernel but rooted with superoneclick.
Fault developed about 3 weeks after last flash.
I install loads of apps so can't really say if anything in particular hurt it. I did once get a charging pause for heat while playing dungeon defenders.
Sent from my GT-I9100 using XDA App
Hey gang. Wondering if you can help me with a strange issue I've got. I can't seem to connect my phone to my computer via USB. Wondering if anyone could help determine if there's a possibility of repair or not.
==Possibly relevant info==
Samsung Vibrant
Android 2.2 Froyo
Kernel version: 2.6.32.9
Build number: FROYO.UVKB5
==End relevant info==
So, my USB is acting up. Has been for the last few months, where although it charges when I plug into a computer, it's not connecting to allow data transfer. I haven't been too terribly worried, this happened before then cured itself when I did a factory reset. The thing that usually triggered it was when I tried connecting to my linux box in media player mode. The phone would just kinda lock up, I'd have to restart it and then USB wouldn't connect. But after doing the factory reset it was fine the first time. THIS time however, since I'd just left it stock now I'm itching to flash a custom rom but even though I've wiped the whole stupid phone totally clean, it still won't connect. I'm trying to follow a method on an article on dkszone.net called "Bypass E - Signature Verification Fail Issue On Vibrant Android 2.2".
So it seems I need USB for the debug mode.
What I think my main problem is, and where I need some help, is that I had purchased a super cheapwad USB car charger off of amazon for like 35 cents. Every time I connected to it, my phone would ask me if I wanted to connect using kies, media player mode, or data storage mode. I thought it was annoying but didn't really think it would harm anything. But now that I'm thinking about it, it's probably been pumping 5v into each pin on the USB port, whereas the USB functionality uses probably 3.3v, and I'm afraid I cooked something.
The random part here though is that about 5% of the time it does (sortof) connect. It brings up the USB options when I want to connect, brings up the icon saying it's using debug mode, etc. But my machine just reports that it didn't recognize the USB device. So, I'm thinking it's on the fritz, but figured I'd check here to see if I could possibly find a ray of hope.
Scooterx3 said:
So it seems I need USB for the debug mode.
Click to expand...
Click to collapse
You can use ADB over tcpip. There's many a free widget to enable it on google play or Do it manually
Thanks for that, I had been wondering if there wasn't a way to do that over wifi. My issue is that sadly I haven't even rooted the phone. ADB over WiFi requires root, whether I do it manually or use one of those super sweet apps.
So anyways, guess what?
My brother in law has the same phone I do, and I grabbed his phone one day and plugged it into my computers and had the same problem as with my own, and I know he hadn't plugged into the same charger I had, hadn't tried the media mode on Linux like I had (which were what I thought were the big contributing factors). SO, I kept digging, found that I hadn't quite yet cleaned it properly. I'd tried my best to clean it but until I swabbed it well with some rubbing alcohol it wasn't as clean as it should've been.
Long story short, I'm backing everything up so I can start flashing roms.
Moral: just clean the dang thing really well
Hey xda, i just got this used Mytouch 4G and rooted and flashed, etc... but know i notice that my wifi does not turn on at all, it just says turning on..... but it never does, when i was on 2.3.4 it would turn on but would say cannot scan. is there any fix for this? is the wifi just dead? throw the phone away?
symbiiote said:
Hey xda, i just got this used Mytouch 4G and rooted and flashed, etc... but know i notice that my wifi does not turn on at all, it just says turning on..... but it never does, when i was on 2.3.4 it would turn on but would say cannot scan. is there any fix for this? is the wifi just dead? throw the phone away?
Click to expand...
Click to collapse
What ROM are you using. Sometimes its just a kernel problem, I've been testing different kernels with different ROMS, and some of them wont let you turn on wifi. Try a different ROM then.
Hey thanks for the replie, ive used pac rom 4.2.2 and cfx rom 4.2.2
symbiiote said:
Hey thanks for the replie, ive used pac rom 4.2.2 and cfx rom 4.2.2
Click to expand...
Click to collapse
Try a stable-ish rom. Like Dark Unicorn Resurrected --->http://forum.xda-developers.com/showthread.php?t=1716558 (Second Post). Or Synthetic.Nightmare's AOSPX -----> http://forum.xda-developers.com/showthread.php?t=1684835
If that doesn't work try going to a stable GB rom (Like CyanogenMod 7.2.0) If that doesn't work then just donate your phone to someone that could use the parts
I tried cm7 and no go, click on turn wifi on and it reboots... x(
Sent from my A500 using Tapatalk HD
Awwwww man. That sucks. If it won't work on cm7. Then try a GB sense rom. If it doesn't work there. Revert to stock all over again. If it won't work there. Then your screwed man. Hope it works though.
Sent from my myTouch 4g using xda app-developers app
N_otori0us_ said:
What ROM are you using. Sometimes its just a kernel problem, I've been testing different kernels with different ROMS, and some of them wont let you turn on wifi. Try a different ROM then.
Click to expand...
Click to collapse
symbiiote said:
Hey xda, i just got this used Mytouch 4G and rooted and flashed, etc... but know i notice that my wifi does not turn on at all, it just says turning on..... but it never does, when i was on 2.3.4 it would turn on but would say cannot scan. is there any fix for this? is the wifi just dead? throw the phone away?
Click to expand...
Click to collapse
N_otori0us_ said:
Awwwww man. That sucks. If it won't work on cm7. Then try a GB sense rom. If it doesn't work there. Revert to stock all over again. If it won't work there. Then your screwed man. Hope it works though.
Sent from my myTouch 4g using xda app-developers app
Click to expand...
