T3 Failing to record - going to standby

[Paul_oz53]

The T3 is misbehaving. It keeps going to standby at random intervals after months of steady operation. The messages log reports each time that it is responding to a key press 116. However, there is no key press and there is no obvious phantom source. Has anyone seen something similar?

After upgrading to 19.3 a more serious problem has emerged of it failing to record, reporting a write error insufficient disk space despite having 676 GB free. It also did this once before on 17.5 so it's not the firmware. Rebooting hasn't had an effect on this behaviour.

I have tried to run the hdd check from the GUI but it fails reporting the resource unmount failed because the resource is busy. It is showing the correct path though.

It has a Warkus refurbished PSU. Obviously, I suspect a hardware or PSU fault but I would like to force a HDD check before I open it up or ship it back to Perth.

Is there a way to run ithe HDD check via either the GUI by unmounting the drive or via telnet?

Thanks

[MrQuade]

If it's registering a keypress then it is almost certainly the front panel capacitive sensors acting up.

New cables or devices installed nearby?

Any internal cables straying too close to the front panel circuit board?

[Paul_oz53]

If it's registering a keypress then it is almost certainly the front panel capacitive sensors acting up.

New cables or devices installed nearby?

Any internal cables straying too close to the front panel circuit board?

Nothing external but I take your point it's likely the FP sensor is at fault. Could be internal.

Too much happening ATM to pull it out and open it up but hopefully later in the week.

[prl]

Is there a way to run ithe HDD check via either the GUI by unmounting the drive or via telnet?

MENU>Setup>Storage>File system check.

To allow the HDD to be dismounted for the check, you have to do it when no recordings are running. You also have to disable timeshift if it is on (disable MENU>Setup>TV>Timeshift>Start timeshift after, then change channels to stop the current timeshift recording) and you need to disable debug logging if it is enabled (disable debug logging in Open Webif, then restart GUI).

If the HDD can't be unmounted to do the check, you'll get an "Error: umount: can't unmount /media.hdd. Device or resource busy." popup.

[Paul_oz53]

Thanks prl,

I hadn't disabled logging in my earlier attempt to run the check nor did I change channels after turning off timeshift. I'll give it another go and see how it goes.

Cheers,

[MrQuade]

There is also that bug with the latest front panel firmware that sometimes causes the HDD to not be detected when waking from deep standby.

[adoxa]

(disable debug logging in Open Webif, then restart GUI).

Not really helpful to the discussion at hand, but I'd like to remind people that with 19.3 this can be done from MENU>System>Log settings.

[Paul_oz53]

Thanks prl,

I hadn't disabled logging in my earlier attempt to run the check nor did I change channels after turning off timeshift. However, with timeshift disabled, no active recordings, logging turned off and channels changed to another network still unable to run the check. Same error message. After a reboot still fails with the "unable to unmount" message.

A USB drive plugged into the front panel did work long enough to reject it as an NTFS drive so it is not file check per se at fault. I tried "simple unmount" from the menu but it returned the same error message as the picture above.

The internal HDD is still readable but failing constantly with a disc full write error. Deleting a file did not free up usable space.

Cheers,

PS I found the setting Adoxa mentioned and turned logging off there.

[Gully]

Not really helpful to the discussion at hand, but I'd like to remind people that with 19.3 this can be done from MENU>System>Log settings.

Missing one step there.

MENU>Setup>System>Log settings

[prl]

(disable debug logging in Open Webif, then restart GUI).

Not really helpful to the discussion at hand, but I'd like to remind people that with 19.3 this can be done from MENU>System>Log settings.

Useful to me: I hadn't noticed that it was in the GUI settings now

[Grumpy_Geoff]

(disable debug logging in Open Webif, then restart GUI).

Not really helpful to the discussion at hand, but I'd like to remind people that with 19.3 this can be done from MENU>System>Log settings.

Useful to me: I hadn't noticed that it was in the GUI settings now

It was in the "release" notes -

• logging can be controlled from System setup;

Also, the log cleanup is now functional.

[Paul_oz53]

Turned it off at the switch overnight. Today it refuses to mount the HDD

Oddly, I could read and write to the hdd over the network last night so it seemed the controller and hdd were ok. Trip to Perth on the cards.

[peteru]

Before you ship it anywhere, ensure you are on the latest 19.3 and ssh into the T3.

Run smartctl -x /dev/sda to see what the deal is with the HDD.

If it is just a dead HDD, there's no point in shipping it back and forth.

[Paul_oz53]

Before you ship it anywhere, ensure you are on the latest 19.3 and ssh into the T3.

Run smartctl -x /dev/sda to see what the deal is with the HDD.

If it is just a dead HDD, there's no point in shipping it back and forth.

Thanks Peteru. It is on 19.3 and I will give that command a try.

Tonight it did mount the HDD but it still comes up with the write error message if I attempt to record to it. I also tried recording to the USB drive and it also gave the write error message. But, from a PC, it still operates to read or write. It gives me some hope that its not the HDD but then that means something on the mainboard. The PSU is by Warkus so not an immediate suspect.

I will have to tidy up a spare HDD to swap over when work allows some spare time. There's an awful lot on this HDD I want to keep.

I'll also try the HDD in a docking station and see if the T4 can read it that way.

[Paul_oz53]

No joy I'm afraid.

It briefly returned to life after I replaced the HDD but it has died again. Not mounting the HDD. I tried to run the smartctl command but I couldn't get past the login, despite resetting the password. Doubt that 19.3 is a factor because this started on 17.5. Network access is also erratic between reboots.

