disk, ps or firmware? T3 reboot loop

Russ57 · Post by **Russ57** » Thu Oct 26, 2017 12:15

g'day
a few weeks ago, my T3 suddenly decided it had no hard disk (again). Last time this happened, it appeared it really was a hdd issue, and so i sent it back to WD for replacement. So, i sent this one back also.

The replacement turned up yesterday, so i refitted it, formatted etc, all good.

all seemed well for about 30 mins, started a recording. a few minutes later i noticed an error on screen, to the effect the disk could be full.
i then started to do a disk check, but it reported 'no disk'. i then decided a reboot may be in order. I did a shutdown from the menu. A crash report briefly displayed, then the system shut down. A cold power cycle results in a beyonwiz- date- booting message, which then disappears and repeats.

options...
1. power supply - the unit has already been back for repair, but i recall that was the main board regulator issue. COuld well be the well known PS issue, but i havent read of this reboot cycle.
2. the new hard disk has suffered an early failure. tonight, i will remove it along with the usb tuner and see if that allows booting. doesnt really clear the PS of course, as it may well be fine with lower load. I think i can plug the disk into a PC and see what that shows about the disk.
3. firmware corruption. seems a bit co-incidental as the system was functioning fine after the new disk install, for a little while.

after i have cleared the disk of blame, i'll contact beyonwiz, but i am not keen on paying again for a well known fault. the thing is now 2 1/2 years old. If it seems to be a PS issue, i'll probably replace the caps myself.
the t2 at current price seems ok, especially if i can keep my current remote as a spare. even though the disk will require an external case and PS i guess i can put up with that.
or spring for the U4? same issues as the T2 i guess. the big question is, are either of them more reliable than the T3? I really, really, really, expect consumer equipment to last well past 2 years.
thanks for any tips/hints/ thoughts...

IanSav · Post by **IanSav** » Thu Oct 26, 2017 12:22

Hi Russ57,

If your unit wasn't sent to WA for Mark (Warkus on this forum) to have a look then it is likely that other components not replaced in the last repair have now failed.

Regards,
Ian.

MrQuade · Post by **MrQuade** » Thu Oct 26, 2017 12:34

This is actually an issue with the latest front panel firmware.

The failure to detect the HDD on a cold boot only seems to affect those units with up to date panels.
I believe that there is a fix in progress for this, so there should be no need to get any hardware repaired.

In the meantime, all you need to do it power off the T3 for a short time and then boot up again. You need to be quick once the T3 has booted.....go to the POWER menu and select "Shut down".
When you start the T3 again, the harddisk should be detected. I would recommend that you leave the T3 in standby when not in use to avoid the problem happening again.
Once the front panel fix is in, then you can safely put the T3 into Deep standby when not in use again.

warkus · Post by **warkus** » Thu Oct 26, 2017 12:47

Sounds like your unit has multiple issues, which unfortunately is very common with T3 units.

The reboot loop is caused by a moisture sensitive power regulator on the main-board that is prone to failing regularly. So if your unit is just reboot looping, first you will need the regulator changed out. To do so requires a very high level of technical ability, and also requires a new regulator that has been stored and maintained correctly too - if it hasn't been and you install it into the wiz, the first bit of heat applied to it will result in instant failure of the new regulator also. Whoever invented such chip needs a good public floggin lol...

Sooooo, once you get that sorted, it is likely you also have more caps failing in your PSU if it was not repaired properly the first time and this is likely why your HDD isnt working correctly too.

So basically it needs repairing... and no its not firmware related.

Mark

Oh and lastly just to finish off, once you get past your main-board regulator issue and power supply issues, then yes see above from MrQuade, the HDD not being detected on first boot issue is also a problem, and does appear to be linked to V20057 front panel firmware, and as mentioned above can be easily worked around, but the unit needs to be functioning correctly first for that to happen!!!

warkus · Post by **warkus** » Thu Oct 26, 2017 14:34

Actually thinking about this now and going over your original post...

