Unfrickinbelievable!!!!
Yiipppeeee!!!!!!
I HAVE A FADER NOW!!!!!

You have now been knighted.

"Sir Lord The Kernel God"

All in favor say "Here, Here"!!

Okay, v54 works well for forcing DC powermode, except.. it shows the wrong bootup logo. So I've done a more complete job of it this time in v55 (now available).

Your force DC code seems to save its setting in flash... Does this mean it stays forced until unforced even when removed from the car and plugged into mains... and then put back...?

Cheers

Kim

>Your force DC code seems to save its setting in flash...
>Does this mean it stays forced until unforced even when removed from the car
>and plugged into mains... and then put back...?

Yep. And more importantly, it stays forced when you turn off the car to pump gas and then hop back in and start it up again.

Cheers

-ml

OK... I am just trying to work out how to fit my ioctls in with it... I think for my own purposes I might just remove the saving of the setting... but I will remove the option to force AC mode and just use your hijack_force_dcpower variable...

Cheers

Kim

Hi Kim... Dredging up a really old thread here... I also seem to be cursed with a power connector that forces my Empeg to think it's on A/C power all the time. I know there's Hijack's "force DC power mode" function, but I'd like to have it automatically detected, such as you've described here, where it uses the presence of the tuner to decide which power it's on. Did you get this working?

Yes I got this working... very hackish though... and I no longer use it as I replaced my power socket on my empeg when I had to replace the fuse that blew and sent my display haywire...

What I had working was a patch to the kernel that tried to initialise the tuner and if it got a response over-rode the sensed power source to believe it was in the car... I was aiming to get it to the point where it could give the value of the tuner ID over a /proc entry... but then not needing the patch after replacing the socket and running completely out of time to work on anything empeg it all took a back shelf...

I am currently in the process of packing my life into boxes to head on a plane to the UK as of Monday for a year or so... as I don't have a job lined up to start immediately I get there I might get more time to work on the tuner detection patch and the kernel exec patch again in between looking for contracting work over there...

Until then you are welcome to play with what I had... believe my web pages at my last job are still accessible and probably will be for a while knowing them... http://www.ned.dem.csiro.au/CovilKim/empeg/

All work in progress and all not having been touched for a while... Hopefully I can get back to it all in the not too distant future...

Cheers

Kim

Kim,

Well I had to futz with the patch a little to get it to apply to the newer Hijack source, but it seems to be working now. The timeout on AC mode is really large though, I am going to try to tweak your timeout number a little bit so that AC boots don't take so long.

So how tough was it to replace the AC adaptor socket? Is the part an easy one to find? I'm not a soldering champ or anything, but if it's an easy solder job, I'd like to fix it.

Also, I noticed that the behavior is a little different, in that when I put my empeg in the dash with the engine off, it powers up, rather than waiting for the accessory line to go high. That's to be expected given the nature of the power connector failure. However, does this same problem keep the empeg from powering down when the standby timeout is exceeded? One of the main reasons I want the in-car mode to be DC is so that my empeg doesn't drain my battery if I leave it in the car. I guess I'll give it a shot later by changing the timeout, but I was thinking that I might have problems with it not powering down, given that the power socket is flaky.

Anyway thanks for this code, it's definitely a start, and I'm not sure if I'd be able to do the power socket myself.

yeah the part is easy to find.. I picked one up at at local electronics supplier... Just check you get the correct size as there are 2.1mm and 2.5mm ones... the empeg one is a 2.1mm one from memory...

To solder it to the board requires completely dismantling your empeg and removing the mainboard from the case to be able to get to the underside... Other than that it isn't a particularly difficult item to attach... definitely easier and less stressful than replacing the fuse I had to replace to fix my display...

I never knew what normal behaviour was on the empeg until after I fixed mine... and so I didn't take notice of what it did before I fixed it as anything special to be able to remember well enough to compare to what it does now (if that makes sense)... The places in the code that will be causing the strange behaviour will be places where the code looks at the state of the power detection line itself rather than the variable set by the patch... or those times when the patched code hasn't had a chance to run (ie first power up)...

Mind you I haven't had a look at the code for quite a while... 7 hours til I am on a plane to the other side of the planet and a new life (for a year)... maybe in that life I might get a chance to do the coding I want to get to do with my empeg...

Cheers

Kim

Hmm. I guess I'm going to live with the software fix for now, as I don't trust my soldering skills. I can deal with drive upgrades, but soldering irons and I don't mix with good results.

As for the so-called strange behavior where the empeg powers up immediately upon connection, I don't think it's a software thing, because, of course, the kernel isn't loaded until it boots up. I'm guessing the power circuitry uses the leaf switch in the adaptor plug to decide whether it should power up right away (if it thinks it's in AC mode based on the leaf switch) or wait for the key to turn (if the leaf switch tells it that it's in the car.) So I don't think we'll be able to have any kind of software fix for that particular piece of the puzzle. If I'm wrong, someone will correct me, but I don't see how we could fix that in software... I'll have to learn to live with that (not a big deal) for now.

What is a bit of a bigger deal is that, unless I screwed something up when I tried it out, it doesn't seem to power off when the standby timer has expired, probably for the same reason. In the power circuit, it probably figures "hey, there's an AC adaptor plugged in, why should I power down?" or something. That means that when I turn my engine off, it correctly goes into standby, but it never powers off like it does with a properly functioning adaptor plug. At least that's what I'm seeing.

But in the win column, I did find that lowering the TUNER_TIMEOUT to 2000000 gives me proper tuner detection every time, without an excessive delay when I plug it in at home. So for now, your fix is exactly what I need, until I get the guts to solder a new plug on, or decide to send it to Cambridge. I'm kinda upset that this has happened to both of my empegs, but life goes on. My guess is that more and more people will be seeing this as their empegs get into their golden years.