The front panel misbehaviour has also gone off the scale when it half works with random shutdowns, restarts, channel changes, etc. every few seconds. No amount of realigning the cables or separating the front panel from the unit changes it. Screws are tight too.

Tried to contact Warkus but no answer to an email or a phone call. No good news in any direction.

[Grumpy_Geoff]

Tried to contact Warkus but no answer to an email or a phone call. No good news in any direction.

It's school holidays over here - this week and next, so he may be away.

[Paul_oz53]

Tried to contact Warkus but no answer to an email or a phone call. No good news in any direction.

It's school holidays over here - this week and next, so he may be away.

Indeed, It hadn't occurred to me that he might be human afterall!

[Grumpy_Geoff]

Tried to contact Warkus but no answer to an email or a phone call. No good news in any direction.

It's school holidays over here - this week and next, so he may be away.

Indeed, It hadn't occurred to me that he might be human afterall!

Send Mark an SMS to see if you get a reply. The last time I phoned him I left a message but never got a call back. A few days later I decided to send him an SMS. I get a call back within a few mins - apparently all he had on his phone was the missed call number - no indication of a voice mail so he knew nothing of my call.

[Paul_oz53]

Update - since replacing the HDD the T3 will record, so a corrupt HDD seems to have been at play. Of course, if I shutdown the T3 it takes two or more reboots to get it to recognize it has a hard drive, USB tuner, and network connection. Clearly, it is having a lot of trouble at startup.

After working with Warkus by phone and email since he returned this week, we have concluded the front panel is rogue. No amount of rewiring and screening has fixed the random (phantom) keypress problem. It will be off to Perth shortly for inspection and replacement.

[prl]

It is possible (with a little care) to remove the front panel completely. You may want to check with Mark whether that is a viable option, rather than sending the whole box.

[Paul_oz53]

That's a good thought Peter.

I did remove it to rework the wiring. There is a question whether it is the mainboard too but Mark hasn't seen one quite like this one before. He may not have a working unit to test the panel in though he has a salvaged front panel.

I'll see what he thinks is best.

[Paul_oz53]

An update on this saga: recently another user retired a working T3 (banhambo) with a failed PSU. He very generously sent me his front panel but this didn't fix my phantom keypress problem.

So banhambo followed up with the motherboard. Replacing that too has not cured the problem either!

I am acutely aware of the cable routing issue and retightening the sensor screws. Earlier Warkus reviewed pictures of my layout and passed it.

So, clutching at straws, I fitted ferrite chokes on the PSU lead, the HDD power wires and the front panel main connection wires. Sad to say, no change.

As the HDD was replaced when the troubles began, I don't believe it could be a factor. However, a spare 120 GB SSD swap was tried but gave no improvement.

It leaves the PSU as suspect. That was rebuilt by Warkus a while back and shows no obvious signs of failure yet it seems the only possible source of the problem. So it's off to Perth for the masters attention again.

I've told Mark that he can keep it for spares if it is beyond salvation.

[peteru]

Could be environmental or even some contamination in the plastic. You could always remove the standby button screw. That's assuming that your problem is actually caused by phantom key presses originating at the front panel. Have you tried reflashing the front panel firmware?

[Paul_oz53]

Could be environmental or even some contamination in the plastic. You could always remove the standby button screw. That's assuming that your problem is actually caused by phantom key presses originating at the front panel. Have you tried reflashing the front panel firmware?

Thanks peteru. No, I hadn't thought of either step. I'll pass it on to Warkus as it's on its way to Perth.

Warkus was did say the T3 is terribly sensitive to environmental factors like UV and RF. It used to be fine but I now wonder if it could be the NBN wifi on 5GHz that's the cause. It's the only thing that has changed in recent years.

Warkus also noted that it could be the PSU as it happens with two motherboards and two front panels. Apparently some of the smaller caps can cause problems.

A further side note was that Warkus has had a run of repairs since COVID resulting in high demand and no spare parts. That and delivery delays mean he is discouraging repairs from the east coast.

[Paul_oz53]

Update today:

Warkus replaced two capacitors on the front panel power rails and that has led to no phantom keypress problems for two days.

A couple of other unrelated fixes and it's looking good to go!

[Paul_oz53]

Received Thursday, connected up and working well. Definitely usable again

However, networking is not working smoothly. I'll discuss this problem elsewhere. Briefly, the T4 can access the T3 bùt not the U4. The T3 can access everything as normal.

[Grumpy_Geoff]

Received Thursday, connected up and working well. Definitely usable again

Did you check the front panel version? Having said that, deep standby clock drift on the old firmware won't be an issue for you since your T3 is left in standby.

[Paul_oz53]

Did you check the front panel version? Having said that, deep standby clock drift on the old firmware won't be an issue for you since your T3 is left in standby.

It's 20057. The clock is keeping perfect time and no sign of the missing HDD either. But, as it's in standby, neither issue should arise.

Fstab is proving reliable for access from the U4 too. Can't help but wonder if fstab would also work for V2 access from a U4 but I don't plan on getting a V2.

One issue I did have was with signal overloading. Had to drop the antenna feed by 9dB relative to the U4 to get a picture. I've long thought the T3 was far more sensitive than the U4 and T4. This suggests that I might be right.

[Grumpy_Geoff]

Fstab is proving reliable for access from the U4 too. Can't help but wonder if fstab would also work for V2 access from a U4 but I don't plan on getting a V2.

Mate, I suggested that early in the piece, but I guess it was overlooked in favour of toy chucking.