This comes down to what Peter was saying in another post (prl), and it has to do with the system defaulting to recording on the internal flash memory which is likely to be what has happened to you.

So this is my take on what has happened, it may actually not be hardware related - on THIS occasion anyway lol...

So you installed a new drive, which mean your T3 was off (mains) for a period of time.
You turned it on, and due to issues as mentioned by MrQuade with v20057 Front Panel firmware, on occasion the HDD is not detected on first boot from cold (mains power up). This would have required you to realise this and shut the unit down to DEEP STANDBY, count to 10, and power on again with the remote control or front panel power button. At this point the HDD would have been seen again and would have worked fine again until the unit was powered off at the wall power once more, at which point you would likely need to repeat above.
Because you didn't realise this, it is very likely that the unit defaulted to recording to the internal flash, which would have filled up in a short time, yep about 30 seconds i would say is probably about right, not really sure but it wouldn't take long...
Once full, it is likely to have corrupted the flash without any available memory left, which is why when you shut it down, you got an error message.
It is also likely why it didn't load back up again when you rebooted it, as the internal flash is corrupted and may only need a USB re-flash to fix this.

So whilst everything i said in my post above is true to these units, your issue may well be different and may be fixed by reloading the T3 via USB.

Once you do this, if it does work, and the unit reboots, note the procedure above for fixing the HDD no show issue on first power up.

Mark

Note: Deep Standby = SHUTDOWN in the power menu or LONG PRESS of the power button on the remote control. It is not normal standby, where the unit essentially stays on and just shuts off the video to the unit.

prl · Post by **prl** » Thu Oct 26, 2017 15:10

warkus wrote: ↑
Thu Oct 26, 2017 14:34
... it is very likely that the unit defaulted to recording to the internal flash, which would have filled up in a short time, yep about 30 seconds i would say is probably about right, not really sure but it wouldn't take long...

30 sec seems a bit short. There's about 450MB of free space in a T3 with no non-preinstalled plugins, and recordings are generally between ~1 and 4GB/h, or between ~17 and 67MB/min. So that's somewhere around 7 and 25 min to fill the flash with recording data (the timeshift buffer as well as actual recordings).

Not long, but not nearly as short as 30 sec.

The problem may still have been that the flash filled, but it may have happened over a series of boots without the HDD being found.

MENU>Information>Memory would have revealed whether it was the cause, but it may be too late to check now.

The whole thing is also complicated by the fact that the timeshift buffer is cleared at each channel change, reboot or GUI restart, and that the recording data files created when the HDD can't be found are hidden by the filesystem if the HDD is mounted (but they still take up the space), so they can't be cleaned up by normal filesystem operations unless the HDD is unmounted.

MrQuade · Post by **MrQuade** » Thu Oct 26, 2017 15:27

Russ57 did actually say 30 minutes, so that maths works then.

warkus · Post by **warkus** » Thu Oct 26, 2017 16:03

the 30 seconds was my bad, sorry...

prl · Post by **prl** » Thu Oct 26, 2017 16:04

MrQuade wrote: ↑
Thu Oct 26, 2017 15:27
Russ57 did actually say 30 minutes, so that maths works then.

Ah. I was just going by warkus's post. Yes, 30 min to fill the flash with the timeshift buffer or a recording is perfectly doable.

Russ57 · Post by **Russ57** » Thu Oct 26, 2017 16:19

Thanks everyone.
Some hope then..
On very first power up, with the new disk, it did hang on 'booting'. I powered off, restarted, then it did say new disk found and initiated format, then restarted.
If it didn't detect the disk then, the scenario mentioned could well have occurred.
. So, I'll
Remove the disk and try it on a pc (if there is anything recorded, then looking more like a hw issue)
Measure some voltage rails?
Download latest firmware and update

And see where we are....

Just read the latest posts., while I was writing. Yes, quite a few minutes of recording I think.

