Page 1 of 3 123 LastLast
Results 1 to 25 of 73

Thread: E31 General Module Sleep Explanation, Experimentation, and Bench Testing

  1. #1
    Join Date
    Mar 2011
    Location
    Damascus, Maryland USA
    Posts
    1,548
    My Cars
    E31/E32 VWs Tesla

    E31 General Module Sleep Explanation, Experimentation, and Bench Testing

    As many of you know, (and is documented in many other threads), the General Module (GM) is responsible for managing door locks, windows, wipers, and other comfort items like interior lights, trunk release, etc. It's also responsible for managing the "Protection" relays by monitoring the current from the relay module. If someone sticks their hand in a window/sunroof while it's closing, causing excessive current, it (hopefully) shuts the circuit down.

    In a properly functioning car, with all the doors closed, lights out, etc. The car initially sits at about 500mA. 16 minutes after the last door is closed the General Module then powers down a handful of relays and modules, and the car's current drops down to 40-50mA. Note: Options like the car phone, alarm, etc. may cause a car to be closer to 50mA, However, My 1991 850i drops down to about 40mA. Now for a little quick math.. With two decent ~80AHr batteries and a 40mA load, the batteries should sustain the car for ~4000 hours (2 * 80AHr * 1000 / 40), which is ~166 days. Granted you likely need at least a half-charged battery to start the car, so we'll de-rate this by 75%, so 125days conservatively. Now, 4 Months should get most folks through winter storage without any kind of charger, maybe an occasional start mid-winter for good measure. Realistically, most batteries are 85AHr, and some may even be 100AHr, so things should last longer with new batteries, but slowly decreasing as the batteries age. Note: this is capacity, and is not the same as the 850-1000 Cranking amps rating)

    However, as many of you also know, the capacitors in our General Modules fail after about 15-20 years. (Like batteries, they dry up and loose effectiveness) More pronounced in colder weather (when battery voltages are lower), bad capacitors do a terrible job filtering the power supply the general module needs to execute code reliably. When a GM with bad capacitors powers up a relay, this may cause a brownout local to the general module, and the firmware may jump to a random place in code, likely leading to a watchdog reset. The GM resets, and after ~2 seconds of 'startup' code, it actuates a relay, and the reboot is cyclic causing the all-too-familiar click-click relay noise in the trunk. The resulting 500mA current (It's actually more than that at 'startup') kills the batteries in 10 days or less. (2*80*1000/500 =320 hrs; 240hrs if we apply the same 75% derating).
    In some cases, when folks start their cars in the winter and drive off, the battery voltage remains low for a little while, and the GM may not boot for several minutes (a combination of voltage and temperature), so wipers/lights/windows/locks may be unavailable initially.

    Many of you also know I've been repairing E31 general modules for several years now, and usually just replace the capacitors and ship them out with moderate success rates. I have several neat tools, including a solder station, solder sucker; I use good quality capacitors, and I restore the epoxy conformal coating on modules I fix, and at a reasonable price that most can't be bothered to hack up their modules. However, in the last few years there have been several modules that didn't work after capacitors replacement, either due to some other underlying issue in someone's car, or some other circuit not behaving on the GM.

    Regardless, inspired by aeronotix' efforts to track down a misbehaving sleep gremlin on his car, I'm attempting to create an isolated bench test for the general module testing, so modules I repair can all be tested prior to shipment. I've spent several days digging through the Electrical Technical/Troubleshooting Manual (ETM), and have performed a bunch of testing on my car (and several modules) to arrive at the following conclusions..

    Things confirmed that keep the car awake:
    -Key in the Run position (obviously, the GM should be powered if the car is on)
    -Open doors (via the doorpin microswitch) will not allow the car to sleep
    -Time (<16 Minutes after last door is closed the car is still 'awake', at 500+ mA)
    -Disconnecting the Black Connector to the General Module (which has only inputs that are seemingly innocuous, see next)
    -Having all 3 wiper relays removed (specifically, the GM expects wiper motor inputs on pins 13&14 to be grounded when powered off)
    -Bad general module capacitors, i.e. a misbehaving/perpetually rebooting module

    Comments: The wiper relays was an interesting discovery, I assume the GM is either trying to get the wiper motor to a known position, or there is a firmware bug that it can't handle the implausible combination of inputs. This discovery begs the question, how many other similar issues may there be in the car? Further, if there are any places where 'Hot-at-all-times' is bridged to 'Hot in run-acc-start', this could cause the GM to think the car is running.

    Things that can wake a sleeping car:
    -Opening door (where doorpin is engaged)
    -Operating the outside door handle with car locked (will turn on the dome light for 4 seconds, but reset sleep timer)
    -Operating the outside door handle with car unlocked (will operate window if up all the way, i.e. window droop)
    -Operating the inside door handle with car locked
    -Operating the inside door handle with car unlocked (will operate window if up all the way minus droop)
    -Operating the trunklid handle with car unlocked (will open the trunk)
    -Turning the car on with key in ignition and/or starting the car
    -Locking / Unlocking the car with the key
    -Locking / Unlocking the car by pulling on lock stalk
    -(Assumed) Unlocking the car with a remote

    Comment: If any of the ~30 inputs / microswitches / wires that are responsible for the previous 10 items are bad, this may manifest itself as a bad module.


