Fiat code box repair?

[digge]

Didnt know where to put this so might be wrong forum part but here goes.

A friend of mine have a Fiat Barchetta and something has gone wrong with the imobilizer. As far as i can tell from what the workshop figured out is that its the code box that's at fault. Now i have that striped out on my desk and not sure what i might be looking for. My first gut feeling was looking for a bad cap. Sure enough there is a lytic in there but im unsure that would actually be the problem.

The lytic is an Frolyt EKS 220uF 35V cap
https://cdn.badcaps-static.com/pdfs/...683971ff5d.pdf

In circuit the values looks reasonable (0,26 ESR 200uF). From what the workshop said it doesnt communicate as it should with the ECU. But it seems to work somewhat cause the warning light for the imobililizer does go out as it should, you cant start the car however.

Any guesses what might be a possible suspect in this case?

Also, was looking at replacements for that cap to maybe try it, nothing to loose since a cap is cheap, would it be a problem going up in ripple, would that effect the circuit at all?

[999999999]

Re: Fiat code box repair?

I'd suspect a faulty transistor but no idea which one. If you can figure out the pinout and how it works then trace the circuit back from that pin it might help.

If all else fails there are services where you mail them the faulty unit and they can pull the programmed code off and reprogram into a new box for much less than a mechanic's shop would charge. For example,
http://www.autotronics.co.uk/cscart/...y-service.html

[b700029]

Re: Fiat code box repair?

The '4013 is a CMOS circuit and may be easily damaged by ESD. Might be worth checking it. There's not much else on the board that's interesting, except maybe the large QFP44 which I can't quite read the part number of, but it seems to be an NEC 78K series MCU.

[digge]

Re: Fiat code box repair?

The QFP44 reads, if its to any help.

Nec Japan
FI IM-1-0
9517XX022
E60

One thing i forgot to mention is that the crystal looks like its half fallen off, maybe does not show that good in the top down image. The reason might be that they wanted to see that the solder had taken under the crystal, i wouldn't expect a crystal to get very hot, at least not to melt any solder?

Ill look into that cmos device, a replacement doesnt sound very expensive.

Im guessing all the programming stuff goes in the QFP right? So anything other then that might be at fault?

[b700029]

Re: Fiat code box repair?

Google gives the weirdest results for that part, so it's probably a custom OEM design.

A crystal should not get hot at all. If the leads are connected it should be alright, although they are shock-sensitive and this being in a car, it's not unlikely it may have been damaged.

You can check the discretes but it'll be very time-consuming (at least they put them all in nice rows ) If you have a 'scope or LA, you could power it up and test its operation. I don't know its schematic or how it's connected to the rest of the car so you're on your own there.

[digge]

Re: Fiat code box repair?

Thats some wierd results indeed I think ill go over it, measure the caps out of circuit and reflow the joints for that crystal and see what that does, maybe clean the card a bit just to make sure. Not that it looks very dirty though.

Later if that doesnt work i might try my luck with that cmos part.

[digge]

Re: Fiat code box repair?

Had in idea, my friend got fooled into buying a new code box even though there is no way he could pair it with his ecu (missing the red key). Would it be possible to lift the programmed chip out of the old box and into the new one? That would rule out anything but the chip i take it.

I must add though that while i do have a hotair rework station i havnt used it much and im not very good at it, maybe i should get some scrap boards and try it out on.

[999999999]

Re: Fiat code box repair?

^ don't know what significance the "red key" is, but if the two boards are functionally the same then yes you could swap the EPROMs. It appears to be the 93C46 chip next to the crystal.

Practice on scrap boards is always a good idea but that looks pretty easy.

[b700029]

Re: Fiat code box repair?

I would rather swap the EEPROM than the main chip (which may be damaged), and rather read the old EEPROM contents and write them to the new one (after doing a backup, of course) than swap a possibly defective component in.

[999999999]

Re: Fiat code box repair?