Given the thing won't boot, there seems no option to remove files in flash short of reload, which I assume (hope) completely replaces flash file system.

I'll check in when I have had a play.
Ps, I never switch off in normal use.

(a bad couple of weeks with domestic tech, my computer died, my wife's tablet died, now this.

prl · Post by **prl** » Thu Oct 26, 2017 16:23

Russ57 wrote: ↑
Thu Oct 26, 2017 16:19
...
Given the thing won't boot, there seems no option to remove files in flash short of reload, which I assume (hope) completely replaces flash file system.
...

A normal firmware update/reload replaces the contents of the firmware partition of the flash, which contains the system root filesystem. There are a couple of other partitions (e.g. the boot loader), but they don't contain file systems.

Russ57 · Post by **Russ57** » Sat Oct 28, 2017 20:50

i do believe the mystery may be solved.
firstly, when i restarted, all was well. no reboot loops, or hangs. but still no hdd. so, i attempted several times to restart as advised above, but still no hdd. so, i finally pulled the cover off, mounted the drive in my pc. healthy partition.....

i then measured the power supply rails. methinks 2.5 v off the 5v rail (at least, the red on the hdd power) is a tad out of spec.

and that big bulging C31 is a good hint as to why...

so, DIY or Beyonwiz or Warkus????

Gully · Post by **Gully** » Sat Oct 28, 2017 21:14

Russ57 wrote: ↑
Sat Oct 28, 2017 20:50
so, DIY or Beyonwiz or Warkus????

DIY is your decision as it depends on your skills, equipment, etc.

If not, then definitely Warkus/Mark.

Mikew_wa · Post by **Mikew_wa** » Sun Oct 29, 2017 14:41

No amount of coaxing seems to get the HDD to be recognised, it "appeared" yesterday after the unit was left on for a few hrs, but has not been seen since another power outage last night. (the third major outage in 3 weeks, not bad for suburban Perth.
The disk power supply is 12 & 5 and all the caps in that area look the right shape.

Now it's on the bench even if it behaves itself, I still have to take it back to the TV room !

Mikew_wa · Post by **Mikew_wa** » Mon Oct 30, 2017 11:41

Did a restore to last beta and the HDD re-appeared ! !!!

Russ57 · Post by **Russ57** » Thu Nov 02, 2017 23:13

Success!
Installed reworked ps, booted, shut down, booted shut down etc as noted above and finally I have a disk back...

I never shut it down, so hopefully it will stay back.
(whole house ups, so power outages a lot rarer than they used to be in suburban Melbourne..)

MrQuade · Post by **MrQuade** » Fri Nov 03, 2017 00:42

Fingers crossed we will see a front panel update soon to fix this little bug.

IanSav · Post by **IanSav** » Fri Nov 03, 2017 09:41

Hi MrQuade,

MrQuade wrote: ↑
Fri Nov 03, 2017 00:42
Fingers crossed we will see a front panel update soon to fix this little bug.

Where are you hearing about this?

Regards,
Ian.

MrQuade · Post by **MrQuade** » Fri Nov 03, 2017 09:46

IanSav wrote: ↑
Fri Nov 03, 2017 09:41
Where are you hearing about this?

I don't know for sure what the status of an actual fix is, but the fault itself was finally tracked down by warkus shortly after he repaired two T3's that I have given to him. One with up to date front panel firmware and one that was out of date. After he noticed the difference and once we'd done some additional testing, the cause of the fault was tracked and reported to WizHQ. I haven't really heard much else since then, but I remain hopeful.

prl · Post by **prl** » Fri Nov 03, 2017 10:04

MrQuade wrote: ↑
Fri Nov 03, 2017 09:46

IanSav wrote: ↑
Fri Nov 03, 2017 09:41
Where are you hearing about this?
I don't know for sure what the status of an actual fix is, but the fault itself was finally tracked down by warkus shortly after he repaired two T3's that I have given to him. One with up to date front panel firmware and one that was out of date. After he noticed the difference and once we'd done some additional testing, the cause of the fault was tracked and reported to WizHQ. I haven't really heard much else since then, but I remain hopeful.