    Things that WILL NOT wake the car NOR keep the car awake (Important for creating a bench test):
    -Closing the hood
    -Opening the hood (unless the alarm is armed)
    -Opening the trunk manually with key (unless the alarm is armed)
    -Missing/Disconnected EKM (BTW, with my EKM removed, my car was sleeping at only 27mA!)
    -Operating wiper switch position (wipers, wash, etc; none woke the car)
    -Wiper home position (i.e. manually moving the wiper arms with car sleeping)
    -Turning on/turning off headlamps
    -Windows: Attempting to operate the windows (while sleeping)
    -Sunroof: Attempting to operate the sunroof (while sleeping)
    -Disconnecting the washer (or intensive wash) pump (causes GM inputs to float instead of be grounded)
    -(Assumed) Wiper pressure control CAM position (my car doesn't have this option, so it's floating)
    -(Assumed) Wiper pressure control Relay missing (my car doesn't have this option, so it's floating)
    -Missing Crash-Control-Relay

    Comment: In short, the only inputs on the black connector the GM seems to care about is the wiper relay feedback(13/14), and presumably the ground (26) and 'power for electronics' (1).

    Twilight Sleep States
    -Operating the power mirrors with the car in 16 minute sleep raises the current ~+200mA for 4 seconds after operation, then returns to sleep current
    -Operating the power seats with car in 16 minute sleep state raises the current ~+200mA for 4 seconds after operation, then returns to sleep current (not surprising, it's the same module)
    -Operating the power steering wheel raises current +120mA for about 10 seconds after operation, then returns to sleep current

    Ideas/Items remaining to test (help appreciated):
    -Disconnecting wiper motor
    -Disconnecting sunroof module (assuming it should handle this gracefully)
    -Disconnecting one or both door modules (no idea how it will respond without the door module to 'collect' the dozen or so door inpets)
    -'Hot in start' input without either 'Hot in run/start/accy' or 'Hot in run/start' inputs (implausible signals)
    -Missing relay module (I assume this will keep the car awake without the protection relay feedback among other things)
    -Missing DWS/DWA module (Drive away protection, which I assume is an option that can be deleted)
    -Coding certain features to a GM that aren't present/removed from a vehicle
    -Bad 1K serial EPROM on the GM (not able to read coding, log faults, etc)
    -Door lock motor confirmation disconnected/shorted (likely expects either LOCK or UNLOCK to be engaged, but what about neither or both?)
    -Door lock switch stuck in position (is there any kind of timeout or input error detection? what about a simultaneous lock and unlock key position?)
    -Low washer fluid (note, short to simulate empty reservoir)
    -Measure open circuit floating voltage of all GM inputs to confirm suspicion that they're internally pulled up to +5 or +12V and then grounded to 'activate'

    To summarize:
    I have a few more information gathering tests/experiments to do, and will likely need a spare relay module to complete my bench testing setup (I already have a set of extra door modules).
    My hope is that I can develop a bench test to initially confirm modules are behaving properly.
    Eventually, with enough other inputs expanded (switches/buttons/indicator lights to emulate systems) I hope to be able to use my bench setup to troubleshoot and repair other general module issues (and deal with the 2 bum modules I'm sitting on and possible other modules floating around in the community)
    I'll add that this is all pretty slow moving when everything you do requires a 16 minute validation. :-)
    However, while this information seems primarily useful for developing a GM bench test, it may be instrumental in determining why a car with an otherwise tested-good general module, may be misbehaving in another car.

    Eventually, in addition to my bench-test setup, I hope to create a parasitic leak flow chart, which can include some of these obscure conditions ( including gems like "Are all 3 wiper relays present? Y/N?" )

    If you have any tidbits to add, please do!

    '89 735i, '91 850i, '81 MB 380SLC (For Sale), Tesla Model 3, and VW Passat TDI -- Yes, I still repair General Modules, DM for info!

  2. #2
    Join Date
    Jun 2011
    Location
    NY, NY
    Posts
    298
    My Cars
    84 633, 91 M5, 91 850
    Wow, great post. Why would low washer fluid cause the GM to stay awake? My car actually had low fluid until 2 days ago, but I didn't think to re-test after that.

    Regarding the open doors- are the interior lights supposed to stay on as long as the door is open? Or will they shut off after 16 minutes?

    Also, I have a spare relay module I could send to you, if it would help finish your bench tester. PM me.