Click to collapse
It's not an issue with the ROM, it's an issue stemming from the kernel modules. You'll see this most commonly when a kernel and the subsequent modules weren't compiled together or the modules were compiled against a different kernel version or configuration. If it's always done this out of the box, however, you may have a problem on your hands but that doesn't necessarily mean it isn't still software related; it just makes that likelihood far less.. likely. You said it was used, so it's possible the previous owner screwed up the kernel modules and the kernel. It's not a hard issue to test for, you just need a known working kernel with known working modules. If you'd like, I'll build you a kernel that's mostly stock just so you can eliminate a majority of the suspects but I am only prepared to build kernels for Jellybean based ROMs, which isn't a big deal because most of the ROMs are already 4.2.2. LiquidSmooth is also a good ROM to try, it worked very well off the jump for me. In the event that you find that you still have an issue, certainly don't throw it in the trash. I happen to need a MT4G because my camera button, volume rocker, and the connector on the side with the three contacts that I assume is some sort of dock connector err.. fell off.. I'd advise against disassembling it unless you pay very close attention to the instructions because those three items are attached to the outer casing and if you aren't careful when you remove it then I'll have to make space in the boat because you'll be in the same one I am, haha. If you have any other questions, by all means ask and if I'm incorrect in any of my statements then someone please correct me so that I don't go around giving out bad information out of ignorance, but I'm fairly confident in my statements. You'll find the modules in /system/lib/modules if you care to look but just because it's empty doesn't mean that they aren't simply compiled directly into the kernel as there isn't much reason not to unless it's something that not everyone would use or someone might see fit to change. The firmware for the WLAN, however, will be in /system/firmware/ no matter what and will be something along the lines of bcmdhd.bin or bcm4329.bin
zygh0st said:
It's not an issue with the ROM, it's an issue stemming from the kernel modules. You'll see this most commonly when a kernel and the subsequent modules weren't compiled together or the modules were compiled against a different kernel version or configuration. If it's always done this out of the box, however, you may have a problem on your hands but that doesn't necessarily mean it isn't still software related; it just makes that likelihood far less.. likely. You said it was used, so it's possible the previous owner screwed up the kernel modules and the kernel. It's not a hard issue to test for, you just need a known working kernel with known working modules. If you'd like, I'll build you a kernel that's mostly stock just so you can eliminate a majority of the suspects but I am only prepared to build kernels for Jellybean based ROMs, which isn't a big deal because most of the ROMs are already 4.2.2. LiquidSmooth is also a good ROM to try, it worked very well off the jump for me. In the event that you find that you still have an issue, certainly don't throw it in the trash. I happen to need a MT4G because my camera button, volume rocker, and the connector on the side with the three contacts that I assume is some sort of dock connector err.. fell off.. I'd advise against disassembling it unless you pay very close attention to the instructions because those three items are attached to the outer casing and if you aren't careful when you remove it then I'll have to make space in the boat because you'll be in the same one I am, haha. If you have any other questions, by all means ask and if I'm incorrect in any of my statements then someone please correct me so that I don't go around giving out bad information out of ignorance, but I'm fairly confident in my statements. You'll find the modules in /system/lib/modules if you care to look but just because it's empty doesn't mean that they aren't simply compiled directly into the kernel as there isn't much reason not to unless it's something that not everyone would use or someone might see fit to change. The firmware for the WLAN, however, will be in /system/firmware/ no matter what and will be something along the lines of bcmdhd.bin or bcm4329.bin
Click to expand...
Click to collapse
Wow, thanks for the detailed reply. ill defiantly try this out, ill get back to you. [Update] ive checked and theres no firmware folder in system, and in system/lib/modules there is a bcm4329.ko file. (im on CM7)
Are you interested in a trade? I cant use phone with no Wi-Fi
Sent from my A500 using Tapatalk HD
Fixed my wifi!
symbiiote said:
Are you interested in a trade? I cant use phone with no Wi-Fi
Sent from my A500 using Tapatalk HD
Click to expand...
Click to collapse
Just FYI, I was in the same boat. I bought a used and heavily abused Mytouch 4g just to pull the motherboard and use it to fix my wife's hard bricked (no power whatsover) mytouch. It worked fine for a day or so, then started having random reboots and wifi failure (i.e. on the stock ROM it would cycle on & off repeatedly, then reboot, and on CM7 it would just not turn on, or occasionally show 'error' if I forced the driver to load.
I tried everything: rooting and installing multiple ROMs, multiple kernels, flashing to stock unrooted, flashing to stock Froyo unrooted, and it never resolved. In the dmesg I would get an error "dhsdsio_probe_attach: si_attach failed!" every time I tried to enable wifi. I even tried different radios, not really expecting much, but just to cover my bases. I took the phone apart multiple times looking for loose cables, damaged ribbons, etc, but never found anything, although occasionally wifi would return for a few hours, then disappear again. Often it came back right after flashing a new rom or factory reset, but even without installing any apps, it would eventually fail, so it wasn't an app issue.
Then... I discovered another thread where people with the Motorola Atrix discovered that a well placed piece of plastic would resolve their failing wifi by applying pressure in just the right spot - most likely due to a loose or failing solder connection. I dug around online and found a photo showing exactly where the Broadcom 4329 chip is in our phones (it is right behind the microphone on the underside of the motherboard), and discovered that by squeezing hard with one finger just to the left of the trackpad, and the other directly behind on the back of the phone, I could get wifi to enable and scan. Letting go would drop the scan. Give that a shot and see if it makes a difference. If so, read on...
I found that people with iPhones with a failed Broadcom 4329 module (same chip) had had some success in reflowing the motherboard. Not having any expertise in that area, I was surprised to find that a common solution was to use an oven to heat the motherboard up in order to soften the solder and do a poor-man's "reflow". I can't post links yet, but if you google "iphone WiFi Repair: Oven Fix" you'll find a great video describing how to to it and why it works.
So I went for it. I took it apart one last time (plenty of take apart videos online to show how this is done), pulled out just the motherboard, stripped everything off that wasn't permanently attached (including removing/unlatching the metal covers on front & back & all ribbons), got a cookie sheet and some aluminum foil, put the foil loosely around the rest of the motherboard (shiny side out) and ripped an opening just around the chip. Then raised the mother board off of the cookie sheet with little foil balls (per instructions online for laptop reflowing) and baked at 360 for 8 minutes, shutting off the heat and cracking the oven door slightly for 20 minutes to cool it slowly, then open all the way for another 20 minutes, out, reassembled, and now I have a working wifi chip.
Crazy, yes, but if you are about to toss the phone, it might be worth a shot as a last resort. Not a 100% success rate when you search online for people trying this, some people have destroyed their phones, but it worked for me!
Keep in mind, I just did this about 2 hours ago, so there's still the possibility it was a short-lived solution. I will try to check back in here in a week or two to confirm that it 'took'.