Some documentation of the issue in the form of an issue tracker item may help keep it in mind.

IanSav · Post by **IanSav** » Fri Nov 03, 2017 10:29

Hi MrQuade,

I seem to remember reading, from PeterU or maybe Jai, I think, that as the T3 is EOL there is to be no more work done on hardware code for that model. While I really hope this bug *is* fixed I suspect that all efforts are now directed to the U4.

Regards,
Ian.

MrQuade · Post by **MrQuade** » Fri Nov 03, 2017 12:54

IanSav wrote: ↑
Fri Nov 03, 2017 10:29
I seem to remember reading, from PeterU or maybe Jai, I think, that as the T3 is EOL there is to be no more work done on hardware code for that model. While I really hope this bug *is* fixed I suspect that all efforts are now directed to the U4.

I'm hoping that the EOL policy applies only to enhancements, and not to critical faults like this one, but you may be right.

Paul_oz53 · Post by **Paul_oz53** » Fri Nov 03, 2017 13:11

MrQuade wrote: ↑
Fri Nov 03, 2017 12:54

IanSav wrote: ↑
Fri Nov 03, 2017 10:29
I seem to remember reading, from PeterU or maybe Jai, I think, that as the T3 is EOL there is to be no more work done on hardware code for that model. While I really hope this bug *is* fixed I suspect that all efforts are now directed to the U4.
I'm hoping that the EOL policy applies only to enhancements, and not to critical faults like this one, but you may be right.

+1 - especially since it seems the cause is known.

Mikew_wa · Post by **Mikew_wa** » Fri Nov 10, 2017 17:31

Power off again for 30 secs and NO HDD on the T3..

So can anyone tell me that the restore I did to the last beta firmware was a "fix" or just another random event that may have triggered the beast to find the HDD before I launch another USB re-boot ?

MrQuade · Post by **MrQuade** » Fri Nov 10, 2017 17:38

Mikew_wa wrote: ↑
Fri Nov 10, 2017 17:31
Power off again for 30 secs and NO HDD on the T3.. So can anyone tell me that the restore I did to the last beta firmware was a "fix" or just another random event that may have triggered the beast to find the HDD before I launch another USB re-boot ?

No, I don't think the upgrade to beta was a factor in your previous success.

If your HDD issue is front panel related then all you need to do is boot up (presumable the HDD will not be found), and then immediately shut down the T3 via the Power menu.

Don't remove power from the T3, but simply wait 10 seconds or so, then start it back up again. the HDD should be found on this second boot.

Mikew_wa · Post by **Mikew_wa** » Fri Nov 10, 2017 18:29

Thanks MrQuade

I have tried this 3 times and to no avail... I'll give a long sleep with no power tonight and maybe magically it will find the HDD tomorrow. If I could afford a ups I would use it but then to install the UPS requires a shutdown so even that is a risk ! Having just got the NBN, as opposed to the dialup speed ADSL we have been on for a decade, I have been looking that the plugins to see if there is anything else worth watching other than the local FtA and there seems to be a lot of content, but the plugins need the HDD !

If it does not wake up tomorrow I can try another HDD. I have never seen a HDD go intermittent, but there is always a first time for everything. I don't have a Linux device with an accessible HDD interface to test T3's disk.

MrQuade · Post by **MrQuade** » Fri Nov 10, 2017 18:34

Following the conversation in the other hdd thread, I assume that you managed to find the correct Shutdown option and were not using standby mode?