  3. #3
    Join Date
    Feb 2010
    Location
    Hampshire UK
    Posts
    8,864
    My Cars
    99 840ci Sport Indv
    Thanks for all your work on this, it is fascinating!
    Timm..2007 E64 650i Individual Sport..1999 E31 840ci Individual Sport..ex owner of 2000 E38 740..1999 E38 740i V8 M62..1998 E38 735i V8..1993 E32 730i V8..1988 E28 518i


    My BMW Repair YouTube Channel
    My Current 840ci Sport Individual
    My Current 650ci Sport Individual
    My E31 Repair and Information Website
    My E38 Repair and Information Website
    My E63/E64 Repair and Information Website

    Chase - Heroes to a generation

  4. #4
    Join Date
    Mar 2011
    Location
    Damascus, Maryland USA
    Posts
    1,548
    My Cars
    E31/E32 VWs Tesla
    Quote Originally Posted by David9999999 View Post
    Wow, great post. Why would low washer fluid cause the GM to stay awake? My car actually had low fluid until 2 days ago, but I didn't think to re-test after that.
    Regarding the open doors- are the interior lights supposed to stay on as long as the door is open? Or will they shut off after 16 minutes?
    Also, I have a spare relay module I could send to you, if it would help finish your bench tester. PM me.
    I would love your spare relay module! I'll send you a message.. and we can discuss terms! I'd also be happy to take a look at your misbehaving modules again (you did have one or two that were acting up, right? check your PMs)

    To answer your questions... In my testing, leaving the doors open will prevent the car from going to sleep, and the interior lights (specifically, the interior lights relay) will stay on beyond 16 minutes. (If you set your switch to the center position, your lights won't come on, but your sleep current will still be elevated by ~50-100mA by powering that relay)
    The purpose of my post was to start to identify all the things that may impact the GM's sleep schedule. You might not leave your door open, but if the door-open switch is bad, or the wire for said switch is shorted to ground in your door hinge, or if the input inverter on the GM is bad from an aftermarket remote start someone installed 15 years ago, (or any number of other faults), the GM may 'think' the door is open, and could prevent your from sleeping. Granted, The door input is only one of about 25 different inputs to the GM, and that easily expands to 50 inputs by the time you include the peripheral modules (door modules, sunroof module, etc).

    I basically reviewed the input/output map for the general module, and initially focused on the black connector being disconnected.. Reviewing the inputs on the black connector (mostly wiper related) nothing jumps out as being a potential problem..
    Low washer fluid was just a hypothetical example. Hopefully low washer fluid doesn't impact the general module's 'beauty sleep', or the sleep current in any way. If it did, hopefully someone would have discovered it by now. That said, I didn't actually test it (yet), but it's on the black connector!.

    Now consider the attached diagram of the wiper relays..
    wipers.png
    You'll see the relay module is responsible for switching the wiper relays on (likely with a transistor-type output), and it works by grounding the 4 pin to the relay coils.. The relay outputs are sent to the wiper motor, but are also monitored by the GM. Based on the 'internal' schematic, these are probably pulled up to some common rail voltage inside the general module, and the arrow indicates it goes to a microprocessor or somewhere else in the module. Theres also a rotational type cam sensor indicated on the wiper motor itself, but the exact details are unknown from this diagram (does it pulse as the motor spins, when it gets to some 'home' position, or something else?What is the gear ratio of the wiper motor to the wipers to the cam sensor? Who knows!

    The general module is responsible for all the logic of the wipers (among a dozen other subsystems), including making sure they go to the bottom when you shut them off.. There is also a service position for changing wiper blades (press the wipe after turning the car off?), low seed, high speed operation. But here's where things are over-engineered.. Intermittent operation.. You'll notice there's no delay switch for intermittent operation like other cars; That's because if you turn your wipers off and then back on when it needs to wipe again, it remembers that time (bounded by a min and max value I'm sure); really, read your manual! And if that isn't enough, the GM also knows the vehicle speed, and it uses that on the intermittent wipers (besides pressure control).. In intermittent mode, it will double the time between wipes when you're stopped at a red light! (or below a threshold speed) Grossly over-engineered!

    Anyway-- Reviewing the attached diagram, which is just 1 of 10 pages of diagrams related to wipers, you can see it monitors the wiper motor position, in addition to feedback from the 3 relays (which select speed and direction in different combinations). Normally, with the wipers powered down, you can see that inputs 13&14 would normally be grounded. The 'bug' I discovered is that if the relays are removed (and the inputs float high), the general module doesn't sleep, probably because it's trying to park the wipers at home or something. For all it's complexity, the 1980s technology in the general module only really has a few kilo-bytes of program memory 16KB IIRC, and with a 1MHZ CPU, they didn't anticipate/accommodate a mis-use case where relay modules were simply missing (or failed, or wiring failed, or dozens of other scenarios... ).

    One more thing to add, the 'state' of the vehicle is important too... Was the battery just connected and it's polling other modules to discover them? Does it know where the windows/wipers are? Does it know if the doors are unlocked? should they be locked? Did the operator just get out, close the door, and lock it from the outside? (Realistically, a 'smarter' module would shut things down immediately, not wait 16 minutes, but there's that 16KB/1MHz limitation again!). Most of my tests were performed after disconnecting and reconnecting the battery, then opening the door, closing the door, and waiting 16 minutes, a sequence I know results in sleep mode(barring no obscure firmware bugs!)

    In another related example.. E32 and E34 modules would retry a door lock if actuating a lock didn't yield the confirmation signal when the lock hit the desired position.. e.g. A cold/frozen lock might not make it on the first try, etc.. But, the early E32/E34 had a firmware bug where a lock that failed to confirm would result in the opposite 'retry' being attempted, and you'd end up with a lock-unlock-lock-unlock cycle when your driver's door lock inevitably broke or developed enough slack to miss the LOCKED limit switch on a regular basis.. (This was fixed in later E32 GM's)

    Anyway-- I'm just highlighting the fact that Electrical gremlins may manifest themselves in very obscure manners, and we haven't even gotten into different potential permutations of configurations.. (Missing sunroof module, implausible combinations of signals, etc)!

