Hi All,

I've just discovered a really weird problem when my Empeg is in the car.

When I first insert the Empeg into the sled, then it powers up fine and Hijack shows that it's in DC mode. I then turn off the ignition, and let the player go into sleep mode (I've set the timeout to be 3 seconds).

When I then turn the ignition back on again, the player always boots up in AC mode. If I remove the unit from the sled, and re-insert, it always boots back up in DC mode.

I've checked all the obvious causes, including my ACC and +12V lines. I've changed the DC power jack, just in case the 'home' sensor was faulty.

I've also noticed that when the player boots up into AC mode in the car, I often get no sound from the line outs. Pulling it from the sled, or simply rebooting it via Hijack fixes the problem.

Has anyone seen this? Any hints as to where I might look next?

My player is an original Mk2, and I'm running 2.01 software.

Thanks,

Andy.

So it only happens if you leave the player in the sled and then turn off the ignition, then turn the ignition back on again?

In other words, if you yank the player from the sled without turning off the ignition key, then put the player back in the sled, then it doesn't happen?

If so, I wonder if it's some new previously-unreported symptom of the negative spike bug ?

You said you already soldered in a new DC power jack, so I'm not going to try any troubleshooting down that road. Are the steps you described 100 percent reproducible, or is it intermittent? Intermittent results might indicate some kind of intermittent hardware issue (such as the DC power jack or a loose wire in the case), but you've been in there already, even going as far as fully removing the motherboard, so that's unlikely to be the issue.

How about This one? Again, doesn't really explain your symptoms, but worth looking at.

Thanks for the reply, Tony.

Your description is totally correct. It's been driving me nuts, and I'm still no nearer to finding the cause.

I knew that my DC jack was a bit flaky, so for a long time, I figured that this was the cause. When I replaced it, I figured the problem would go away, but it didn't.

The only other information I should add, is that I have removed some of the wires from the sled connector. These are the tuner connections (as I never use it), and the serial port. I did this to reduce the wire count when installing in very tight surroundings. At the same time, I also resoldered all the other connections.

I might try making up a diode lead, and see if that helps. I'm not seeing any of the other symptoms, but you never know.

Andy.

Since your player is an original MK2, and you say the steps I described are accurate and reproducible, I would have high hopes for the diode trick working. Do you have any of the other symptoms of the negative spike bug?

Remember that you can do it internal to the player if you don't want to dig behind your dash again.

Another option to consider is that your REPLACEMENT dc power jack is duff...

Also make sure you don't have any special behaviors set in Hijack, such as "if tuner=xx force AC power".

Hi Tony,

Thanks for the other suggestions.

I've checked out the HIjack behaviour, and it's definitely set to 'normal'. I'm going to install a diode internally today, to see if this helps.

The other thing I'm considering is that temperature could be a factor. I've only actually investigated the problem, once I've seen it happening. This usually occurs after a long drive, when the internal temperature of the Empeg (as seen by Hijack) is in excess of 40 degrees. I'll run some cold pull tests, to see if the behaviour remains the same. If it does, then I might think about putting a fan in.

Thanks again,

Andy.

It looks like this could be temperature related. Doing cold tests, the Empeg always powers up in DC mode. As I've seen the temperature creep up to 54 degrees recently, I reckon this could be a factor.

I've now performed the negative spike mod, and also installed a 40mm fan. It's a bit noisy, but the sound is masked as soon as the engine is running.

In my tests, the max. temperature I'm seeing is a much more reasonable 28 degrees C.

I did have a nervous moment, when the display stopped working. Fortunately, it was just a frayed display cable, which was easily fixed. My heart was in my mouth for about half an hour, though.

Andy.

Thanks to Tony and Mark for the suggestions.

I'm going to run with the fan (albeit somewhat slowed down), and see if this solves the problem. If not, I guess I'll have the change the DC socket again (I have got a spare).

Thanks,

Andy.

I wonder if it's not the jack, but perhaps something else. Maybe others can chime in and say what it might be...

But let me make one point clear: You shouldn't need a fan to make the player boot into the correct mode. Temperature shifts might make an intermittent failure easier or harder to reproduce, but that's just a diagnostic, that's not a repair strategy. You have to remember that it is still a *failure* and its root cause needs to be narrowed down and repaired.

If the player isn't booting into the correct mode, then either the PIC that controls poweron behavior is getting the wrong signal, or Hijack is being confused by something and it thinks it's being told to change the player's mode into AC mode after bootup.

If the wrong signal is going to the PIC, the most common source of that is a bad AC adapter jack. But I'm wondering if there are other ways this might go bad. Those who know the motherboard circuitry: What happens to that control line after it leaves the AC adapter jack? Does it go straight to the PIC or some other IC? Is it possible that a dry solder joint on one of the IC pins could cause intermittent AC-power-detection? If so, which one should we look at?

In the MK2, after the AC adapter switch is a 12K resistor followed by a hex inverter IC 74LVX04 which often needs be replaced in MK2s if the detection switch is okay but AC/DC detection fails. In MK2As there is a protection diode that follows the AC adapter switch and sits in between the 12K resistor and switch. See this thread.

Hope this helps,
Stu

Ah, excellent. Just what I was looking for. Yeah, Bagpuss: What he said. check that stuff.

Okay, reading that thread Stu linked... there doesn't seem to be a final solution in that thread yet. It's left dangling.

There is discussion in that thread, describing a the player is behaving fine at bootup with a stock kernel, but having the AC detection problem with a Hijack kernel.

Bagpuss: If you install a stock kernel instead of a hijack kernel, do you get normal behavior out of your player?

So you were pretty sure the faulty serial hardware was the cause of that one?