Need help diagnosing a no crank no start issue on a 2000 323 E46 BMW

[poot36]

I have been helping my dad attempt to fix this issue and we think it may be a EWS issue but we are not sure.  My dad has posted on these two forums but has not really got any good advice: https://www.e46fanatics.com/threads/intermittent-no-crank-no-start.1329824/ and https://nam3forum.com/forums/forum/main-forum/e46-2001-2006/283008-no-crank-no-start-issue-need-help

Not really sure where to take it from here.  Any advice would be welcome.

[m k]

Connection problem and a car is not very nice combination.

If problem is a regular connector its measuring can change the situation for quite a long period.
Means that if probing is difficult its practicalities can lead to false good connection.

CAN error and error between controller boxes indicate that it is not a simple switch thing somewhere.

One possibility is a loose connection that leads to a lower voltage and therefore uncertain operation.
RFID is a good candidate for being affected by those kind of things.

Dropped voltage is not necessary very straight forward thing.
I had a Citroen once that had a lifted fuse, by me, for electrical cabin heater.
That fuse didn't burn, but had to be replaced every now and then since its plastic case was melting.
Other Citroen had ashtray light as a headlight height motor fuse, maybe BMW is not that sneaky.

[DavidAlfa]

Normally the starter is directly engadged by the key, then the ECU Will enanle injection or not, but should crank.
Maybe BMW is different so do take it too serious.
I'd check the wiring schematics and trace it out.
In most cars there's a hierarchical power distribution, one big beefy relay powers several other and so on, it's not that hard unless it's modern crap full of control units and magic boxes.

If you randomly have 12V at the starter, trace follow tha wire out, at some point you should find something.

Fuse boxes fail on some cars, the internal connections burn, break or whatever, so don't trust only the fuses.

Having said that, I've checked BMW WDS and it's the worst utter piece of crap I've ever seen, no way to track a signal without becoming crazy.

[coromonadalix]

some cars  now with an transponder

the motor will crank,  no fuel pump    until  the transponder acknowledge  the key / security,    to test  just put aluminum foil to cover completly  the keyfob ...    unless you have in the key itself a small electronic part

for sure   if @op is able to get some electrical schematics ... he could follow  the acc line  etc ...       you have better cars forums then here ...

a random 12v      could be some relay / contacts,  fuse box   intermittance,  corroded   wires, pinched wires in the harness ...   ground issues

good luck

[CapLeaker]

I think you either got water and corrosion in a connector or EWL, wiring, fuse box or a simple key issue. Does it have any codes? Weird to have like 6 keys registered to the car. Does the car complain about the key? Hard to tell without seeing what happens on the cluster.
I had the same problem with a 2002 VW back in 04. The problem was that my key wore out, but I could see a key symbol blinking in the dash.

[poot36]

We have checked all the connections and fuses and they are good and clean.
There are no damaged wires.
If we bypass the relay in the EWS we do get the engine to crank but it will not start due to I suspect the EWS not fully identifying the key as valid and telling the computer to not enable the spark and fuel.
The DME to EWS error code was caused by us when we accidentally turned the car on with the EWS disconnected.  As far as I know when this error occurs the car will crank but not start.
Right now we do not even get a crank.  All the lights on the dash operate normally but when you go to try and crank it nothing changes.  There is no click sound at all.  I do find it interesting that the EWS identifies it's self as for a automatic transmission vehicle but the car is manual.  The other interesting thing is that the key that is normally used in the car (key 5 in the scanner software, key 6 in the AK90+ software does not have any VIN or KM reading stored in it anymore.  I wonder if the section of the EWS that writes to the key is failing and that somehow wiped that section of the key.

[CapLeaker]

Sounds like something went corrupt somewhere. The car wouldn’t happen to have a start inhibit if a gear is engaged or clutch not depressed, or automatic transmission is in Drive?

[poot36]

It does have a clutch switch that has been confirmed to be functional.  According to this document: http://www.e38.org/EWS.pdf (also attached below) that is one of the last things the EWS checks before enabling the DME to allow the spark and fuel pump.  We have the EWS 3.3 version.  I suspect we are failing on the step that the EWS sends a code to the key to get back the rolling code from the key but I am not sure.  We currently have the EWS in the freezer to see if the issue changes with temperature as it seems more likely to start when left overnight.

[amyk]

I'd open it up and check for bad caps and solder joints (especially in a car where it's normally subjected to vibration.)

There are various "EWS delete" modules you can find online if you can't figure this one out.

Of course, there is also the LS swap option if you're really stuck and still want to keep the car

[fzabkar]

I suspect we are failing on the step that the EWS sends a code to the key to get back the rolling code from the key but I am not sure.

I notice this section in the manual:

• The DME will ignore loss of the ISN after the engine is running.

 • The DME retains the ISN information from the EWS II module for 10 seconds after the ignition is switched off.

 Restarting or switching the ignition on within the 10 seconds cancels the key identification process.

As I understand it, you can still sometimes get the car to start. If so, then I would try to fool the DME into believing that the ignition is on even after you switch it off. You may need to fudge one signal pin, or provide external 12V power to the affected circuit. This means that the DME will never again check the ISN.

If the car then behaves for one month, say, you can be reasonably certain that the problem is in the key identification process.

To be certain that the DME is indeed ignoring the ISN check, you might try starting the car with a foreign key, or a new key that hasn't yet been programmed. (Can that ever work?)

Without doing any modifications, try restarting the car within 10 seconds using a foreign key.

To identify the power pins in the DME harness, locate those pins which have 12V when the ignition key is in the Off, Acc and Ign positions. Probe the harness with the DME disconnected.

[poot36]

We have opened up the EWS and all the solder joints look good and the caps are not visibly leaking or bulging.  The EWS did work after putting it in the freezer but also continued to work when we warmed it up with a heater.  We are leaving it overnight to see if it fails again.  @fzabkar This part did cause us confusion as well but there are two rolling codes, one between the key and the EWS and another between the EWS and the DME with the ISN.  As far as I know if the code between the EWS and the DME gets messed up the engine will crank but not start due to the DME not enabling the spark and fuel.  I don't think this is the issue right now.

[poot36]

We had the EWS fail to start the car after about 2 days of starting it stopping it and driving it around the block just in case it broke down again.  So sticking it in the freezer did make it a bit better.  We decided to re-solder all the through hole components on the PCB and so far it has lasted for more then 2 days.  Hopefully it is just a micro fracture in the solder that we did not see.

[poot36]

Well it has been functioning correctly after re-soldering all the through hole components and connectors on the PCB.  I guess it was a micro fracture in a solder joint after all.