UPDATE 1: One day later, and zero reboots, 100% reliable wifi. It has never been so stable.
zygh0st said:
It's not an issue with the ROM, it's an issue stemming from the kernel modules. You'll see this most commonly when a kernel and the subsequent modules weren't compiled together or the modules were compiled against a different kernel version or configuration. If it's always done this out of the box, however, you may have a problem on your hands but that doesn't necessarily mean it isn't still software related; it just makes that likelihood far less.. likely. You said it was used, so it's possible the previous owner screwed up the kernel modules and the kernel. It's not a hard issue to test for, you just need a known working kernel with known working modules. If you'd like, I'll build you a kernel that's mostly stock just so you can eliminate a majority of the suspects but I am only prepared to build kernels for Jellybean based ROMs, which isn't a big deal because most of the ROMs are already 4.2.2. LiquidSmooth is also a good ROM to try, it worked very well off the jump for me. In the event that you find that you still have an issue, certainly don't throw it in the trash. I happen to need a MT4G because my camera button, volume rocker, and the connector on the side with the three contacts that I assume is some sort of dock connector err.. fell off.. I'd advise against disassembling it unless you pay very close attention to the instructions because those three items are attached to the outer casing and if you aren't careful when you remove it then I'll have to make space in the boat because you'll be in the same one I am, haha. If you have any other questions, by all means ask and if I'm incorrect in any of my statements then someone please correct me so that I don't go around giving out bad information out of ignorance, but I'm fairly confident in my statements. You'll find the modules in /system/lib/modules if you care to look but just because it's empty doesn't mean that they aren't simply compiled directly into the kernel as there isn't much reason not to unless it's something that not everyone would use or someone might see fit to change. The firmware for the WLAN, however, will be in /system/firmware/ no matter what and will be something along the lines of bcmdhd.bin or bcm4329.bin
Click to expand...
Click to collapse
I think WiFi should have worked on cyanogen7 or a sense Rom....
That's why I figured it was a hardware problems.
Also can you compile kernels for me? I don't know How to properly. :'(
Sent from my HTC Glacier
A little more info in case it helps anyone diagnose a different issue, below were some handy-dandy commands I used to check the driver situation:
Try to turn on wifi (obviously it will fail)
Then in terminal emulator (download one from the play store if not installed), type
$ lsmod
if nothing is returned, the driver isn't loaded. If you get "bcm4329" listed as 'live', then your driver is loaded. Its probably not a driver compatibility problem (not 100% on that, since I'm new to all of this).
If your driver isn't loading, it still might not be a bad driver. It might just have been because the chip wasn't detected. Type the following to give permissions and manually load the bcm4329 driver:
$ su
# mount -o remount,rw /dev/block/mtdblock3 /system
# insmod /system/lib/modules/bcm4329.ko
if it returns an error/fail, you've got the wrong bcm4329.ko file/driver installed or maybe a corrupted file. Try reflashing the ROM, or replacing the bcm4329.ko file with one that is compatible with the ROM/kernel you are using.
If not, then
# exit
then check again with
$ lsmod
If the driver shows up now, then at least you know it can load. Then try connecting again. If it still doesn't connect, this might help the process of elimination.
go back to terminal emulator immediately after trying to connect
type in dmesg
scroll up a little ways and look for the a series of phrases 'wifi_set_power: 1', etc. If you see it trying to connect, then a few lines down, see the error "dhdsdio_probe_attach: si_attach failed!" shortly after it loads the Dongle Host Driver, then it isn't finding the chip. You probably have the same loose connection problem I had (above), and may have to do something physically to your phone if you want wifi back. I'd try the 'squeeze test' first to confirm before taking it apart.
I hope that is helpful. I spent way too much time figuring this out, and would have loved to have found this info somewhere in one spot online.
its def a kernel problem because i rely on wifi everyday since i tether to my pc and once i tried a kernel that should fix headphone heat up. didnt know if it did but it broke the wifi. would not switch on so i switch back to a next kernel n wifi works. roms like carbon, cm10.2 everlov gives kernel with wifi working. hope this help if someone end up in this boat again
http://forum.xda-developers.com/showpost.php?p=41652708&postcount=3
Just a quick link to N_otori0us_'s posts on kernels if you wanted to try other ones.
tonyMEGAphone said:
http://forum.xda-developers.com/showpost.php?p=41652708&postcount=3
Just a quick link to N_otori0us_'s posts on kernels if you wanted to try other ones.
Click to expand...
Click to collapse
There's more kernels for Gingerbread in the second post too. The one's in the 3rd post were just common things I've seen people asking for,
my friend just gave me a new motherboard, works fine now.
followup on the oven solution
After a week - wifi still working, but not without some intervention. It crapped out again after 2 days, and I re-baked it (the first time I chickened out and didn't keep it at 385 for the full 8 minutes). It has stayed fairly consistent, but started acting up again over the weekend, so I grabbed a tiny square of some sort of soft compressable insulation material from a tablet I was fixing, and placed it right over the broadcom chip to keep it compressed and applying a little pressure. So far that has held and worked well.
I have a feeling its a matter of time before this eventually fails again, too. I don't think there is anything that can be done to permanently fix lead-free solder paste that has started to develop micro 'whiskers', according to what I've dug up online. Reflowing and pressure seem to work temporarily to re-establish a connection. If I knew how to remove the solder and re-apply a solder paste with some lead content, I'd do it. That would probably be a permanent fix. That or a new motherboard ala simbiiote.
[UPDATE 2] Killed it. Apparently re-baking every few days at slightly hotter temperatures eventually will fry your motherboard. Don't go past 400 degrees, and use an oven thermometer to get an accurate idea of what the real temp is. Now back hunting for broken ebay phones to replace the motherboard. Oh well, it was good while it lasted.
By now folks realize there are different hardware variances for the touchscreen which is one reason why some people have issues and others do not.
Don't you hate it when you have a workable system, then an update comes along which fixes something, let's say GPS, but then your touchscreen goes down the tubes? Suddenly you have dead spots, dead bands, ghost/phantom touches, stuck touches, broken drags, etc. Touchscreen works ok when held or plugged in, but is unusable on a table?
Never fear, I've put together packages to handle 3 different variances of the touchscreen.
I was able to make the touchscreen usable again on 3 of 4 test units. The 4th was always bad even out of the box so it might be hopeless for that one.
