Thread: Renault "Tuner List" Head Unit/CD changer hacking - Controls

Originally Posted by pippolippi

Well, my quest for a suitable circuit hasn't ended yet
I have an inverter to power my laptop in the car (a belkin ac anywhere 140w) and as soon as I connected the laptop to the inverter, with the interface connected, the laptop shut itself off. I thought I fried my laptop, but luckily that wasn't the case: after removing the battery (and the serial port) it went back to life.
So I thought I fried the serial port, but, no, it's working (phew), but as soon I connect the serial port to the interface it disappears (i.e. it doesn't exists anymore, like the usb adapter is freaking out).
I'll probably have to put optocouplers in both directions (and for the spdif too) if I want to use the inverter.

Jeesh, scary... Can it be a grounding problem ? Do you have the problem even if you don't connect the HU side ?

Something like this (you'll have to visit this page first to agree to the GPL license)

Yes, opto-isolation is always a good precaution. In this case however it would need 3 optocouplers (Rx/Tx/Remote)...

Originally Posted by pippolippi

Well, since it's a simple communication I find it more understandable if it is in a single, short, file, instead of being scattered in 50 classes in 200 source files
Hope you'll appreciate that the program is rather simple, even if you don't grok python.

Yes, undoubtedly, it's rather straightforward.
I hope you won't come into concurrency problems if you have to both retry a frame and receive another one for example, but future will tell.
In my case, there are already about 15 classes that simply represent the kind of frames that can be received or sent. Add the threads for listening and sending frames, a watchdog running every second, the finite state machine to store the emulator status (BOOTING/PLAYING/PAUSED/CUEING/etc), a few utilities for dumping pretty hex debug logs and the plugin architecture... Maybe not 200 files but definitely not far :-)

But I admit python is particularly clear and concise.

feel free to upload it.

OK. I'll do it as soon as you have tested it.

Now a question, as you see from my other post my problems aren't over yet.
Are you powering your car pc with an inverter?
If so, are you having the same problems I had or you just tested the circuit with another pc?
(yes, I know you've already connected the spdif output of the car pc, but maybe there's an isolation transformer, either in the pc or in the HU input, if the latter an optocoupler wouldn't be needed for spdif. The fact that spdif connector in the HU has a different gnd seems to confirm this possibility).

You are right, maybe I should describe my setup a bit.
The carPC isn't running on an inverter. It has a M1-ATX as DC/DC converter and has worked with SPDIF (and the actual CDC for Data) for nearly a year now without a single problem (but SPDIF GND is not connected to the common GND of the PC.
BUT :
All the hacking tests I've done on this project are not using the carPC. I'm using a 1GHz PIII desktop PC powered on mains (220V). The only precaution I take is to always touch both the PC case and the HU at the same time before connecting the adapter to avoid ESD. The adapter has continuous ground from the DATA GND PIN of the HU to the GND pin of the DB9.
And I haven't had such a problem until now.

Now I don't think connecting the adapter to the car PC can make things worse, except for a ground loop maybe...

Slimmegane, can you describe your setup a bit ?

Originally Posted by Vicne

Slimmegane, can you describe your setup a bit ?

I've Epia 10000 and M1ATX, but my spdif is from soundblaster audigy 2 nx (usb 2.0 port) I've connect the ground to the rs232 ground and the pin 15 of mini-iso to the ground of the PC, all PC ground is the same of CAR Ground.

I think that the problem of pippolippi is the ground that is more different from PC and car.

About the girder and java try to read someone in the girder forum http://www.promixis.com/phpBB2/index.php I think you can use the LUA plugin to send simple command to Girder or Internet Event, now I'm wait an answer from girder forum

Originally Posted by pippolippi

Well, that's the actual program I used to do the test, so it works
However I'll do no more tests until I find some optocouplers...if I can resist doing no further tests

So, your setup works or it doesnt't ? I admit I'm a bit lost... :-)

Well, that's an inverter, but it's possible that it keeps the same ground as the car. In my case is a double inverter: 12V to 220v (and maybe the ground goes straight to one of the 220V socket) and then the laptop brick, so I'm not surprised the grounds (and all other levels) are messed up. Until today I had only tested with the laptop battery.

Oh, I see : it works with laptop on battery and doesn't once you use the inverter ?

you mean the epia has an isolation transformer? or is the HU that has an isolation transformer on input?

No idea :-)
I just let this ground alone. Maybe I could have connected it to the car's ground, but I didn't see the point in doing that...

well, it depends on how your power supply works, if it's as descripted in the VWCDPIC project it's not going to work (though I don't think you'll fry your serial port, I didn't fry mine after all and with a worse setup)

Well, as far as I remember, the M1 ATX connects the car ground to the motherboard ground, contrarily to what is explained on http://www.k9spud.com/vwcdpic/invertedpower.html