[trgz]

Yeah don't worry about that. Happened a couple of times to me as well. Quite scary until the device cools down and does the magical resurrection trick.

In summary, the RK3188 does not have a temperature sensor (appalling I know) so FiiO wisely put in a few temperature regulators on the board. When the board temp reaches a cutoff temperature the power to the board is just cut off, but if the charger is still connected the device will remain hot. If this happens to anyone, unplug and leave it for 6h (or pop it in the fridge) and the device will boot either to bootloader or normally to Android.

The black part of the picture is where the RK3188 CPU is, obviously it goes back to normal after a bit.

I've just discovered this post in this thread and it's the same fault that I'd noticed twice before whilst trying out two earlier kernel versions (here's the first time I encountered it https://www.head-fi.org/threads/fii...rs-and-much-more.906004/page-13#post-15125409). TBH I'd rather not push my player beyond what I would deem as 'safe' operating parameters.
Questions:
1. Is this custom kernel specific or has it been observed in the stock too? I will probably revert back to stock again if it's the custom kernel (even though I've spent a few days converting my 2 SDs to exFat and re-installing the custom kernel)
2. How many users have seen this?
3. Has it been observed during simple playback, or standby, or scanning for media? Personanly, I suspect that mine was scanning on both occasions. nb I used a smart suppy the first time it happend (advised as safe by FiiO) but a standard supply the second time.
4. Is there any app (preferably free) that can detect a dangerous rise in temp and reliably shut down the FiiO before the screen starts to cook, ie before the cutoff that doesn't actually work if powered up (ie when sat in the DK1, or even the K5)?
BTW the general advice is to not put a hot device in the fridge or freezer
Thanks in advance

 

[jkbuha]

1) This is quite a rare bug, and is actually more prevalent in the vanilla kernel than in the custom. The good news is it's harmless. FiiO's hardware has a thermal cut out that bypasses any CPU freeze conditions. I suspect it's one of the reasons why FiiO limited the original clock speed but actually it could happen at any low frequency given that it's the CPU that locks up. With the custom kernel I've removed quite a few of the conditions that cause a lockup, especially in the old kernel code.

2) There is little that the software can do to monitor this, simply because there's no temperature sensor on the CPU. In the custom kernel I've written a watchdog timer to use the battery temperature instead, so it actually makes things better but battery sensors don't work as fast as CPU temp sensors.

3) This is a very rare issue (and to reiterate, harmless), only experienced in a specific race condition. It is caused when there is a software lock up, most likely due to a corrupt filesystem, which could happen on all players, vanilla or custom. The only issue here is that if the device is on the charger the hardware does not cut off the charging circuit, but (safely) cuts off power to the rest of the board. In kernels 3.x and above I've implemented an extra watchdog that monitors temperature so it's even better than vanilla.

 

[trgz]

1) This is quite a rare bug, and is actually more prevalent in the vanilla kernel than in the custom. The good news is it's harmless. FiiO's hardware has a thermal cut out that bypasses any CPU freeze conditions. I suspect it's one of the reasons why FiiO limited the original clock speed but actually it could happen at any low frequency given that it's the CPU that locks up. With the custom kernel I've removed quite a few of the conditions that cause a lockup, especially in the old kernel code.

2) There is little that the software can do to monitor this, simply because there's no temperature sensor on the CPU. In the custom kernel I've written a watchdog timer to use the battery temperature instead, so it actually makes things better but battery sensors don't work as fast as CPU temp sensors.

3) This is a very rare issue (and to reiterate, harmless), only experienced in a specific race condition. It is caused when there is a software lock up, most likely due to a corrupt filesystem, which could happen on all players, vanilla or custom. The only issue here is that if the device is on the charger the hardware does not cut off the charging circuit, but (safely) cuts off power to the rest of the board. In kernels 3.x and above I've implemented an extra watchdog that monitors temperature so it's even better than vanilla.

Thanks for the feedback - I feel a bit more confident now

 

[Preachy1]

i installed the Neon optimized Neutron, and it's working fine. the UI appears to be a bit more responsive than the previous version. thanks again for the recommendation.