I do NOT claim this will fix everyones' touchscreen and that isn't even the goal as it might be an impossible task.
The goal is to make 70-80% of the problem units usable (again), but even if only 50% become usable this is an improvement over the current situation.
On to the packages:
In the attached boot-ts.7z file you'll find 4 files:
boot-ts10-lock.img
boot-ts20-lock.img
boot-ts30-lock.img
boot-ts-unlock.img
the ts10, ts20, and ts30 packages handle 3 variances on the touchscreen
You basically try each one to see which works best for your particular touchscreen.
Once you find one (and hopefully you do) then it is locked in place so future updates won't mess up your touchscreen again. No more OTA performance anxiety.
Now let's say there is an update in the future which has touchscreen fixes you want to use, that is where the 4th package comes into play. Once you use the ts-unlock image, you will once again be using touchscreen updates from the manufacturer. If that update doesn't work out, you can go back and use the ts10/20/30 package that was working for you.
Ok, so how does all this work?
I tried to make the process as simple as possible (has been tested on JWR66N, JSS15J, and JSS15Q, will work on LTE and custom ROMs also):
Your bootloader must be UNLOCKED to use these packages. Unlocking your bootloader is covered in this post.
unpack the boot-ts.7z package using 7-zip
you'll be left with a boot-ts directory with the aforementioned files
connect your tablet by USB
make sure USB debugging is enabled
Settings->About tablet [tap "Build number" 7 times]
Settings->Developer options [check "USB debugging"]
you can find adb/fastboot drivers HERE
in a command prompt in the boot-ts directory type (commands in italics)
adb reboot bootloader
wait for the bootloader screen to come up
alternatively use Power+VolDown to get to the bootloader/fastboot screen
fastboot boot boot-ts10-lock.img
this is how you try out package ts10
the screen will hang for 18-24 seconds, this is NORMAL, do not touch
this does NOT flash over your existing kernel
wait until Android is COMPLETELY booted
Note: if you are on Android 4.4 the stock 4.3 kernel will not get past the BouncingBalls
wait 120 seconds for the boot to settle down
press/hold the Power button 10-15 seconds until the tablet reboots
proceed to next step
adb reboot
after booting to Android you MUST reboot at least once
IGNORE any instability up to this point, it is expected
if adb reboot does not work for some reason, use the power button to shutdown/reboot
once you are booted into Android the 2nd time, try out the touchscreen and see if it feels usable
if there is instability, give it one more reboot back into Android before judging
if things are not working acceptably
repeat step #3, with ts20, and ts30 if needed
hopefully between ts10, ts20, and ts30, you will have found a package that works for you
there is nothing else for you to do
you can reboot and use your normal ROMs, custom ROMs, stock kernels, custom kernels, etc.
your touchscreen package is locked in place and won't change unless you want it to
if you want to go back to the way it was before and/or you want a future touchscreen update from the manufacturer
repeat step #3 using boot-ts-unlock.img
upon first reboot you will have the touchscreen software from the installed ROM/kernel
Please post feedback with:
the version of software you are running
what was/is broken
whether any of the packages helped
Remember to click thanks if you find this useful.
That's it folks, hope this helps somebody.
very interesting! lots of FC after loading the file, cant really "Power off using power button" because of FC-windows, using adb reboot instead.
current ROM: [ROM][CM10.2][4.3][CyanogenMod FLO Kangs][DJL10.2][LINARO][Build 10-TEST][2013/08/27]
ts-10: no more jumping in google maps while zooming/panning, but still jumping in other apps. after second reboot, starts jumping in maps again. after third reboot, no jumping on rightside of screen.
screenshot using two fingers
ts-20: lots of jumping on right side of the screen, zooming/panning in gmaps is still jumpy. after third reboot, still jumpy in gmaps.
ts-30: still jumps of right side of the screen, just a few mm from the right bezel,zooming/panning in gmaps is still jumpy. after second reboot, same effect. after third reboot, still jumpy on right side of screen.
so what exactly are in those files?
neonlite said:
very interesting! lots of FC after loading the file, cant really "Power off using power button", using adb reboot instead.
tried ts-10. no more jumping in google maps while zooming/panning, but still jumping in other apps, will try others.
so what exactly are in those files?
Click to expand...
Click to collapse
When you say lots of FCs, can you tell me what ROM/version you are running?
You should IGNORE any instability in that first boot. While you can boot into android using the files I'm providing (and in some cases run fine), there could be mismatches in versions depending on what you have currently installed, which is why I say in RED you must reboot at least once before the system will become stable.
Once you do that reboot, you are back on your original software with just the touchscreen controller updated.
neonlite said:
ts-30: still jumps of right side of the screen, just a few mm from the right bezel,zooming/panning in gmaps is still jumpy. after second reboot, seems most stable.
Click to expand...
Click to collapse
You must IGNORE any results from that first boot for all the packages, especially since you are running CM because (on that first boot) you are essentially running CM with a stock kernel without any kernel mods CM may need.
On CM I would actually not pass judgment until perhaps a second reboot.
One shouldn't have as much instability in that first boot if they are running a stock ROM.
Can you compare the behavior to what you had before trying them out?
I presume you had some issues before, hence trying these out.
sfhub said:
You must IGNORE any results from that first boot for all the packages, especially since you are running CM because (on that first boot) you are essentially running CM with a stock kernel without any kernel mods CM may need.
On CM I would actually not pass judgment until perhaps a second reboot.
One shouldn't have as much instability in that first boot if they are running a stock ROM.
Can you compare the behavior to what you had before trying them out?
I presume you had some issues before, hence trying these out.
Click to expand...
Click to collapse
yes, ignored all FC's. no problem.
ts-10 is running perfectly, no more jumpy stuff. maps/right side of screen, hold and drag working great too.
I really appreciate how much effort you've put into helping others with troubled devices here.
Could we get some clarification on what exactly the 3 packages are? Are these touchscreen settings customized by you or simply the settings from each of the 3 versions of 4.3 we have had available to us?
You sent me here from the other thread and i must thank you, the ts10 one worked fine and now no screen problems, cancelled my replacement. Just a question though, if you tap hard on the camera corner of the tablet whilst holding it in the air, do you hear a small rattle?
