MK4 Attempted Repair

[Avora]

The other day I was browsing eBay for my wishlist of things, I came across a faulty MK4 Sat Nav drive for a relatively large sum of money but with an offer option.

Having read much about these sat nav drives, I thought I'd make a daft offer. Clearly this has been for sale for some time as I got it half price, which was around Mk3 money. I was naturally very pleased with myself after all, it could just be an input capacitor.

Well, the item was shipped at a blistering rate and arrived from Ireland the next day. I spent a good few hours yesterday diagnosing and tracing power around the board, there is a nifty high side power switch controlled by 5V. And initial tests showed the eject button lit up and the power light did nothing.

I flipped the board over and saw what was clearly a very small amount of water damage matching with the metal surround in one corner of the board, solder resist appears to be flaking off, and some other nasty corrosion occurring.
The components in amongst all this is a 5V regulator and some smoothing caps. I cleaned the board off with a fibre abrasive pencil and checked for shorts and damage. Clearly someone had been here before and had written the drive off as water damaged.



After some more poking around and generally not finding anything despite the corrosion, I soon gave up and reassembled the thing for in car testing.

Today, I did some testing...the fun really started then. The drive was soon connected up, I opened up the car door and BMW was plastered on the screen. Hopes were raised, but nothing seemed to happen.



The drive presumably crashed. After a power cycle the screen was taken over by the TV module in typical 80's styling as all the BMW forums suggested would happen (alarmingly similar to the standard Rover Mk3 I thought)



Discs would be taken in and spin for a while then be ejected. Not much seemed to be happening. I retreated inside for some more research and decided I needed to attempt a firmware upgrade. After a while of looking around for the firmware, and buying some blank CDs and DVDs I had myself a standard V32 update disc, and a couple modified discs with a Rover splash screen just in case.

To get the car to play ball, I had to put the original Mk3 drive back in and make sure the car saw it, then swap it with the Mk4 (after the Mk3 powered itself down) Once this was done the drive would make one attempt at booting up, in went a disc and low and behold an update screen came up.



Yet another hope rising result only to bring me back down to Earth with another disappointment. My drive already has this software version. Hmm. Back to the drawing board...

Quick as a flash, (at least I bought the bigger pack of CDs) a V29 update disc in hand, and the musical sat nav drives completed. The downgrade went perfectly. Here I noticed the real problem...the screen comes up remove disc and press ok to confirm.



No matter what I pressed or how long I pressed it, ok would not be confirmed. Now, how likely is it that this last piece of the puzzle gets this drive out of its purgatory state just so happens to be impossible.

I have an incline as to what the problem may be. Many drives would have been scrapped long before now, but no, I am now determined .

Just to be sure, I put the V32 disc back in after the usual jiggery pokery and hey presto loaded the software back just fine, but still couldn't confirm with an OK.

I have a sneaky suspicion [read hope] there is a problem with the IBUS connection hence a button press can't work, and the other systems can't see that nav drive. I sincerely hope there is a discrete transceiver or something to interface with the bus I can get at, or at least test. Because if there isn't, this could well be the end of the road for this board.

The board is very well designed, and the extreme last hope is an unpopulated connector with a few lines leading off a serial chip...must be a debug or hopefully a program port.

For now, the saga continues.

Here are a couple more photos of around the board, and under the shielding.