My hard drives have been flaky. The other day, I got the "hard disk not found" error on boot. When I turned off the truck, the empeg didn't power down. I had to pull it from the sled to get it powered down.

Is this something fixable via hijack?

It's probably fixable by reseating the ribbon cables...

Ditto...

Out of all of my drives (lots of drives, lots of Rios), I have had one that got really trashed in a Rio.

Neither of those things are software problems, so neither of them can be fixed with Hijack. They are both hardware problems.

The two things might not be related. Then again they might, if the hard disk ribbon cable is frayed and scraping against a component inside the player.

The usual reasons for it not turning off when you turn off the ignition are listed in the usual place here, and the hard disk stuff is in the usual hard disk place. As always, each of those lists several possible causes for each problem, go through all of them carefully.

My money is on: Loose or frayed +12v continuous wire for the latter problem. Or, if the problem is happening continuously since the other day, blown fuse on that same wire.

Actually, for the not-turn-off thing, my money is on the AC sensor switch actually.

Actually, I've seen the same problem on my Rio when the drive started going AWOL. The thing would either randomly not start up correctly, shut down or best of all would start up of its own free will. It drained the battery a couple of times.

Once I'd resoldered the IDE connector the drives behaved correctly and I never saw the odd startup/shutdown problem and consequently the battery never drained again.

Adrian

The IDE connector soldering is what trashed my drive as well. Lost some of the ground pins.

Ross

Hmm... talk about an ambiguous question on my part.

I know the disk issue needs to be fixed by checking the cables, etc, and hijack can't fix that. The "can hijack fix this" question was in regards to the "not-turn-off" thing.

So, in effort to restate the issue...

When the Empeg is in a state where it's displaying the "hard disk not found" error, it doesn't correctly sense the loss of power on the ignition line, and drop into standby mode. When it boots normally (i.e. finds the hard disk and starts playing music), it operates as expected, dropping into blinking-blue-dot mode when the ignition is turned off. That's why I don't think this a hardware/wiring issue (beyond the obvious "fix the disk problem").

My thought is that it's the player application that's reading the sensor switch, and issuing the call to drop into standby mode when the ignition is turned off. When the hard disk isn't found, the player application doesn't start (is it on disk? or in flash?), and consequently, there's nothing monitoring the ignition line, to turn off the power to the empeg. Is this correct? And, is it the kernel that's displaying the "hard disk not found" error? Or is it some bootstrap code on the flash? If it's the kernel, then should the kernel be responsible for monitoring the power sensors, and dropping the empeg into standby when it's in an error state?

Ah, I see what you're saying. That's an interesting description of the issue.

I'm not sure it's true, though. I thought standby mode code was in a PIC that was independent of the player app. Mark Lord and the Empeg guys would probably know for sure.

But if standby mode code was in a PIC independent of the player app, then you'd think that it would operate regardless of the player state. But it clearly doesn't. Or, if it does, it needs the player app to be initialized, before it can do the job correctly.

For a quick demonstration that doesn't involve having bad disks, put the player in the car, and turn on the ignition. The player boots, and starts initializing. Before it finishes, turn off the ignition. If memory serves, the empeg does not drop to standby -- it finishes initializing the player app, and starts playing music.

Hi,

Yeah, that's what happens when it doesn't find the drive(s). It just shows the no hard disk drives found message, indicates X's in the eyes of the Linux penguin, and stays on. It doesn't power down on its own.

You have to remove the player from the sled while it is live. When you do that, you will often hear the player internal drives spin down.

That's what mine have done even when a drive cable has become loose.

I have never thought this was abnormal.

Ross

What power mode does the player think it's in when in the car? If the sensor is busted, it should show "AC/Home" in Hijack. In which case you can force "DC/Car".

It shows car and the DC Battery symbol sometimes during the transition from ignition to accessory with the key-switch.

Ross

Yes. Definitely car power.

Another data point... I tried to use the top button to power it down, as I would when I want to turn it off with the player running, and that didn't work, either.

IMHO, it's abnormal. If it starts up with the ignition sense, I expect it to shut down (or drop into standby) on loss of ignition sense, regardless of error state. It's just symmetry. If it can't do standby, because that requires the player app, then I'd expect it to do the next-closest thing, which is power down completely.

What called my attention to it, was that I'd made a short trip, and didn't realize that it had powered into the error state. When I came out the next morning, it was still powered on. Thankfully, it didn't kill my battery, but still not happy.

I've seen this no-power-off behaviour when it doesn't find the hdd as well. Reseating the ribbon cables always fixed both conditions.

Okay, the relevant FAQ entry has been updated to indicate that the no-standby-mode behavior can be caused by hard disk trouble.

Mmm.. I'd think it more likely that just plugging/replugging the AC power cord might tweak the internal "sensor" (switch) into working from time to time.

Mine never sees ac power. It's been years since I uploaded anything to it.

Yeah, the power-off button doesn't work when it doesn't recognize the drive in "car Power" mode.

Ross

It should still work.

Maybe I should restate that... my player hasn't been on ac power in years, yet had the hdd cable problem a couple of times. During those times, it would also not power off until I reseated the ribbon cable. I never touched ac power during that time, so the problem had nothing to do with the ac sense.

Right. The lack of response to power signals (other than complete loss of power), or button presses, is not a hardware issue. My bet is that the HD init happens before the kernel starts monitoring the various sensors (power and buttons), the error state prevents it from reaching that monitoring portion of the kernel, and that it's fixable in Hijack.

But I'm not exactly familiar with the internals...

I've forgotten how that works, but at least some of the hardware is tied directly to the sensor. I'm not sure that Hijack can bypass it all.

But my memory is fuzzy, and I'm too lazy to look any closer at it for now. Replacing the DC power jack ("AC") is a reasonably easy repair usually.

Cheers