Mikew_wa · Post by **Mikew_wa** » Fri Nov 10, 2017 21:44

Yes, thanks..

just put it to sleep for the night, maybe there is a pesky cap holding a charge and causing something to flip the wrong flop

Mikew_wa · Post by **Mikew_wa** » Sat Nov 11, 2017 09:47

the deep sleep did not work ! I'll try a new disk this afternoon.

prl · Post by **prl** » Sat Nov 11, 2017 10:50

Mikew_wa wrote: ↑
Sat Nov 11, 2017 09:47
the deep sleep did not work ! I'll try a new disk this afternoon.

If you really want to make sure there are no capacitors holding stored charge anywhere, you need to go to shutdown/deep standby and then turn off mains power to the box (e.g. at the power switch on its back panel, or at the mains for a T2).

Myself, from the behaviour you've reported, and the known issues with the T3, I'd be suspecting the power supply.

Grumpy_Geoff · Post by **Grumpy_Geoff** » Sat Nov 11, 2017 12:19

Mike,

It looks like a trip to Thornlie. Give Mark/warkus a call.
http://decisivecomputersolutions.com.au/contact.html

Mikew_wa · Post by **Mikew_wa** » Sun Nov 12, 2017 14:58

Got back later than expected yesterday so did a HDD swap this morning.... Bingo.!!! The T3 detected the new DHH and happily reformatted it. All I can suspect is that the power section of the WD Green 2TB had experience one too many power outages and the past few times it appeared were just curtain calls before grand exit. I have never seen a HDD go intermittently incognito like this. Anyway l'll leave the 750 GB HDD in for a while as I save for a bigger disk and a UPS !

Thanks to all.

Mikew_wa · Post by **Mikew_wa** » Sun Nov 12, 2017 15:00

prl wrote: ↑
Sat Nov 11, 2017 10:50

Mikew_wa wrote: ↑
Sat Nov 11, 2017 09:47
the deep sleep did not work ! I'll try a new disk this afternoon.

If you really want to make sure there are no capacitors holding stored charge anywhere, you need to go to shutdown/deep standby and then turn off mains power to the box (e.g. at the power switch on its back panel, or at the mains for a T2).

Myself, from the behaviour you've reported, and the known issues with the T3, I'd be suspecting the power supply.

The deep sleep was just that....no power for 12 hrs

prl · Post by **prl** » Sun Nov 12, 2017 15:03

It doesn't power off everything. If it did, you couldn't start the box using the remote or the front panel.

Mikew_wa · Post by **Mikew_wa** » Tue Nov 21, 2017 14:53

It was the power supply and not the HDD. Thanks to warkus all is good the T3 is running like it should.

prl · Post by **prl** » Tue Nov 21, 2017 15:24

Good to hear it's sorted

warkus · Post by **warkus** » Tue Nov 21, 2017 19:17

Well technically it was mainboard and power supply issues causing his system problems, crashing and rebooting.

But the HDD no detect on first powerup of mains was definitely the front panel firmware v20057, i downgraded him to v20038, problem gone...

Mark

MrQuade · Post by **MrQuade** » Tue Nov 21, 2017 19:47

Woohoo trifecta!

No wonder these diagnoses are so much fun

Grumpy_Geoff · Post by **Grumpy_Geoff** » Tue Nov 21, 2017 20:42

warkus wrote: ↑
Tue Nov 21, 2017 19:17
...
But the HDD no detect on first powerup of mains was definitely the front panel firmware v20057, i downgraded him to v20038, problem gone...

... but perhaps now suffer from the clock drift issue. 20052 had it, I wonder if earlier front panel f/w had it?

peteru · Post by **peteru** » Tue Nov 21, 2017 20:51

Anything before 20057 has a whole host of subtle issues and reliability problems, but they are a lot harder to spot than a missing HDD. Downgrading to 20038 just swaps one lot of problems for another.

I've got an inkling of what may be the source of the problems and if I am right, then the HDD missing issue will become more obvious as the PSU works better. If the PSU is at the lower end of the spec or below, then the problem is less likely to occur. I've asked Warkus to run some diagnostics and let me know his findings. I'm still waiting for the results.