^ I agree trying to read the old EPROM and write it would be good, but if the old/original EPROM is damaged or corrupted... It'd a lot quicker to swap the chips, though if the new box is left untouched it may have higher resale value (though on the other hand, hard to speculate what the market is for something like that and finding a potential buyer for a one-off sale except maybe listing it on eBay).

[kc8adu]

Re: Fiat code box repair?

is there a big danger of the car getting stolen?
i would just eliminate the thing.
a well hidden switch to power off the ecm and or the fuel pump would do just as well to stop a thief.

[999999999]

Re: Fiat code box repair?

^ I think it needs that working to start the car OR you'd have to build an equivalent circuit to fool the ECM if you remove it. I recently had to build something similar for a friend's GM car with PassKey II, fortunately all the circuit needed to do was send a 50Hz signal to the ECM but I have no idea what this Fiat needs.

[bluto]

Re: Fiat code box repair?

PassKey... what a joke. The best was the first gen that could be defeated with a single resistor that matched the resistance of the 'chip' in the key.

[999999999]

Re: Fiat code box repair?

^ In theory yes, but in practice if you didn't have the key and were trying to steal a car? Good enough for the time period.

How easy is it really to insert a resistor of a specific ohm value into a fake/duplicate key right where it needs to be and know what value it needs to be ahead of time? Someone with those skills probably isn't stealing cars...

Not a joke at all, I bet it's enough to deter you today. It's essentially the same as knowing the combination to a lock, sure "if" there's a way it seems easy, but that doesn't translate into no risk or deterring someone who'd move on to an easier target. Remember these where GM cars, low value!

[bluto]

Re: Fiat code box repair?

A deterrent, yes. But technically a joke especially when the marketing folks would have you believe that the chip key contains some super-duper unhackable military grade crypto processor 9000.

[999999999]

Re: Fiat code box repair?

^ It makes no difference. Even if it were a super duper unhackable crypto link, the time efficient and non-destructible way to bypass it is exactly the same. Instead of hacking the key interface, replace the black box (passkey module) with another.

Granted they could have made that link to the ECM encypted instead... but someone who knew all this would just aim the hack at the ECM itself then. A thief with this level of sophistication could just bring their own hacked ECM, plugging that in is a lot faster than rewiring the passkey II box itself and that type of hack would defeat the security on the majority of vehicles on the road.

[cadiman]

Re: Fiat code box repair?

I think I see a low voltage regulator.

just above the connector on the left side (lower part of the board) looks like a small transister. (3 legs with the top soldered to the board)

Most test equipment uses the 12v and ground of the aldl on the car. the code reader needs less voltage for it`s processor normally around 3.3 volts I am guessing, but that is where I would start.

[digge]

Re: Fiat code box repair?

Just home from trying out some component swaparoo on that box. Started out with swapping the lytic, just for good measure, it measured good but why not. The new one was in the higher tolerance levels uF wise while the old one was in the lower regions.

Turned the key to the first notch and what do you know, the code light turns off??!!?? Start the card and it starts right up but the code light comes back on again. But the car keeps running.

Thinking it must be a grounding fault at this time i start pulling the fuses one at a time to see if it turns the code light of (since it wasn't lit when in first position of the key). However the light stays on.

Turns the car of and tries to start it again and now all of a sudden the code light stays of instead for no apparent reason. I mean i didn't do anything special. Anyways the car is working and all that and my friend is a happy bunny. Just hope it stays in working order cause seems like its something odd going on somewhere.

Oh well, if it ain't broken, don't fix it they usually say

An finally, thank you all for your input!!

EDIT: Spoke to soon, now he called and said it wont work again, but im guessing this means the problem is not inside the box but elsewhere somehow?

[999999999]

Re: Fiat code box repair?

^ That's the other problem with aging cars, you may end up finding intermittent contacts from dust, dirt, corrosion, heat or cold induced mechanical deformity, friction fit connectors lost their friction, wires are brittle or flexed too many times and broke internally, etc, etc.

Examine the connections, and it could help to have a long jumper wire handy with needle probes on both ends to create parallel wiring to the existing wiring as an internally broken wire can sometimes be hard to find as just moving it to get a continuity or resistance reading can (re)complete the circuit.