    '89 735i, '91 850i, '81 MB 380SLC (For Sale), Tesla Model 3, and VW Passat TDI -- Yes, I still repair General Modules, DM for info!

  5. #5
    Join Date
    Jan 2011
    Location
    Orlando
    Posts
    87
    My Cars
    840 Silver/Silver/Silver
    Potential additional item for your list: I believe that the 840 comes out of sleep for current changes in *either* direction -- both when current is drawn (e.g. lift door handle) **AND** when it is supplied (e.g. connect a charger)!

    At least that's what it looked like to me when I put an ammeter in series with the battery. Of course you want to verify my assertion (for which you need a battery that's good, but not just recently fully charged)...

    Here are the details of how I arrived at this belief:
    I came home from a month-long trip and even though my 840 was on a Deltran Battery Tender, the batteries were dead and its LED was flashing instead of solid. How is this possible??? Well,... My 840 was drawing too much current when idle, out of sleep: 1.8A. The Deltran supplies about 1.2A. But still, when it sleeps 16 minutes later, the draw goes down to 50mA, so the Deltran ultimately charges the battery. BUT in the middle of a Florida storm with the AC power going up and down for very short periods, every time the power comes back on, the car wakes up because the charger does, for another 16 minutes (@ -1.8+1.2A=-0.6A). Plus if the power goes out soonish, the car stays awake for the rest of its 16 min. (@ -1.8A in my case) If the batteries are old enough, and the AC fluctuations are frequent enough, you can ruin a cell with an unlucky run of 16-minute discharges, before you get enough recharging time in sleep. And once old cells go bad, they stay bad. So finally the Battery Tender decides it cannot charge up a junk battery. (Of course whatever-it-is that's consuming the excess current in the first place is the underlying problem. But the old batteries probably wouldn't have died in a month if I had just left it off the battery tender, because it would have gone into sleep and *stayed* there!)

  6. #6
    Join Date
    Nov 2007
    Location
    Belgium
    Posts
    2,481
    My Cars
    Made in Germany
    Excellent information, EEDegreeToDrive! Finally some hard facts on the GM and its gremlins!

    In my testing, leaving the doors open will prevent the car from going to sleep, and the interior lights (specifically, the interior lights relay) will stay on beyond 16 minutes. (If you set your switch to the center position, your lights won't come on, but your sleep current will still be elevated by ~50-100mA by powering that relay)
    I'm not sure whether the GM will go to sleep but I'm quite confident that the interior lights will go out after 16 minutes with the doors open. This is also mentioned in the electrical troubleshooting manual diagnostic procedures (ETMDP) as delayed switch-off condition for the interior lights.

    The wiper relays was an interesting discovery, I assume the GM is either trying to get the wiper motor to a known position, or there is a firmware bug that it can't handle the implausible combination of inputs.
    The GM does not "know" the position of the wipers other than the neutral position (that's the position the wipers rest in in interval mode - not the parked position). This position is read through the reset contact on the wiper motor. Its only purpose is to make sure the wipers do not rest somewhere in the middle of the windshield in interval mode (and to detect blockage). To park the wipers the GM runs the wiper motor in reverse for a brief moment from the neutral position. Once parked, the reset contact is not activated so the GM does not really know the wiper's position. In other words, the GM can sleep with the wipers anywhere on the windshield (as you also discovered).

    The GM however monitors the wiper relays. When the relays are pulled the GM thinks the relays are faulty (sticky) and stores a fault code. I'm not sure why that should prevent the GM from sleeping - perhaps a firmware bug as you mention. Great discovery nonetheless!

    Another possible battery drain related to the wipers is with the pressure adjustment system (ADV) (ECE-spec only). If the K12 ADV module (front power distribution box) is pulled and reinstalled without disconnecting the batteries, it may go into "potential equalization mode" which causes both internal relays to be permanently engaged - even with the ignition switched off!

  7. #7
    Join Date
    Jun 2011
    Location
    NY, NY
    Posts
    298
    My Cars
    84 633, 91 M5, 91 850
    Quote Originally Posted by revtor View Post

    I'm not sure whether the GM will go to sleep but I'm quite confident that the interior lights will go out after 16 minutes with the doors open. This is also mentioned in the electrical troubleshooting manual diagnostic procedures (ETMDP) as delayed switch-off condition for the interior lights.
    It's interesting that there are 2 different conclusions here. Could it be a difference between European and US models?

  8. #8
    Join Date
    Nov 2007
    Location
    Belgium
    Posts
    2,481
    My Cars
    Made in Germany
    Quote Originally Posted by revtor View Post
    I'm not sure whether the GM will go to sleep but I'm quite confident that the interior lights will go out after 16 minutes with the doors open. This is also mentioned in the electrical troubleshooting manual diagnostic procedures (ETMDP) as delayed switch-off condition for the interior lights.
    I did a few experiments: The interior lights do indeed go out and the GM does go to sleep with the doors open but it takes a little longer than expected...