It is the touchscreen controller software.
pashinator said:
Just a question though, if you tap hard on the camera corner of the tablet whilst holding it in the air, do you hear a small rattle?
Click to expand...
Click to collapse
Every unit has the rattle. Some louder than others. It is normal. Possibly the lens or focusing mechanism.
sfhub said:
Every unit has the rattle. Some louder than others. It is normal. Possibly the lens or focusing mechanism.
Click to expand...
Click to collapse
cool i thought i broke it by bashing it to get the screen to work again
i've been a long time lurker at the xda forums and your multitouch fix made me register and become a member just to say thanks. ts10 fixed my multitouch issues, going to observer for a few more days and update you if there are bugs. thanks a lot!
My tablet was working perfectly on the JSS15J then JSS15Q came along and I found letters "sticking" while typing and the screen was unresponsive to touch while laying flat on a table. I flashed the ts-10-lock image and so far, so good. I have been typing for awhile now with no "sticking" letters, and while laying flat on a table. I have not encountered the unresponsive screen at all flat on its back or otherwise. I will continue running through scenarios that used to give me problems and report back if I encounter anything. Also, running Bulletproof kernel and no impact with kernel functions.
Thank you sfhub!
Can anyone verify if placing the N7 on a Macbook Pro keyboard still triggers the multitouch craziness with the patch?
This is the video of my N7 placed on my MBPs' keyboard.
http://www.youtube.com/watch?v=H05N1P0dpTo
Can this really be only a software problem?
I am not really experiencing other touchscreen issues with my N7 except for some very rare stuck touches. This one I triggered by accident.
bandit_knight said:
Can anyone verify if placing the N7 on a Macbook Pro keyboard still triggers the multitouch craziness with the patch?
Click to expand...
Click to collapse
So this guy goes to the doctor's office and says my head hurts everytime I hit the wall. The doctor says, then stop hitting the wall.
Couldn't you just slide it over and don't use the tablet on top of the keyboard?
Have you considered the MacBook Pro might be emitting interference? Stray currents and voltages? It doesn't take much to throw these capacitative touchscreens off.
On this thread, most people have it much worse than you making the tablet very hard to use. We are trying to get these units back into a state that is at least usable.
Your case, while an interesting science project really isn't something someone would feel any sense of urgency to help address.
Excellent.
I'll try these now and put them through there paces.
Sent from my Nexus 7 using XDA Premium 4 mobile app
Great work
Incredible!! tried the ts10 and OMG everything is fixed!! Great work, sfhub!!
you should work in Google, seriously! shouldnt he, everybody?
My device is:
1) 4.3 JSS15Q
2) two finger touching registered as one finger, and muiti-touches ghosted. hold on the Left side of screen and the touch is registered repeatedly
1) TS10
Yea I booted 10 first and it is great. I can't reproduce any of the bugs I've seen before using YAMMT.
Google Maps Pinch/Zoom is working flawlessly. I couldn't even zoom in before.
The OP says to "fastboot boot boot-ts10-lock.img". Should we eventually "fastboot flash boot-ts10-lock.img" to make it permanent?
My N7 is:
HW rev_e
Bootloader....... FLO-03.14
15Q - CM 10.2 Kanged Build 9
Sitting the N7 on my Logitech Keyboard doesn't make it go crazy either! LOGITECH FTW
sfhub said:
It is the touchscreen controller software.
Click to expand...
Click to collapse
How do you "lock" it in so that it won't get updated? Is it just a file in the filesystem, if so how do you prevent "fastboot erase" from getting rid of it? I've looked at ektf3k.c and know where I need to make a change so that it will replace any non-matching version, but I don't really know where the actual firmware is stored. It doesn't seem to be in the vendor part of the kernel tree or in the actual ektf3k.c or ektf3k.h files. Where is it kept? kept in fw_data.b
I'm new to Android, but I'm not new to Linux or reverse engineering so please be patient with me. Just looking for a technical description of how you did what you did. What exactly is in the .img files?
EDIT: I haven't booted any of them yet, but I'm guessing these are Linux kernel images with a modified version of the elan driver. Sorry for being stupid, I'm just trying to put this all together in my head and I can't cope with anything that seems like magic. Thanks for your time.
EDIT: Ok, I figured out where the firmware is stored. I also successfully built a ROM with my compiled kernel and flashed them to the tablet. No brick, it booted the second try and seems to work. Of course there's no apps on it hardly.
I'm kinda thinking out loud here, please correct anything that I get wrong:
I started off with the assumption that there was a closed source driver, but that seems to have been incorrect on my part. After looking at the code a little more, I see that ektf3k.c appears to be the lowest level driver, outside the embedded firmware. The driver does only update the firmware in the case that the flash is empty (FFFF.....) or the firmware presented is a higher version (just as you said of course). It appears that a simple change will make it flash any version presented that is different than the one currently flashed in. I plan to try this on my tablet and see if there are any unexpected repercussions from doing so.
I wish I had a datasheet for the controller (hint hint anyone).
DISCLAIMER: I know that you have already solved this problem in an arguably better way, I just want to tinker around a little. Thanks again. :good::good::good:
sfhub said:
By now folks realize there are different hardware variances for the touchscreen which is one reason why some people have issues and others do not.
Don't you hate it when you have a workable system, then an update comes along which fixes something, let's say GPS, but then your touchscreen goes down the tubes?
Never fear, I've put together packages to handle 3 different variances of the touchscreen.
I was able to make the touchscreen usable again on 3 of 4 test units. The 4th was always bad even out of the box so it might be hopeless for that one.
I do NOT claim this will fix everyones' touchscreen and that isn't even the goal as it might be an impossible task.
The goal is to make 70-80% of the problem units usable (again), but even if only 50% become usable this is an improvement over the current situation.
On to the packages:
In the attached boot-ts.7z file you'll find 4 files:
boot-ts10-lock.img
boot-ts20-lock.img
boot-ts30-lock.img
boot-ts-unlock.img
the ts10, ts20, and ts30 packages handle 3 variances on the touchscreen
You basically try each one to see which works best for your particular touchscreen.
Once you find one (and hopefully you do) then it is locked in place so future updates won't mess up your touchscreen again. No more OTA performance anxiety.
Now let's say there is an update in the future which has touchscreen fixes you want to use, that is where the 4th package comes into play. Once you use the ts-unlock image, you will once again be able to get touchscreen updates from the manufacturer. If that update doesn't work out, you can go back and use the ts10/20/30 package that was working for you.
Ok, so how does all this work?
I tried to make the process as simple as possible (has been tested on JWR66N, JSS15J, and JSS15Q):
Your bootloader must be UNLOCKED to use these packages. Unlocking your bootloader is covered in the stickies.
unpack the boot-ts.7z package using 7-zip
you'll be left with a boot-ts directory with the aforementioned files
connect your tablet by USB
make sure USB debugging is enabled
you can find adb/fastboot drivers HERE
in a command prompt in the boot-ts directory type (commands in italics)
adb reboot bootloader
wait for the bootloader screen to come up
alternatively use Power+VolDown to get to the bootloader/fastboot screen
fastboot boot boot-ts10-lock.img
this is how you try out package ts10
the screen will hang for 18-24 seconds, this is NORMAL, do not touch
this does NOT flash over your existing kernel
adb reboot
after booting to Android you MUST reboot at least once
IGNORE any instability up to this point, it is expected
once you are booted into Android the 2nd time, try out the touchscreen and see if it feels usable
if there is instability, give it one more reboot back into Android before judging
if things are not working acceptably
repeat step #3, with ts20, and ts30 if needed
hopefully between ts10, ts20, and ts30, you will have found a package that works for you
there is nothing else for you to do
you can reboot and use your normal ROMs, custom ROMs, stock kernels, custom kernels, etc.
your touchscreen package is locked in place and won't change unless you want it to
if you want to go back to the way it was before and/or you want a future touchscreen update from the manufacturer
repeat step #3 using boot-ts-unlock.img
Please post feedback with:
the version of software you are running
what was/is broken
whether any of the packages helped
Remember to click thanks if you find this useful.
That's it folks, hope this helps somebody.
Click to expand...
Click to collapse
So if we do this can we lock the boot loader backup. If for some reason it doesn't work will I be able to take it back for exchange. After unlocking the boot loader. Do people leave the boot loader unlocked . Is the software staying the same meaning will I be ok downloading future updates.
Is there a way to find out what variance is installed right now on the device? I got JSS15Q and my tablet is useable but sometimes there are ghost touches or the touch is stuck or gets interrupted and the likes.
I wanted to start a thread to solve some minor headaches regarding NuVision tablets. Specifically the TM800W560L, but may include others.
The manufacture is TMAX, and the brand is NuVision. Some of you have these. I picked up the 560 at the Microsoft store, a couple of days before Christmas, 2016. I ghosted the factory image as backup, and wiped clean to start from scratch, using Windows 10 Pro.
Let's say my hack started at the Microsoft Store, social engineering a sale of $25 for the unit. Yes you can get these for free, if you were buying a Windows Surface, but why would you want this if you have a Surface? It's such a terrible device to support for the sod who got one...
This unit is the most similar to the HP Stream 7, MSRP $79. I have a couple of these as well. I a bit miffed that, although the CPU is 64-bit, we can only use 32-bit Windows. This produces some problems as only 64-bit drivers are readily supported via Windows Update. I have made use of some of these drivers for the 560 to solve some problems.
Lets get down to the issues at hand:
1. The Camera drivers.
OV2680 ( OVTi2680 Ven_OVTi&Dev_2680 )
OV5648 ( INT5648 Ven_INT&Dev_5648 )
Problem: For every Windows Feature Upgrade, the drivers are wiped out and replaced with some non-functional version. Frankly, I don't think I have been able to get the cameras to work at all. At least, not by using the Windows 10 Camera App. I pulled a litany of drivers from various sources and I can get them to at least seem okay, as far as the device manager is concerned, but not much beyond that.
2. Intel Chip Set drivers.
Problem: Much like the camera drivers, for every Windows Feature Upgrade, the chipset drivers are updated as well, but then the supplied drivers are non-functional. This results in the lost of touch and gravity sensors, among other functions. This requires the use of a USB OTG adaptor, (Yes you can have USB and power at the same time!) and reinstallation of the original chipset drivers.
All in all I get the sense this is abandoned technology, and I feel that no-one particularly cares. In addition, I am a little bored with this thing, and I would like to see if I can make it better.
If anyone has questions or advice, I am open minded, but it should be sincere.
Thanks,
I picked the TM800W610L. I will see if the camera drivers and chipset drivers from this model will work for the 560. However NuVision has reduced their driver set to bare bones. In some respects this is admirable, but in my case, it leaves little hope that there's any 32bit support.
For the uninitiated, the 560, although has a 64bit CPU, the BIOS only boots 32bit Windows. This is the same situation on the same CPU, for the HP Stream 7, which I am also leveraging for support.
I have been able to dig around a little bit, regarding the BIOS firmware, for both models. Geekbench is a nice repository of data, for a shallow dive.
This is what I have, for known BIOS revisions:
I have the installer for version H, supporting the 610.
I can backup version R, installed on my 560.
If anyone is interested in an "update" from an older version, it would be interesting to get a backup of what you currently have in exchange for an update.
TMAX TM800W610L
JK-BI-8-HLK80CR100-C34A-101-F-LCD3 08/01/2016
JK-BI-8-HLK80CR100-C34A-101-H-LCD2 03/02/2017
TMAX TM800W560L
JK-8-BI-PX5S10TCR100-X534A-001-G
JK-8-BI-PX5S10TCR100-X534A-001-H
JK-8-BI-PX5S10TCR100-X534A-001-L
JK-8-BI-PX5S10TCR100-X534A-001-O
JK-8-BI-PX5S10TCR100-X534A-001-P
JK-8-BI-PX5S10TCR100-X534A-001-Q
JK-8-BI-PX5S10TCR100-X534A-001-R 08/25/2016
JK-8-BI-PX5S10TCR100-X534A-001-T
I just picked one of these up myself (the TM800W560L) and am working on getting it upgraded to Pro (and put on the Insider preview program, because I like my devices consistent) at the moment. I was a little sad to see that these machines have 32-bit UEFI effectively making the 64-bit enhancements useless and limiting software support to 32-bit only (that immediately threw out my plans for using WSL/Ubuntu on the thing since Windows Subsystem for Linux is a 64-bit only feature), but after a couple hours with this tablet, I'm really enjoying it. It took about 4 hours to install the latest batch of updates, but unlike your own experiences, all my drivers still worked after the upgrade from 1511 to 1709 (Build 16299).
I did see some very negative reviews for the product on Amazon, mostly about customer support, so I knew what I was getting into when I ordered it.
EDIT: My gripes about the device were apparently for a defective device. Got a new one after the note below caused a brick and have had a much better time with it.
Important! For my first real contribution to this thread, I wanted to give a warning, don't mess with the BIOS options if you can help it on these devices. I simply changed the "Boot to State when Power is resumed" from G0 to S5 and the tablet no longer powers on. It's charging, but nothing is happening from the buttons. So I guess it's to the exchange with this one.. Learned for the next one though :/ [Update: Turns out that it caused an actual brick, lesson learned, don't mess with the Power Management firmware settings! There's no good way back from a brick caused by UEFI firmware settings, you apparently have to reflash the BIOS using an SPI programmer to fix those. These manufacturers really should be designing their firmware better (and not using 64-bit firmware on a 64-bit processor, shame on them for that travesty!)]
Edit 2: I ran into an actual issue that I can reproduce pretty frequently. With the native Windows auto-rotation on, going from Landscape to Portrait (or vice-versa) leaves the touchscreen in this weird state where it's still looking for touch input from the previous orientation. Not sure if this is a Windows 10 version 1709 build 17043+ bug, but it appears to be, as the issue does not happen if I turn rotation lock on and manually change orientation. It's an interesting issue to say the least. [It appears after browsing the feedback hub that this is indeed a Windows 10 Insider bug, if you are affected by this, turn off auto-rotation and then manually rotate the screen using Settings, the touch input works as expected then.]
Hi, thanks for the great info! I'm trying to start a little robot makerspace and wanted to mount these little cheap tablets on mobile robots. The only issue appears to be not being able to use USB devices while supplying external power to the device. It sounded like you mentioned this is somehow possible, could you elaborate on that?
Thanks!
Jesse
ShadowEO said:
Edit 2: I ran into an actual issue that I can reproduce pretty frequently. With the native Windows auto-rotation on, going from Landscape to Portrait (or vice-versa) leaves the touchscreen in this weird state where it's still looking for touch input from the previous orientation. Not sure if this is a Windows 10 version 1709 build 17043+ bug, but it appears to be, as the issue does not happen if I turn rotation lock on and manually change orientation. It's an interesting issue to say the least. [It appears after browsing the feedback hub that this is indeed a Windows 10 Insider bug, if you are affected by this, turn off auto-rotation and then manually rotate the screen using Settings, the touch input works as expected then.]
Click to expand...
Click to collapse
I can confirm the strange rotation behavior, on the 610L. I got out of it by hooking up an OTG cable, keyboard and mouse.
Boot up and log in, rotate to landscape and reboot, not messing with the touch screen. After the reboot, the touch screen worked in landscape, and seems to be in sync again.
I installed 17083, that was a chore, and all seems well. However I am getting app crashes, which I think is an OS issue. (currency ran sfc /scannow - no probs. And dism /online /cleanup-image /startcomponentcleanup /resetbase) Next is to run chkdsk and reboot... I have no real hope this will take care of things until the next fast build, which could be in a week or two at the moment.
To get the recent build installed, I had to run a custom script, to clean up all the known temp and softwaredistribution/downloads folder. Additionally I removed the upgrade assistant and one drive. Then ran the Windows Update trouble shooter, and ran (compact /compactos:alwasy) and set NTFS compression on the C drive, ignoring all errors. I was able to get 16GB free, and finally able to install the latest Fast Insider's Build.
Some of these things are not recommended, however extreme conditions require extreme measures... 32GB MMC Crive is a bit restricted.
jjurban55 said:
Hi, thanks for the great info! I'm trying to start a little robot makerspace and wanted to mount these little cheap tablets on mobile robots. The only issue appears to be not being able to use USB devices while supplying external power to the device. It sounded like you mentioned this is somehow possible, could you elaborate on that?
Thanks!
Jesse
Click to expand...
Click to collapse
I am having mixed results here. I plainly am not getting good results with an OTG/HUB combo with three power settings. I am not getting power and USB/date to the micro USB port, as the same time. It's either/or.
610L - Nope, one or the other. I originally thought it was possible, but it's not.
560L - Nope, just confirmed, same result as the 610L - will charge in power+data mode, while completely off.
I have confirmed as well, 17063 build, on both the 610L and the 560L, has this strange screen rotation, mouse, touch screen orientation lock thing. That is, the screens rotate, but the touch alignment does not. The mouse pointer only moves in the same space as if it was prorate mode. Touch is fully prorate mode, while the screen rotates to landscape. Then got the BSOD (green screen)... Rebooting while in landscape mode, (requires a mouse), does fix the issue. Touch alignment comes back with screen rotation. (works on both models)
I see a future problem which I think we will have to address on our own. No help from TMax/NuVision I suspect.
When the microcode is finalized by Intel, we will have to come up with a solution to update the code, to protect against Meltdown and Spectre.
I have doen this before, prior to Compaq merger, on the HP VLi8 and VL400. But I don't know if it's possible to inject new microcode on the TMax firmware/bios update.
As you can see above, I have listed a number of firmware for our tablets, I am hoping we can work with some recent purchasers and see if there's an arrangement, where we could come to trade.
The USB OTG ACA support is weird. There are times it does work and times it does not. I have two USB OTG hubs that support supplying the tablet power while USB OTG is in use. With one hub, this works fine; with the other though, it will either not charge, or it will claim to be charging, but I assume not receiving enough power on the tablet side to charge the tablet enough while it's in use. It does however work, I have used it multiple times already with my 560L.
As for storage space, yes, I recommend running
Code:
compact /compactos:always
after each update. You can also get away with storing your user profile on the MicroSD card (and applications, thus rendering the internal storage an OS only drive)
As for my previous complaint about screen rotation, this disappeared after the upgrade to 17083. It was confirmed to be a problem with the Desktop Window Manager service by Microsoft and even had a fix posted to the Feedback Hub entry for the issue.
@CraveTech It may indeed be possible to modify our firmware, but the risk of doing so may outweigh the benefits. Windows already has user-mode mitigation of Spectre and Meltdown that isn't dependent on the microcode to be installed. We also have to worry about the possibility of bricking the devices without recovery (at least in my case, as I don't have access to an SPI programmer, nor do I feel comfortable cracking this thing open) seeing as a single option change in the BIOS effectively bricked the device.
anyone try to unrar the current image from nuvision for model w610L
test the archive and i get errors.
same with drivers.
tried newest unrar and also older version.
wtf??
ShadowEO said:
The USB OTG ACA support is weird. There are times it does work and times it does not. I have two USB OTG hubs that support supplying the tablet power while USB OTG is in use. With one hub, this works fine; with the other though, it will either not charge, or it will claim to be charging, but I assume not receiving enough power on the tablet side to charge the tablet enough while it's in use. It does however work, I have used it multiple times already with my 560L.
Click to expand...
Click to collapse
What is the model of the one that works for you?
I just got a TM800W610L the other day and have been looking for one that works with this tablet.
Thanks
netstat_EVO said:
What is the model of the one that works for you?
I just got a TM800W610L the other day and have been looking for one that works with this tablet.
Thanks
Click to expand...
Click to collapse
the exact hub I had the best experience on so far (I am still buying and trying hubs) was this one: https://www.amazon.com/gp/product/B00LTHBCNM/ref=oh_aui_detailpage_o02_s01?ie=UTF8&psc=1
It charged when it said it was charging and had no qualms with the other devices. There's a Cerrxian branded 4-port rectangular hub which doesn't seem to work well at all for my devices. Even when the Cerrxian says the system is charging and the system shows it, it still dies. I assume that it just isn't supplying enough power to the VBUS :/
ShadowEO said:
the exact hub I had the best experience on so far (I am still buying and trying hubs) was this one: https://www.amazon.com/gp/product/B00LTHBCNM/ref=oh_aui_detailpage_o02_s01?ie=UTF8&psc=1
It charged when it said it was charging and had no qualms with the other devices. There's a Cerrxian branded 4-port rectangular hub which doesn't seem to work well at all for my devices. Even when the Cerrxian says the system is charging and the system shows it, it still dies. I assume that it just isn't supplying enough power to the VBUS :/
Click to expand...
Click to collapse
Interesting. I had seen that one on Amazon and quickly glossed over it since it had a switch for OTG/Charge, so I assumed it couldn't do both simultaneously. At $7 I'd say it's a cheap enough experiment though. Thanks for the info.
netstat_EVO said:
Interesting. I had seen that one on Amazon and quickly glossed over it since it had a switch for OTG/Charge, so I assumed it couldn't do both simultaneously. At $7 I'd say it's a cheap enough experiment though. Thanks for the info.
Click to expand...
Click to collapse
Ya I was surprised too, but it worked with my Nexus 7 and LG G2 with their ACA enabled kernels. I had to fidget with the switch a little before charging started, but I was able to confirm in Ubuntu Touch on the Nexus 7 that it still saw the USB devices via lsusb. I seem to have lost that exact hub within the past week or so, but I had used it last month to charge the tablet while reinstalling Windows from my flash drive with a keyboard and mouse attached. I can't testify that it actually charged the device since I hadn't been able to see percentages (being in Windows setup and all), but the charging indicator was on, the screen was as bright as Windows normally makes it when charging, and it didn't die while the device was at ~15% battery before starting the installation. The Cerrxian hub couldn't do that and the tablet died while attempting installation, So I assume it was charging the tablet, or at least provided enough power to keep the device's battery stable.
I also just bought another 3-port hub from Acasis, another Chinese company making these cheap devices, but so far, I have yet to get that hub to charge the device either. So it seems that this is a very specific type of configuration inside the adapter, but the device does support OTG+Charge.
That said, Windows itself (at least the desktop versions) does not actually support this feature according to Microsoft's USB driver documentation. That it works at all seems to be a bug in the USB driver, or perhaps the adapters are doing something strange to trick the device into pulling down it's own VBUS and using the external one, possibly independently from the OS (Firmware USB drivers maybe?).
Does anyone have screenshot of the BIOS main menu and the save and exit screen? I'm trying reset to the bios. I was messing MIPI and eDP. I thought eDP was meant for external display port.
lol I fixed it. I gamble on different version of this tablet. I press F3 and Left arrow key for yes and enter. Then I press F4 and Left arrow key for yes to save and restart.
snkchaos said:
Does anyone have screenshot of the BIOS main menu and the save and exit screen? I'm trying reset to the bios. I was messing MIPI and eDP. I thought eDP was meant for external display port.
lol I fixed it. I gamble on different version of this tablet. I press F3 and Left arrow key for yes and enter. Then I press F4 and Left arrow key for yes to save and restart.
Click to expand...
Click to collapse
Glad to see that you fixed it. I would recommend against modifying the firmware settings if you can avoid it, especially any power management settings as changing one of those can completely brick the tablet leaving you unable to even enter the firmware.
EDIT 2/27: I came across my other OTG hubs and have some bad news, it appears that the OTG+Charge works on the 560L, but doesn't on the 610L. Perhaps NuVision changed USB chipset revisions between models?
Just picked up the 610L last week for $66 as a small option to use a program I need mobile. Updated to 1709 after a struggle (why Microsoft, why?).
Ran great for a week and now it’s starting to freeze and BSOD with mainly “Clock Watchdog Timeout”.
Not really getting any indication of why, I know I can run a log dump, but I often can’t get it to stay on log enough. It will even hang and crash when trying to wipe/revert.
Anyone have any suggestions?
Been a long time since I was active around here
update:
Was able to locate the stock ISO files and made a bootable usb. It’s currently restoring back to OEM status.
The installer is another language, but it’s working.
I think I won’t be install updates until I finish this project I have to do.
Another update:
Getting an intel framework and thermal error in the logs, which I notice while the device gets hot while charging and operating. If I let things be without charging it stays cooler.
I’ve only had had one BSOD and it was for System Service Exception but the logs don’t show much.
I let things update overnight with it plugged in and the screen off and it was still operating this morning.
Not sure if this is a driver issue or not. Seems that way.
Been trying to restore this but get an hour into the nuvision software and it times out. Anyone have an idea how to get this back to stock?