Russ57 · Post by **Russ57** » Sat Dec 02, 2017 17:48

So, as the op of this thread, I'd like to acknowledge warkus's efforts - his rebuilt ps sorted my original issue.

However, it only stayed sorted for a month.
Now, the T3 takes it upon itself to operate the remote control...
It frequently changes channel (down 1), occasionally adjusts the volume, and has been known to turn itself off. (standby).

I checked I had done nothing silly in reassembly. The cable to the front panel was pressing against the lower chassis lip, but I've addressed that with no improvement.

Any thoughts, anyone? Dry joints?
I even covered the ir sensor in case it was a flickering light (I saw that 25 years ago..)

Firmware issue?

Regards

MrQuade · Post by **MrQuade** » Sat Dec 02, 2017 18:50

Russ57 wrote: ↑
Sat Dec 02, 2017 17:48
Any thoughts, anyone? Dry joints?
I even covered the ir sensor in case it was a flickering light (I saw that 25 years ago..)

Firmware issue?

It's usually a result of RF interference with the front panel capacitive touch buttons.

Since your T3 has been openedv and fiddled with recently, this is a likely scenario.

Open the T3 and make sure no power or SATA cables are running near the front panel circuit board. You may also want to ensure that the screws holding the front panel on are nice and secure.

Russ57 · Post by **Russ57** » Sat Dec 02, 2017 21:03

Thanks!

I'll open again and work on getting the cables as far away as possible...

Russ57 · Post by **Russ57** » Sun Dec 03, 2017 16:11

Ok, rerouted all cables so they are as far away from the front panel.

Its certainly significantly better than before.
It did shut down to stand by overnight, and I have seen one 'block' warning pop up whilst watching recordings.

In 1 60min recording last night it had gone to standby about 6 times..

I'll see what happens..

Russ

prl · Post by **prl** » Sun Dec 03, 2017 16:23

Russ57 wrote: ↑
Sun Dec 03, 2017 16:11
... 'block' warning pop up whilst watching recordings. ...

What's that? What did it actually say?

peteru · Post by **peteru** » Sun Dec 03, 2017 17:24

If it appears to be temperature sensitive, then it's most probably the screws that hold the front panel in place.

You can try tightening those screws (each screw also doubles up as the sensor for a button), or put washers under them. Either way, they need to make good contact. Be careful not to overtighten the screws - doing so can damage the front panel.

Russ57 · Post by **Russ57** » Sun Dec 03, 2017 20:22

prl wrote: ↑
Sun Dec 03, 2017 16:23

Russ57 wrote: ↑
Sun Dec 03, 2017 16:11
... 'block' warning pop up whilst watching recordings. ...
What's that? What did it actually say?

Sorry, poor explanation.
It's the "diagonal line across a blue circle" you see when a function is not allowed in that state, eg, change channels while in playback.

Russ57 · Post by **Russ57** » Sun Dec 03, 2017 20:28

peteru wrote: ↑
Sun Dec 03, 2017 17:24
If it appears to be temperature sensitive, then it's most probably the screws that hold the front panel in place.

You can try tightening those screws (each screw also doubles up as the sensor for a button), or put washers under them. Either way, they need to make good contact. Be careful not to overtighten the screws - doing so can damage the front panel.

Ok, I'll try removing each in turn and refitting. I didn't realise that row of screws also held the panel in, I had guessed they were part of the touch buttons only.

The issue probably started when it got hot last week, but it's now quite cool, so it's not temperature sensitive in the normal sense.

MrQuade · Post by **MrQuade** » Sun Dec 03, 2017 20:50

Russ57 wrote: ↑
Sun Dec 03, 2017 20:22
Sorry, poor explanation.
It's the "diagonal line across a blue circle" you see when a function is not allowed in that state, eg, change channels while in playback.

That could be called the "invalid keypress icon".

About the Forum

Join us