    Approximately 16 minutes after opening the door the interior lights did go out. The total current draw of the vehicle however was still considerably higher than normal. So even though the interior lights did go out, clearly the GM did not go to sleep.

    I had the interior lights in automatic mode. In manual mode the interior lights still lit up. So that confirmed the GM was not sleeping (if it was sleeping the K72/K73 consumer cut-off relays would be disengaged and there would be no power for the interior lights at all). All the GM did after 16 minutes was deactivating the automatic interior lights through relay K29.

    It did not make sense to me that an open door could prevent the GM from sleeping so I decided to wait longer... Another 16 minutes later, the interior lights (now in manual mode) went out and the current draw dropped to around 50 mA. Finally the GM was sleeping.

    Quote Originally Posted by David9999999 View Post
    It's interesting that there are 2 different conclusions here. Could it be a difference between European and US models?
    There are no different market versions of the GM but the module had several hardware and software revisions over the years. It's possible the early modules have slightly different behavior but in this particular case I think EEDegreeToDrive simply did not wait long enough.
    Last edited by revtor; 08-01-2017 at 04:20 PM. Reason: typo

  9. #9
    Join Date
    Mar 2011
    Location
    Damascus, Maryland USA
    Posts
    1,548
    My Cars
    E31/E32 VWs Tesla
    That's a good & interesting discovery, thanks! (I didn't have my interior lights in auto mode, I had them off to save my battery, but I was using a map light for simple sleep confirmations, in addition to my meter, of course). As noted in my testing, the memory seats/memory mirror/power tilt steering modules also caused a brief increase in current for a few seconds, but did not actually wake the GM or interrupt it's sleep cycle. If anyone has their doors apart in the near future, I'd love to know what happens if one (or both) of your door modules is disconnected! Does it still sleep without a door module?
    Last edited by shogun; 04-09-2020 at 10:01 AM.

    '89 735i, '91 850i, '81 MB 380SLC (For Sale), Tesla Model 3, and VW Passat TDI -- Yes, I still repair General Modules, DM for info!

  10. #10
    Join Date
    Jun 2011
    Location
    NY, NY
    Posts
    298
    My Cars
    84 633, 91 M5, 91 850
    Bumping this thread- I recently fixed my GM "awake" issue on my car. With the help of a friend of mine and some files from revtor, I was able to get INPA/NCSExpert set up on a dedicated computer to read the errors on the modules in my car. I found that the GM had an error for the windshield wiper pressure control (my car is US version and does not have this option), as well as an error for the "DRM 2 ADV" (not sure exactly what that is, but I also found references for it on euro cars only).

    I used NCSExpert to reset the module to factory, and my car now enters sleep mode correctly, with a resting current between 50-60 mA. My car did not exhibit any electrical issues before or after the coding process. Really happy to have solved this on my car!

  11. #11
    Join Date
    Jul 2014
    Location
    Westfield, IN
    Posts
    1,313
    My Cars
    '91 e31 M70
    This is very interesting. I have been enjoying driving mine in the nice weather we've had, but plan to dive back into this issue this winter. So bookmarking this as something to try.
    Last edited by shogun; 04-09-2020 at 10:01 AM.
    '07 e65|N62 - Alpina B7 - Black Sapphire|Black|Loaded|GONE
    '02 e46|M54 - Orient Blue|Beige|'Vert|GONE
    '00 e39|M52 - Anthrazit Metallic|Grey|GONE
    '01 e38|M73 - Anthrazit Metallic|Schwarz|Shadow-Line|GONE
    '91 e31|M70 - Schwarz|Schwarz|M70|FOR SALE

    '13 328i Auto - Black Sapphire|Coral Red|Sport Line|GONE
    '05 545i SMG - Jet Black|Black|Bamboo Anthracite|GONE
    '07 335i Auto - Montego Blue|Grey|GONE
    '05 545i SMG - (Euro Delivery 7/22/05)|Black Sapphire|Auburn|GONE
    '97 528i Auto - GONE

  12. #12
    Join Date
    Jun 2011
    Location
    NY, NY
    Posts
    298
    My Cars
    84 633, 91 M5, 91 850
    Just did some quick googling and figured out that ADV is the "Anpressdruckverstell," or "adjusting contact pressure" in English. So, my theory is the GM was replaced by the PO with a unit from a European car. In my car, it was looking for the wiper pressure adjustment and could not find it.

    For anyone with electrical issues, I would highly recommend getting INPA. It shows a huge amount of information. If anyone is local to the Fairfield County CT area, I would be happy to share my setup.

