Random battery drain overnight but only sometimes

[StephenFord]

0.3A does indicate about 4W of power. That does seem a tad high for a 'dash' light...

[markjw]

is you car fitted with keyless entry? ex-boss had an issue with a Jag XJ he owned where he would find the battery flat some mornings. What was happening was the table where he put the keys at night was just on the edge of the car's detection - if the key was on one end of the table, no problem in the morning but if at the other end then flat battery. The car was picking up the presence of the key and powering up, waiting a while and powering back down, repeat all night long and dead battery.

[djsubtronic]

It's not keyless, I think it's the instrument cluster that is causing the drain when the dash light stays on (and whatever else that stays on along with it). I've been checking the dash light every night and when it is off, the car manages to start up fine the next day. I have a feeling that the next time the dash gets stuck on will probably lead to a non starting car the following day.

[JW1982]

The instrument cluster display is controlled by CANbus activity. The display is ON when the CANbus network is active. The display is OFF when the CANbus network is inactive (approximately 20 to 30 minutes after the car has been locked) and all CANbus modules go into sleep mode. 

The display not switching OFF indicates that there is a problem that causes the CANbus network to remain active and preventing the CANbus modules from entering sleep mode. In this case every single CANbus module that is installed in the car will remain constantly powered (during operation the modules draw approximately 95% more current than in sleep mode) which causes the battery to drain pretty fast.

Every single CANbus module that is installed into the car can basically develop a defect or software fault and cause the CANbus network to remain active and prevent the CANbus modules from entering sleep mode. 

[djsubtronic]

1 minute ago, JW1982 said:

The instrument cluster display is controlled by CANbus activity. The display is ON when the CANbus network is active. The display is OFF when the CANbus network is inactive (approximately 20 to 30 minutes after the car has been locked) and all CANbus modules go into sleep mode. 

The display not switching OFF indicates that there is a problem that causes the CANbus network to remain active and preventing the CANbus modules from entering sleep mode. In this case every single CANbus module that is installed in the car will remain constantly powered (during operation the modules draw approximately 95% more current than in sleep mode) which causes the battery to drain pretty fast.

Every single CANbus module that is installed into the car can basically develop a defect or software fault and cause the CANbus network to remain active and prevent the CANbus modules from entering sleep mode. 

Ah, thanks for the explanation. That makes a lot of sense. Interestingly enough, since I removed all the fuses one by one the display has been going to sleep 20-30 minutes after locking up the car as expected (so far - been 3 days). I'm hoping it stays that way but I am keeping a much closer eye now that I have a little more insight into wtf is going on.

[djsubtronic]

Well it's been a few months and I thought I'd report back. The battery has indeed died a couple of times, but it was a good month or so before it happened in April. Happened twice in the same week in late May, and then it happened again once this last Saturday. So it seems the occurrence is completely random. I've tried doing drain tests by removing fuses on the days when it has died, but no particular fuse seems to drop the current (as reported by the clamp meter) - except the one that drives the instrument cluster but that is to be expected, and it's a tiny amount of drain around 0.15A.

Wonder if anyone has any ideas what could cause this completely random erratic behaviour on certain days every few weeks / months... and what are the odds it might actually be a faulty battery (battery is a new Bosch S5 that I got in January).

Thanks

[djsubtronic]

Another quick update - I think I have narrowed down the culprit fuse with 99.9% confidence, it's 107 (Instrument Cluster battery Supply and Onboard Diagnostics). When I remove the fuse, there is no drain. I can confirm that the ODO does NOT stay on, it switches off 20-30 mins after I've locked the car as normal. I also used ELMConfig to check all modules for any DTCs, there was one random one which I cleared, but the drain did not stop.

Note, it doesn't seem to draw a constant amount but rather it fluctuates. It stays constant for about 3 seconds, then drops for 1 second, then back to the higher amount. Testing the voltage across the fuse when the drain is occurring shows 2.2 mV (for 3 sec) then drops to 2.1 mV (for 1 sec) and back to 2.2. This translates to a current draw of around 300 mA.

Does anyone have any ideas what could cause this particular circuit to be drawing current even though the ODB is off and no DTC errors are pending?

[Hooded]

On 4/3/2020 at 8:34 PM, djsubtronic said:

Maybe the odo staying on is normal but should it be drawing 0.3A??? That's the main concern here...

On My 2007 car it goes off roughly 20 minutes after locking it. 

[djsubtronic]

10 hours ago, Hooded said:

On My 2007 car it goes off roughly 20 minutes after locking it. 

Yea same, after about 30 mins, but the drain continues through fuse 107 regardless.

10 hours ago, djsubtronic said:

Another quick update - I think I have narrowed down the culprit fuse with 99.9% confidence, it's 107 (Instrument Cluster Battery Supply and Onboard Diagnostics). When I remove the fuse, there is no drain. I can confirm that the ODO does NOT stay on, it switches off 20-30 mins after I've locked the car as normal. I also used ELMConfig to check all modules for any DTCs, there was one random one which I cleared, but the drain did not stop.

Note, it doesn't seem to draw a constant amount but rather it fluctuates. It stays constant for about 3 seconds, then drops for 1 second, then back to the higher amount. Testing the voltage across the fuse when the drain is occurring shows 2.2 mV (for 3 sec) then drops to 2.1 mV (for 1 sec) and back to 2.2. This translates to a current draw of around 300 mA.

Does anyone have any ideas what could cause this particular circuit to be drawing current even though the ODB is off and no DTC errors are pending?

As expected, I left fuse 107 removed last night and what do you know... no drain this morning. So now it just a matter of figure out WHY the ODB system / instrument cluster battery supply is draining and what do I need to do to fix it?!

[djsubtronic]

Ok so I was wrong the whole time. Turns out the alternator was the problem all along. Had a new Remy alternator fitted in spring of 2019 after the original one finally died. I guess it was ok until winter came and that's when the car was often unable to start in the mornings vs during the warmer months. Replaced with an OEM alternator now and haven't had a drain since.

[Kevin#95]

The alternator might have had a bad diode that was leaking current. It's good to know that you've fixed it. It might have been possible to put the current clamp around the b+ red wire and check for leakage with the ignition off. I think it should be a few mA for the alternator with good diodes, but not sure what the exact values should be.