mister pepper

Has anyone had an issue with corrosion on the positive battery cable causing a battery drain? I ask because I'm trying to track down a drain that keeps leaving my battery too dead to start the car every 2 days. The first time it happened I checked voltages running and not and the alternator looked good so I put in a new battery, assuming it was an internal short. In hindsight I should have had it checked first because when it happened again a few days later I had the (new) battery, alternator and starter checked and everything was good.

When the battery is charged everything works normally. No sluggish starting or wonky electronics or anything. I spent some time today checking for phantom drains at the fuses using 2 different voltmeters, after "latching" the doors and trunk and waiting for the car to fully shut down. I couldn't find anything. I also checked with a clamp meter on the positive cable and pulling fuses, but I didn't know then about tricking the car into thinking the doors, trunk and hood were shut.

A few years ago I had issues with what I thought at the time was a bad battery but turned out to more likely be wiring related of the positive cable harness that goes from the firewall to the starter to the alternator. I was getting ~.8V less at the alternator than at the alternator so I ordered a new harness, but when I saw what a pita it was going to be to thread the harness past the starter (catalytic converter needed to come out to get to the engine mount bracket which needed to come off) I decided to run a fat supplemental cable from the firewall to the alternator, effectively making a loop. My assumption then was that there must be corrosion in the positive cable between the starter and the alternator, or possibly before the starter even. At any rate, it worked for a few years.

Fast forward to today. I have a pretty major phantom drain and I can't seem to find it. My only clue to where it might be is that if I disconnect the positive harness at the firewall and measure the voltage from the disconnected, unpowered end to the negative test lug I get somewhere around .7V. That suggests I'm getting some bleed from the positive cable from somewhere prior to the firewall, right? I didn't pay close attention to the polarity of the .7V and I'm happy to check whatever to track it down, especially if I don't have to swap that dang front harness so if anyone knows what to check for or where to look I'm all ears.

SMac770

What car are we talking about? Is this a B8? What year? Is this a US spec vehicle?

The battery on the US spec B8 A4 2.0T has only four connections:

negative post -> J367 BEM -> awg 0 wire -> chassis ground connection near battery
positive post -> B/1 -> J367 BEM plug T2x pin 2 --- this is so the BEM knows the battery voltage drop, there should be no (near no) current flow
positive post -> 110A fuse -> D/1 -> awg 3 wire -> car term 30 wiring harness -> lots of stuff
positive post -> awg 3/0 wire -> TV22 (term 30 junction 2, the jump start block in the plenum chamber) E5

TV22 then has only four connections also:

E5 back to the battery
F6 -> AC suppressor (capacitor) -> chassis ground in engine bay left --- there should be no DC current flow here, this drops off any AC voltage riding on the system to "ground"
D4 -> awg 2 wire -> starter -> awg 5 wire -> alternator B+ --- starter and alternator "ground" to the engine, engine block as this massive strap that grounds to the same place as the fan controller
S204 60A fuse -> B2 -> J293 coolant fans control module --- grounds to engine bay front right

So I'd start by better documenting the places of current flow. With VCDS, if you have it, open to address 19, measuring values, group 18, you can see the current flow the BEM is measuring between the battery negative and the chassis. With engine off, DRLs off, climate system off, audio system muted, I tend to show about -10.3A. Do you see similar? Do you measure the same with your current loop on that awg 0 wire?

The other thing is current out of the battery negative is going to equal the sum of the current in at the battery positive. So if I'm reading 10.3 amps out on the negative, I should read a total of 10.3 amps in across the three paths to the battery positive. B1 should always be near 0 if not 0. With the engine off, I'd expect 0 current on the large wire from TV22. So I'd expect to see 10.3 amps in on the wire on D1. Do you see this? Does the tool agree with the car, and does the in equal the out?

Then extend the issue to the connections on TV22. With engine off, there shouldn't be any current flow anywhere on these lines. Are you measuring any? The only thing I see maybe leaking current with the engine off is the suppressor. If you disconnect the line at TV22 D4, do you measure continuity between the end of that line and the chassis? You should not. It would mean an internal fault in the starter or alternator creating a short.

4ringsnewbie

Do you have a thermo cam like Flir or something similar? You could check which fuse/component is giving off heat and locate the drain that way.

mister pepper

4rings, no, I don't have a Flir or similar thermal camera. SMac, that's a lot of info and it might take some thought to digest it fully. I do have a VCDS, but I almost never use it. I'll try to follow your advice and let you know what I find. Along with doing that I'm going to crawl under the car and see if I can spot any physical damage to the positive harness between the battery and firewall. We had some pretty severe rain here recently and the undercladding along the driver's side got torn up from a deep puddle and I'm wondering now if there wasn't a branch or rock or something that might have caused some damage to the wiring. I have the car charging now so that I can get some full-voltage readings tomorrow. Thanks to both of you for your advice.

I forgot to add -- 2009 A4 (B8) USA build.

mister pepper

I just checked the VCDS and I was getting anywhere from -10.5A to -12.6A. It bounced around quite a bit. I'd say an average of 11.2A. That doesn't seem too far off to me. Battery was only 12.2V in the software, so maybe the amps should be a little higher actually. I measured voltage at the alternator, starting lugs and battery and all were 12.25V to 12.31V, so not much drop anywhere. I disconnected both positive wires at the firewall plate but couldn't get my continuity meter to beep between either the front positive harness and negative jump start lug or when disconnecting the battery and checking between the (disconnected) positive cable and the ground lug in the trunk. Oddly, when I disconnected the positive cable at the battery I still read a positive voltage between the cable and the ground lug. With the cable disconnected at the firewall, so I'm just reading the voltage of the first harness, not anything inside the engine bay, I get a voltage that starts at about 1.5V and drops to settle around .5V after 15 minutes. Is that a capacitor backfeeding to the harness? I don't see similar voltages when I measure the positive harness in the engine bay (disconnected at the firewall) to ground.

SMac770

If you had the door open, that red light adds a decent draw, maybe why it was above the typical low to mid -10.x I see. You had climate control off and the audio system muted (and on radio, not anything mechanical like spinning a CD) and the DRLs off (through the car-exterior lighting menu) and the door shut and no interior lights on, I would not expect it over -11.0. I would be curious what was pulling the extra amps.

So trying to sort out the test statements. I'm not entirely following what is what when you're doing what. :-) To make it easier to understand the statements, I made a drawing so we have a common language to use.


Ground point 12 is under the washer fluid filler neck, near the ECM box. I don't know anything about that suppressor, if it's small, big, near the TV22, near the ground point, etc. It's just a capacitor. Capacitors appear as an open to DC current (once they are saturated) and a short to AC current, so this way all AC that might get picked up from noise or whatever gets shunted to ground and dumped. You'll still measure a DC voltage of 12v or whatever across the cap because the electric field is there, but there won't be any DC current flow. Unless it's leaking. That ground wire at 12 has a lot of stuff grounding there, so to check for DC current flow, you'd want to put your DC ammeter clamp on the red wire going to the capacitor or the brown wire right off of the capacitor.
Ground point 615 is the negative post on the left strut tower.
Ground point 685 is on the right frame. If you look up from underneath between the engine and the under the turbo, you should see the massive cable that grounds the engine block to that point.
Ground point 624 is where the battery attaches.

I assume pulling the positive means the capacitor would start to discharge, so you'd see the dropping voltage until it was discharged, and then everything would be at 0V.

mister pepper

Wow, thank you so much for your help here. I hooked up the VCDS again, only this time I did turn off my DRL's. I thought the **** on the dash turned them off but I had to go into the MMI menu. Now I'm getting 10.4A-10.5A at 11.8V of battery. The battery is draining. I did have the doors pseudo shut by triggering the latch with a screwdriver, so no door lights were on. I was able to lock the car this way and check amps on the cables coming off the battery. With the car powered down for a sufficient amount of time I'm still getting almost .5A at the negative cable to ground 624, almost nothing along the main positive cable from F to E5, almost nothing along the black assuming 3GA wire and .45A along the bundle of 2 red (maybe 6GA?) + A few tiny wires that comes off the 110A fuse on the BEM and heads back towards the rear bumper. That seems intermittent, though, because I just tested it again and everything was near zero amps. Maybe it takes longer to go to sleep than I thought.

Okay, so I just threw the clamp meter on the neg battery cable and let the car just shut down like I normally would and observed what happened a few times. For about the first 1m30s the current draw is around 7A, then some shutdown cycles it drops to .45A for ~16 minutes, and other times it drops to zero. The 7A measured at the negative cable is corroborated by measuring 4.5A at the red cable bundle and 2.5A at the black 3GA positive wire. Essentially no current is going to the engine bay during shutdown.

By the way, I inspected what I could of the under-car wiring and couldn't see anything visibly damaged or corroded. The only thing I would like to have been able to see was the positive lug on the starter to see if there was any corrosion there. A few years ago the cable I added went from F6 to B+ and seemed to solve my weak starting in the mornings issue I was having at the time.

I should probably mention, the car is bone stock electrically with the exception of a USB charger plugged into the 12V outlet in the center console, and that outlet is switched so it turns off when the car does.

mister pepper

Well it started again this morning, so that was nice. Thinking more about this, the easiest place to determine if I have an actual problem is at the negative cable off the battery, between it and chassis ground 624. Any drain i have going out from the positive should manifest itself as a corresponding drain on the negative cable. It doesn't help me locate what is causing the drain, but it verifies that there is a drain happening and the scope of the drain. The only other possibility for a drain that could circumvent the neg cable, according to SMac777's diagram, is the circuit between BEM and terminal B on fuse panel A (battery positive lug contraption). If I'm seeing it right that circuit appears to be two ~20GA wires.

If the drain is not constant, i.e. a system is waking up periodically and draining the battery, then I probably need to take a different approach to tracking it down. It appears the car is going to sleep correctly because after about 20 minuted all current draw off the battery slows to less than 0.1A when measured at the negative cable. Is it possible that something is waking up and draining my battery? With the car "off" and the interior lights dimmed I am still seeing a drain in the 7A region with no visible tells that anything is drawing power. I guess the computer draws a lot. If it is waking up from time to time at night a 7A draw could deplete the battery in just a few hours if the battery was not fully charged. Is the next step to hook up a GoPro inside the trunk and video the clamp meter? Are there any likely circuits that would be waking up that I can pull the fuses for? Is it better to just disconnect the battery every night for a few nights just to make sure it isn't something with the battery chemistry itself, even though this battery is brand new and I only bought it to solve this problem because the thought the other battery had gone bad (it hadn't - both batteries test fine)?

SMac770

Do keep in mind the electricity flows from negative to positive. What is electricity? Electrons. What charge are electrons? Negative. So big negative means many electrons and big positive means few electrons, in a manner of speaking. So the electrical current is the flow of electrons from negative to positive.

So your battery discharging, presenting a reduction of voltage, is due to a flow of electrons from the negative to the positive. This could be via external paths, like the car, or internal paths due to physical defect in the battery itself.

If it's truly reliably consistent that if you let the car sit two days, the battery will be low (and by that you mean 12.0v, 11v, 10v, lower?), then just remove the negative cable from the battery and let the the car sit two days. Either the battery is low like it always is, or its not this time. I don't know your physical environment. Mine sits in my garage so leaving it unlocked or even with the trunk open for days is a non-concern. You can latch the trunk lid with a screwdriver so the car thinks it's closed and you'll still get the beep when you lock the car. No beep, the car knows something is up and it won't necessarily move to sleep as quick.

If the battery is not internally discharging, then it must be an external drain. As you've noted, any draw on the battery is going to go out the only path out the negative, through the BEM and to the chassis. The BEM has a two wire plug on it. One wire is the LIN bus connected to the voltage regulator on the alternator and to the J533 CAN Gateway, which also manages the alternator output level, vehicle energy management, etc. The other wire is the red/red one that goes to the B point on the fuse panel A on the positive battery post. This is so the BEM knows the voltage at the battery. Keep in mind, when you remove the cable from the negative post, you're supposed to unplug that plug before reattaching the cable to the battery post, then reconnect it after the cable is in place. I assume this is to prevent possible spike problems on the LIN line back to other electronics.

All the current you measure going out battery negative cable should equal the current coming in on the two cables on the positive post, the D point where all the car electronics are connected and the F point that goes off to the TV22. Another isolation test, if there is a vehicle electronic device waking up and doing something unexpected, it's connected through that 110A fuse on the D position. Pull that fuse and leave it for two days and see if the battery is fine. With the engine off, there's no reason for any current flow to be noted on the F position, the cable from the TV22.

I very significantly advise disconnecting the negative cable from the battery before any manipulation of anything on the positive side. Once you've pulled the 110A or whatever, then you can put the negative back in place. If leaving the 110A out makes the battery voltage persist without issue, then you're looking at something going on with the vehicle electronics. This is a 2009, so won't have MMI 3G; I think the J794 had an update about power draw. I have a 2009 with MMI 2G and no stray draw issue.

As for the 12v sockets, unless the factory setup has been manipulated, they all run through a relay which is switched off by the J393 when the car shuts off. So there should be no power to them if the ignition (term 15) is off. You could pull the relay, or the fuses for each socket, to ensure such.

"With the car "off" and the interior lights dimmed I am still seeing a drain in the 7A region with no visible tells that anything is drawing power." I don't understand what you're saying here. What do you mean interior lights dimmed? Either the car is asleep and near 0 amp should be noted, or the car is awake and whatever is whatever. When you're reading 7A, is the car aware of your presence? The car should think it is all things closed and locked and all CAN buses in sleep mode (takes 15-30 minutes for that, though less if it's locked, and it beeped acknowledging all good and shut down).

If it turns out removing the 110A makes it all good for a few days, then you have to start breaking down the entire car electronics wiring and determine along which fuse panel or unfused device that the issue might be occurring. The other two tests would be leaving the 110A in place and disconnecting the plug from the BEM (to break the connection back to the B point), and disconnecting the super cable at either the F point on the battery (not sure that's even possible; I've not looked inside the fuse panel A in a long time) or at the E5 point on the TV22. One of the three configurations should not have a problem. It just takes a lot of waiting time to get through them all.

mister pepper

It's been 4 days and the car has been starting normally every time I've used it, although I didn't pull any fuses, move any wires, reset any systems or really change anything. It was random when it was happening, now it seems to have possibly, randomly, gone away. Typically the car would either function normally or be completely dead. It wasn't like a typical battery drain situation where it gets progressively weaker. Whatever was draining it was draining it fully.

The comment about the lights off but still seeing a 7A drain was when I would latch the door handles and trunk with a screwdriver so that the car thought everything was shut, but not locked. I kept a clamp meter on the negative cable and watched what happened with the current draw. The interior lights dim after about 30 seconds or so, as well as any displays still on in the dash, so the car appears to be fully shut down, but I was still getting a 7A drain on the battery until the 1m30s mark, and then a .45A drain for another 15 minutes or so. I wanted to see if the car was going to sleep correctly and it appears that it was.

I'm wondering if the car was (is?) waking up unexpectedly during the night and drawing 7A for an extended period of time and that is what has been draining my battery. The fact that I have gone 4 days without a dead battery says to me that it was something computer/software related and not a "mechanical" issue like a corroded cable causing the drain, since a corroded cable doesn't suddenly fix itself like a random computer issue can. It hasn't rained in weeks so it's doubtful that any water got anywhere it shouldn't. If there is a history of systems or components going bad and causing random battery drains like this I'd love to know what the most likely culprits are so I can pull those fuses for a while, assuming they aren't critical systems. Until then I'll just keep my LiFe jump starter in the car and hope the problem has fixed itself I guess.

Again, I appreciate all your help. I feel like I owe you for your time. Hopefully this thread helps more people in the future track down their random electrical issues. I'll post back if the problem crops up again.