  13. #13
    Join Date
    Feb 2010
    Location
    Hampshire UK
    Posts
    8,864
    My Cars
    99 840ci Sport Indv
    ...Or it could be (just like the EGS) that the module suddenly decides it has options that were not originally fitted!
    Timm..2007 E64 650i Individual Sport..1999 E31 840ci Individual Sport..ex owner of 2000 E38 740..1999 E38 740i V8 M62..1998 E38 735i V8..1993 E32 730i V8..1988 E28 518i


    My BMW Repair YouTube Channel
    My Current 840ci Sport Individual
    My Current 650ci Sport Individual
    My E31 Repair and Information Website
    My E38 Repair and Information Website
    My E63/E64 Repair and Information Website

    Chase - Heroes to a generation

  14. #14
    Join Date
    Mar 2011
    Location
    Damascus, Maryland USA
    Posts
    1,548
    My Cars
    E31/E32 VWs Tesla
    Here's the deal-- the module can write to EPROM to store faults, and when coding, it has to be able to program the coding into the serial EEPROM.. When the capacitors fail, the localized power supply for the microcomputer is not a clean regulated supply. At some point the module does a brownout (likely with a high current thing switched on) and all bets are off.. The RAM gets corrupted, maybe the program counter gets reset, stack overflow (what little stack a 1MHZ CPU has), etc.. The CPU may then jump to a programming vector, or might try to log a fault to the (1kilobit=128 bytes) EEPROM at wrong address (a configuration section instead of a logging section) and bam, the car now thinks it should expect the wiper pressure relay, headlight washers, etc.. (note, this would have nothing to do with the capacitor repair, since the EPROM would have been corrupted prior to the fix)
    As I've demonstrated with removing my normal wiper relays, if they're missing the module gets upset due to the implausible state (bad conditional flow logic, programmer forget else/default cases).. It's safe to assume that the same mistake was made for other subsystems, and other modules for that matter (EGS, Timm?) In this respect, a car programmed for, but missing wiper pressure relays/motor (ADV), or any of a few other options could misbehave like my car without wiper relays!
    Sounds like I need to try and get INPA setup and reprogram my other dud modules.. I suspect at least 1 of 3 will be salvageable! The developers could have also gone the extra mile and computed a Cyclic Redundancy Check on the module programming and wiped it if there was a problem, but it seems like that didn't happen either!
    Another possible problem is that the Serial EEPROM could be bad.. a 1K EEPROM is only 128 bytes, but they only have limited write/erase cycles, on the order of ~10000. A 'regular', recurring fault could cause an EEPROM to fail, and then the coding (and faults) wouldn't be saved properly, and it only takes 1 lazy bit to code in an option that doesn't exist..
    The proper solution is likely to reprogram, read back, and if reading back doesn't come back the same, then maybe the EEPROM needs changing too...
    Of course, another option might be to ground the feedback loop on default options not present, like the ADV relay confirm, but that introduces a whole bunch of other unknowns!
    Thanks for everyone's help and input on this topic, I think we're on to something!
    Last edited by shogun; 04-09-2020 at 10:03 AM.

    '89 735i, '91 850i, '81 MB 380SLC (For Sale), Tesla Model 3, and VW Passat TDI -- Yes, I still repair General Modules, DM for info!

  15. #15
    Join Date
    Aug 2009
    Location
    Chicago
    Posts
    56
    My Cars
    BMW X5
    Hey guys, I am glad I have found this discussion. I have been having similar battery drain issues with my e53 / X5 4.4 2001 for a couple of years now and about to replace my gm3 capacitors unless you have different opinions.
    Simply, once the weather is cold outside, the battery is drained pretty much overnight. I have replaced 3 interstate batteries (thank god for warranty) because the drain eventually completely ruins the battery and it becomes unchargable. Now the interstate dealer knows me so I am not sure if they'd replace the battery again, so I have to find a solution for this.
    I checked the drain Amps when car is supposed to go to sleep after 16 mins (when its under 30f or so outside) and I read a little more than half an Amp. I also see the Shift led still on (as well as the little IHKA fan still running) so the car obviously does not go to sleep. The car is fine during summer when its hot outside (I live in Chicago) but when its cold, the battery is dead pretty much the next day or so.
    I checked things with INPA and I do not see any errors other than 'footwell flap motor'. I am not sure if that is related to the same issue.
    Any opinion / help is well appreciated..
    Thanks, Ozzie
    Last edited by shogun; 04-09-2020 at 10:04 AM.

  16. #16
    Join Date
    Mar 2011
    Location
    Damascus, Maryland USA
    Posts
    1,548
    My Cars
    E31/E32 VWs Tesla
    GM Capacitors sound like a good idea, but this is the wrong forum for X5 discussion. The only thing that gives me pause is the shift light and IHKA fan running, which might be a separate issue, anything ranging from a bad IHKA, bad ignition switch, bad/sticky relay, shorted wire, etc. Footwell flap motor might be peripherally relayed if the IHKA is in an unhappy/inconsistent state and may be retrying things. Again, GM caps are probably a good idea, but you'll need to reevaluate and revisit the issues (in the X5 forum) afterwards.

    '89 735i, '91 850i, '81 MB 380SLC (For Sale), Tesla Model 3, and VW Passat TDI -- Yes, I still repair General Modules, DM for info!

  17. #17
    Join Date
    Aug 2009
    Location
    Chicago
    Posts
    56
    My Cars
    BMW X5
    Thanks, I only posted here because the issues would be a result of a similar module malfunctioning and the issues were very similar. I thought the system not going to sleep would be a result of the same (general) module. No sleep issue happens even when the IHKA is not running. (its power unplugged).
    Thanks again, Ozzie
    Last edited by shogun; 04-09-2020 at 10:05 AM.