 

[Preachy1]

This kind of post is likely getting moldy, but just a quick note to say that after installing the latest version of the Kernel (and updating Neutron), everything has been working 100% for around 10 days now. So once again, I will pass on my sincere thanks to all involved with getting me to this point, with extra h/t's to JK and Witzy. You guys rock!!!!

 

[jkbuha]

Doesn't get old. Always happy to receive thanks and/or support for the kernel project.

 

[WitzyZed]

To those who opt to keep Google Play apps enabled/defrosted..

Google Play Store ver. 17.6.19 (has new looking icon)
Google Play services ver. 7.5.73 (factory version, old looking icon

Play Store needs to be updated past factory version. But I found that if Play Services were updated beyond the factory version, it seemed to bring its own flavour of micro-lag to the table. When returned to the factory version these disappear. I have 'Do Not Auto Update Apps' set in Play Store, & separately keep download manager disabled unless necessary. During idle, Play Services will only have maybe 4 processes active in this factory version vs 11 I saw in the updated one.

Has anyone else found the same?

This was the case as of custom kernel 3.4.1 (& I assume stock kernel as well).

 

[techguy777]

To those who opt to keep Google Play apps enabled/defrosted..

Google Play Store ver. 17.6.19 (has new looking icon)
Google Play services ver. 7.5.73 (factory version, old looking icon

Play Store needs to be updated past factory version. But I found that if Play Services were updated beyond the factory version, it seemed to bring its own flavour of micro-lag to the table. When returned to the factory version these disappear. I have 'Do Not Auto Update Apps' set in Play Store, & separately keep download manager disabled unless necessary. During idle, Play Services will only have maybe 4 processes active in this factory version vs 11 I saw in the updated one.

Has anyone else found the same?

This was the case as of custom kernel 3.4.1 (& I assume stock kernel as well).

Yes your absolutely right they are doing more and more stuff in the background because of permissions. You have a few options either don't use Google services or play store download apks offline from apkmirror is a good one nothing illegal all free they have the older play stores if you want to downgrade. Or my favorite option to use greenify but I do use Xposed you can try Francos servicely app for freezing background processes completely. Also different things you've got on your device use Google services and that will depend on you.

 

[WitzyZed]

Yes your absolutely right they are doing more and more stuff in the background because of permissions. You have a few options either don't use Google services or play store download apks offline from apkmirror is a good one nothing illegal all free they have the older play stores if you want to downgrade. Or my favorite option to use greenify but I do use Xposed you can try Francos servicely app for freezing background processes completely. Also different things you've got on your device use Google services and that will depend on you.

I already have them frozen with Titanium Backup. Was just putting the info out there.

 

[techguy777]

My sennheiser ie60 came today and im very impressed they sound great for 100 bucks. I wish the cable was detachable incase they break. Music is really different at first I had to change my settings on viper and all around. Even on My s7 with the Moro Sound control 2.1 kernel control and the thunder tweaks aka modified kernel auditor app you can get good sound without viper. But then with viper bam its sick im going to be back to wired headphones. My fiio just shipped supposed to be tomorrow it took them longer because the black friday sales. Hopefully theres no more delays.

 

[Preachy1]

Are Viper effects available with the custom Kernel? I paid for a few of them a while back, but at some point they disappeared, I think with one of the FiiO firmware upgrades.

 

[brick42]

yes viper went away with a recent fiio firmware upgrade

no big deal people only paid for it :eyeroll:


anyway my fellow hippie(s)
just a quick side note for you:
relisten.net

check it out it's super for dead/phish!

 

[WitzyZed]

Are Viper effects available with the custom Kernel? I paid for a few of them a while back, but at some point they disappeared, I think with one of the FiiO firmware upgrades.

so long as you install https://www.mediafire.com/download/rz2ck1ahtmnxtyo fiio music 3.4.3 they can be reactivated (fiio must have instructions somewhere)

 

[Preachy1]

Muchos Garcias!

 

[Preachy1]

yes viper went away with a recent fiio firmware upgrade

no big deal people only paid for it :eyeroll:


anyway my fellow hippie(s)
just a quick side note for you:
relisten.net

check it out it's super for dead/phish!

Cool site! If you check out a handful of recording from '84 - '91, you'll see my name on some of them. Example:

The "Pauline" in pauline.miller is me!!! My full name is Rick Pauline.