  18. #18
    Join Date
    Aug 2009
    Location
    Chicago
    Posts
    56
    My Cars
    BMW X5
    Thanks, I responded here because the issues were very similar and would be a result of the same module in a different car. No sleep issue happens even when the IHKA module is plugged off.

    Thanks again,

    Ozzie

  19. #19
    Join Date
    Mar 2011
    Location
    Damascus, Maryland USA
    Posts
    1,548
    My Cars
    E31/E32 VWs Tesla
    Yeah, the general module function, operation, and failure modes may be similar between models/modules, so a GM capacitor repair may address your sleep current problems. Your IHKA issue/concern may be a completely separate issue though. Also, Be careful making assumptions with things unplugged/disconnected/fuses removed as some systems are receiving power from different sources, and running one without the other may be unpredictable or lead to false conclusions. (e.g. What happens when you run Radio "Switched +12V" without the "Hot-at-all-times +12V"?)
    Last edited by shogun; 04-09-2020 at 10:05 AM.

    '89 735i, '91 850i, '81 MB 380SLC (For Sale), Tesla Model 3, and VW Passat TDI -- Yes, I still repair General Modules, DM for info!

  20. #20
    Join Date
    Aug 2009
    Location
    Chicago
    Posts
    56
    My Cars
    BMW X5
    Quick update : The GM capacitor repair did not seem to help. As I was testing other things with INPA, I noticed each window module (PM) had a .4 amp of drain even when there was no movement on windows up or down, even when the engine was off (before sleep). I am not sure what would cause that or if that is normal. I will test to see if that continues after the system goes to sleep.
    Last edited by shogun; 04-09-2020 at 10:06 AM.

  21. #21
    Join Date
    Aug 2009
    Location
    Chicago
    Posts
    56
    My Cars
    BMW X5
    Another update :

    I am posting this here since nobody is helping on E53 forums.

    The car seems to go to sleep at 32 mins when its colder outside. Drainage amount starts with .95 Amps, then goes down to .5 Amps for 15 mins. Right before system goes to sleep something seems to wake up the system and it goes back to .95 Amps. It eventually goes to sleep at 32 mins (50f outside, I suspect that time changes when its colder) at .04 amps which I believe is normal. I replaced the capacitors already.

    Again, I am posting here because I am not getting any responses on other / related forums..

    Any help would be greatly appreciated.

    Thanks..

    Ozzie

  22. #22
    Join Date
    Mar 2011
    Location
    Damascus, Maryland USA
    Posts
    1,548
    My Cars
    E31/E32 VWs Tesla
    Sleep current for most cars should be less than 50mA, 500mA is still way too high..
    The specific items I listed in this thread, as they pertained to 8 series, were basically designs from the mid-late 80s, and included some of the idiosyncrasies of multiple module interaction and module coding/misconfiguration. In the case of the E31, certain discrete signals disconnected or things not matching coding could cause problems, likely from poor software practices at subcontractors like Siemens and Bosch once upon a time. Your X5 is a generation or two newer, so there is likely nothing the same, and few E31 folks know anything about the E53, myself included. There may very well be another E53 module besides the GM that is capable of waking the E53 and it may need some sort of service. That said, you'll probably need to start a similar thread for E53 sleep current analysis, and you'll likely need to get some real BMW diagnostics software up and running to figure out what's not behaving properly (a $15 OBDII adapter might not be that helpful). If the E53 forum isn't helpful here on BF.com, you might need to find out where the E53 folks do gather to discuss (check facebook, or other BMW forum sites). Sorry I can't help you any more than that!
    Last edited by shogun; 04-09-2020 at 10:07 AM.

    '89 735i, '91 850i, '81 MB 380SLC (For Sale), Tesla Model 3, and VW Passat TDI -- Yes, I still repair General Modules, DM for info!

  23. #23
    Join Date
    Aug 2009
    Location
    Chicago
    Posts
    56
    My Cars
    BMW X5
    Thank you. I still greatly appreciate your help..

    My car eventually goes to .04 amps in sleep mode but somehow after two tries. So it takes 32 mins to go to sleep. Right around 15 mins, something seems to wake up the system and it briefly goes to .95 amps for a minute, back to .5 amps for another 15 or so mins and system goes to sleep at .04 amps at 32 mins.

    I do have the INPA. I checked many things and only error I see is (as I mentioned before) footwell flap motor error related to IHKA.

    Thanks again..

    Ozzie

  24. #24
    Join Date
    Mar 2011
    Location
    Damascus, Maryland USA
    Posts
    1,548
    My Cars
    E31/E32 VWs Tesla
    Slight Update---
    I finally got an INPA system up and running on an old Dell laptop with serial port, complete with an OpenADS interface from EBay-- For the first time I was able to talk to the EKM, LKM, IHKA, and ZKE2 (GM) and check out real time values!

    (As an aside, when I was playing around with the LKM and activating headlamps/fog lights/etc with my computer, I couldn't get my pop up headlights to work.. Turns out the concealed headlight relay sitting was on my desk in the house. Doh! )

    I then pulled my known good GM out of the car, and installed a local guy's GM that I refurbished but still didn't go to sleep afterwards.. Upon checking the coding in INPA, it seems its programming had become scrambled. (It was set as a European model, with headlight washers, wiper pressure control, etc.. Of course, the best part is that its manufacture date was ~2044, so the EEPROM was definitely scrambled somehow)

    So I decided to figure out how to reprogram this thing. I quickly realized that INPA can't actually recode things, so I messed around a little bit with NCS Expert, which is a bit scary when you first look at it.. (the german doesn't help much);
    After a little fumbling around and trying carefully not to erase all the other modules in my car, I was able to recode Kevin's GM to a US variant GM. (The coding sticker in the fuse panel was super helpful)
    Restarting INPA, I confirmed that the options now looked good and it also stated it was now a US module! (Though still made in 2044, an EEPROM field that would take a little bit more work to try and repgrogram; If I was a motivated engineer, I'd "re-certify" the repaired module with the date that I changed the capacitors, but I digress :-) )

    Anyway-- Repeating the current test in my E31 with Kevin's troublesome GM revealed it was...
    ...still not happy. ~500mA. Hmmmmmph.

    Not giving up quite yet, I remembered a snippet of info I read about coding and some things needing power cycling to take effect. I disconnected the battery; Let the car think about what it had done for a few seconds, (or rather, forget)..
    Reconnected and repeated the current test..

    16 minutes later, it was down to 40mA! Module fixed.

    So here are the conclusions I'm drawing..
    1) Certain inputs, if not satisfied, can cause the module to stay awake, a firmware bug in the GM. The most obvious offense is that removing the wiper relays (or a bad wiper relay) can cause this!
    ...it's not much of a stretch to also assume that
    2) Installing a module coded for a different region or coded with 'missing', options can cause the module to stay awake for similar reasons.. (Perhaps coded with missing options like headlight washers and/or wiper pressure control?)
    3) A failing module may be capable of scrambling its own EEPROM, as evident by the 2044 year of manufacture, and European config in a US module (0x00 is European, 0x01 is US)

    How is #3 possible you ask?? The module's software has the capability of writing to the 1Kbit EEPROM when it is (re)programmed and perhaps for saving logged faults. Now, When the GM starts flipping out due to bad capacitors, and less than ideal power fluctuations on the module, it's certainly possible that the CPU's program counter or tiny stack can become corrupted, and it might jump to the reprogramming vector of code; with a corrupted stack, all bets are off and the module basically scrambles the EEPROM. (It might also log a bunch of very strange faults) Of course, A checksum or some code to handle implausible values could have been used to reset things to some default, but we've already established that BMW/Siemens had less than ideal software practices in the 80s, as evident by the bad-coding-keeps-car-awake bug )

    So-- If anyone has any refurbished general modules sitting around that still didn't solve their sleep current problem, PM me and we'll see if there is anything reasonable we can do about that!

    '89 735i, '91 850i, '81 MB 380SLC (For Sale), Tesla Model 3, and VW Passat TDI -- Yes, I still repair General Modules, DM for info!

  25. #25
    Join Date
    Feb 2010
    Location
    Hampshire UK
    Posts
    8,864
    My Cars
    99 840ci Sport Indv
    That's great news - we always had a suspicion that other modules (such as the EGS, seat and steering-column modules) 'reprogrammed' themselves to activate inputs and outputs for non-existent options and you have confirmed that with your tests. I don't think that is due to a failing module though - as this affects modules that don't actually fail (such as the EGS), it is possibly due to undetermined behavior under low-voltage conditions or something similar.....
    Timm..2007 E64 650i Individual Sport..1999 E31 840ci Individual Sport..ex owner of 2000 E38 740..1999 E38 740i V8 M62..1998 E38 735i V8..1993 E32 730i V8..1988 E28 518i


    My BMW Repair YouTube Channel
    My Current 840ci Sport Individual
    My Current 650ci Sport Individual
    My E31 Repair and Information Website
    My E38 Repair and Information Website
    My E63/E64 Repair and Information Website

    Chase - Heroes to a generation

Page 1 of 3 123 LastLast

Similar Threads

  1. General Modules and EURO LKMs
    By TxGR8White in forum 8 Series (E31)
    Replies: 9
    Last Post: 02-21-2010, 09:29 PM
  2. FS: M20 DME and General Module
    By SpeedsterBek in forum 1988 - 1996 (E34)
    Replies: 0
    Last Post: 05-09-2009, 03:00 PM
  3. Priority: Relay and General modules for E34
    By KBHumphrey in forum BMW Parts Wanted
    Replies: 3
    Last Post: 04-26-2009, 03:56 PM
  4. Difference between Relay and General modules
    By KBHumphrey in forum 1988 - 1996 (E34)
    Replies: 3
    Last Post: 04-26-2009, 03:53 PM
  5. Window Wiper Motor and General Module
    By sloralyn in forum General BMW Mechanical Help sponsored by RM European Auto Parts
    Replies: 2
    Last Post: 01-12-2009, 10:27 PM

Bookmarks

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •