Mdk! Wiring harness

[bjornk]

Replaced mdk a few weeks ago with brand new mdk . Ran like a banshee for about two weeks. I felt great relief. The other night while car is sitting at idle at a store, car goes into eml . Since then has gone into eml several times. I’m certain my connector is on there good, I’ve done this job a few times now,,,,

So I’m guessing to replace a few cheap things : accelerator/ Bowden cable. And possibly harness . Think I read on here you can replace just the mdk part of harness from mdk to dme? I’ve found part numbers for the connector but not a harness.. pleiades .. I see in another post you listed the part numbers, did you build your own harness? Any special wire?


Also planning to run through a check of my grounds, I know one of my grounds looks pretty shot.

Guy who had car before me had a pretty serious system in there. There was a lot of battery corrosion in trunk, I’m wondering if I find some shorts/ drains/ issues somewhere ...everything’s is now gone... but not sure who clean install and de install was

Obviously I’m worried I just threw money at a problem and now don’t know actual problem. I suspect wiring and connectors.

One thought was, new mdk literature says dme must be running latest version or something to that effect. I suspect my car has not been to a dealer in quite a while. Maybe the dme needs updated??



Also thinking of running this car to an electrical guru if anyone knows anyone in ne fl or tampa area,,, please let me know.

I found this but hoping I don’t have to build the harness
http://www.e38.org/breakoutboxesandconenctors.pdf


Pulling all the same codes it was before:




Sent from my iPad using Tapatalk

[effduration]

I am sorry to hear this... I was hoping that the new MDK would solve your problem. Unfortunately, these MDK's in the M52TU cars are very weak "links" (and are no-warning stranding issues) with expensive fixes. They can have several causes: Bad MDK, bad harness, bad connection/ground, maybe a bad throttle cable, etc.

I have been through several used MDK's and have replaced my throttle cable. I just bought a spare used engine harness off ebay that I intend to swap in when I go into MDK-related limp mode. I would remove the intake manifold and swap the heater pipes while I am at it. Yes, a used harness may have the same problems that my original harness has.

Maybe others will respond, but I don't hear of too many people separating the MDK harness from the engine harness. I suppose it's possible, but that harness is an absolute octopus. If we could do that easily, we could (in theory) bypass and test the engine harness and run the 10-wire (is it 10?) round plug directly from the DME to the MDK. If I get another harness, I might try and create a separate harness. Interestingly (or not), the auto and manual version of the '99-'00 528i used the same engine harness - Part #12511439175.. I might also look and see if the separate e39 auto trans harness used that same round connector. That might be a good source of separate MDK lead.

[geargrinder]

I'm with you on the harness problem. Those codes are clearly happening step 1: 'diagnosis' error 118 -> bad resistance readings on the pot sensors, then step 2: because of that 169 -> go into 'limp mode' on the MDK... So unless your new MDK is DOA (unlikely to be exactly same behavior as last one...) then indeed something else wrong. Have you checked DME connectors for any contamination / corrosion?

If car ran OK before I doubt DME update is a big deal... unless somehow in the later revisions they made the diagnosis less overly-sensitive...

BTW 12.04V is pretty low... that might hint you do have a bad ground interfering with your voltage? What voltage do you get w/ engine running at the battery, and, at the jump-terminals in the engine bay?

[pleiades]

Sorry to hear about this. If you installed a new MDK (not rebuilt or something else), you shouldn't be getting any failsafe mode from that part, so the finger then points at the harness, or .... did you really really get that connector socket reset and back on correctly? That thing can be a real beyatch.... I couldn't get a grip on mine well enough to spin it back on by hand... had to grab it with some channel locks.

I found (or rather, was given) part numbers for the harness connector but didn't buy one. I -did- check every pin between the connector and the DME side for continuity, but nothing more.

[bjornk]

I took the thing by a reputed electrical guru.he is so backed up he said he’d be happy to look at it but it would take a week maybe week and half.

I guess I’ll try the Bowden cable and double check those connectors and grounds one last time. I had a tech say the Bowden cable can’t trigger eml. But you’d think the mdk has some way of measuring the plate opening and bases its measurements off that and that play in the cable could be bad. I have noticed the screw that keeps it tight seems to get loose.

Seem crazy to spend 150 on a diagnosis if it’s just a 400$ harness. That’s a big percentage of the total cost harness job sounds straight forward .

GG. I think that test was with the door open ac on and engine off. But I’ll check again. I’m gonna call a dealer and ask about the dme update.

Good idea to think of those heater hoses, definitely seems like I’d be getting up close and personal chasing that harness around the engine.

IACV can’t put it into eml right?I mean my idle seems fine in general, but I guess it has occasionally slipped in to eml while sitting at a light at which point the idle gets super rough.. and spiky, I’ve assumed that’s the the result of eml not a possible cause. Kind of chicken and egg situation,





Sent from my iPad using Tapatalk

[pleiades]

.....But you’d think the mdk has some way of measuring the plate opening and bases its measurements off that and that play in the cable could be bad.....

The pots in your MDK are located internally on both sides of the throttle plate. I believe the DME compares their values with each other. The bowden cable is clutched onto a two-piece wheel on the side of the MDK; it turns the plate spindle open when you press the pedal and a spring wrapped round the inner half of the cable wheel turns the spindle back, closing the plate when you let off the pedal.

I don't understand how the pots could generate faults unless the MDK were worn out, or flawed right out of the factory. If you look on the bottom of your new MDK, there will be a small datestamp (last two digits of year circled by the clockface indicating month) in the metal body. If it's new, that date should be fairly recent, assuming BMW still has them being manufactured.

If you suspect the IACV, it certainly isn't that hard to pull it out and clean it.

[bjornk]

The pots in your MDK are located internally on both sides of the throttle plate. I believe the DME compares their values with each other. The bowden cable is clutched onto a two-piece wheel on the side of the MDK; it turns the plate spindle open when you press the pedal and a spring wrapped round the inner half of the cable wheel turns the spindle back, closing the plate when you let off the pedal.

I don't understand how the pots could generate faults unless the MDK were worn out, or flawed right out of the factory. If you look on the bottom of your new MDK, there will be a small datestamp (last two digits of year circled by the clockface indicating month) in the metal body. If it's new, that date should be fairly recent, assuming BMW still has them being manufactured.

If you suspect the IACV, it certainly isn't that hard to pull it out and clean it.

My “new” mdk was marked 14, the old one was marked 00....

I’ll look up cleaning iacv. Can’t hurt and I’ll be back in there anyway


Sent from my iPad using Tapatalk

[geargrinder]

No that voltage listed w an error is what was recorded with it when the error occurred. 12V sucks. It is bad.

MEASURE YOUR VOLTAGE as requested. This is important.

The dme on these in fact no has no idea what degree you push the pedal except for what it reads from the mdk ... So, no the cable seems highly improbable to be a cause.

But harness and grounds And low voltage issues could cause some funny business w readings...

[bjornk]

No that voltage listed w an error is what was recorded with it when the error occurred. 12V sucks. It is bad.

MEASURE YOUR VOLTAGE as requested. This is important.

The dme on these in fact no has no idea what degree you push the pedal except for what it reads from the mdk ... So, no the cable seems highly improbable to be a cause.

But harness and grounds And low voltage issues could cause some funny business w readings...

I’m measuring 12.6 on battery while sitting which seems a little low. As soon as it is running, there’s 14.0 going to the battery.

I have a very sketchy badly corroded engine ground. The big one on passenger side below the engine. I’ve got a new one on the way. I would be awesome. if it was a ground. I rebuilt a Honda once and couldn’t get it running until I find a little loose ground then all way good. .. maybe......the regular engine grounds look all tight. This scanner danner dude has me doubting my multi meter. The more I watch the more I think I need a scope.

I’m looking at scopes. Any of you guys have any experience with them? Input? Seems they could be of use...in this scenario. If it’s 95$ an hour for a diagnostic and a one week wait, I can almost buy a scope and take a week to learn how it works....wiring diagrams hurt my head....

https://www.scannerdanner.com/forum/...-a-budget.html


Sent from my iPad using Tapatalk

[bjornk]

I’m trying to understand these electrical concepts, so I am making a statement here which may be wrong and hoping someone can correct me if am am.

A while back, I tested my wiring harness for continuity thinking if there were a bad spot in the harness I would show no continuity. But that’s kind of a inaccurate test because a) a single strand of wire could show continuity and b) what happens under load may be different. I took measurements but they were of the resistance of the wire from one point to the next and could likely be similar even if there were a short.

So what I should likely do is check for short to ground with an ohm meter. I would do this by taking lose booth ends of the harness from the mdk to the dme. I would put my multimeter to my Highest setting, I would connect one end to a ground and the other to one end of each pin or wire , I would then look for a voltage drop? If I saw nothing, I would slowly drop settings.

Thus I would identify a short to ground within the wire?

I know I’ve looked at a mdk wiring diagram.but I’m diagram dyslexic. Where is the actual path from the potentiometers to ground? Is that just way too co plicated.? Do they share my big engine ground?I’m thinking that since the mdk is on the engine, that it’s grounded to some nearby ground to the engine. But if the engine itself has a ground that’s showing resistance, then that mdk which is operating likely in small amp gauging by the wire, takes a path of less resistance,, could be anywhere under a variety of conditions and it’s going somewhere else means one potentiometer measures slightly different than the other throwing the thing in to eml but not necessarily blowing fuses or creating other problems( maybe more Corosion)

So barring a fancy breakout box, I can simply check my wires for voltage drop. But the differences could be very small, so I need to have a reference?

This sounds crazy and hurts my head. Are we really all wavelengths?


Sent from my iPad using Tapatalk

[bjornk]

Does anyone know if it’s normal for the mdk to go back in to eml as it relearns it’s adaptations?could this simply be part of a process, and then I make it worse by clearing codes?

Like if I put a new mdk on it, and it starts acting stupid, and I erase my memory errors... do I need to let that memory sit and stay put so the dme can learn it’s adaptations or whatever? How many miles does it take for the process to be done ?




Sent from my iPad using Tapatalk

[geargrinder]

I’m measuring 12.6 on battery while sitting which seems a little low. As soon as it is running, there’s 14.0 going to the battery.

I have a very sketchy badly corroded engine ground. The big one on passenger side below the engine. I’ve got a new one on the way. I would be awesome. if it was a ground. I rebuilt a Honda once and couldn’t get it running until I find a little loose ground then all way good. .. maybe......the regular engine grounds look all tight. This scanner danner dude has me doubting my multi meter. The more I watch the more I think I need a scope.

I’m looking at scopes. Any of you guys have any experience with them? Input? Seems they could be of use...in this scenario. If it’s 95$ an hour for a diagnostic and a one week wait, I can almost buy a scope and take a week to learn how it works....wiring diagrams hurt my head....

https://www.scannerdanner.com/forum/...-a-budget.html

That looks like a cool website, I'll have to look at it more later. I would def not say a scope is super high on my list of required diagnosis tools - software like BMW Tools / INPA (for cars of this era) would be way way way up there before that. Seldom in most civilians cars are you into something where a 'scope is really going to be required to diag it, although there's certainly places where it can help on occasions. Still using the scope even in those cases tends to be a rare event. I have a proper scope and never lug it out to the garage (though I've been meaning to get one of the mini-scopes and/or a PC-based one for ages...)

~14V running is perfect (normally its like 13.x but ~14 - great). 12.6 fully charged is weak - you may need a battery.

That ground might be your problem. Good investigation.

I’m trying to understand these electrical concepts, so I am making a statement here which may be wrong and hoping someone can correct me if am am.

A while back, I tested my wiring harness for continuity thinking if there were a bad spot in the harness I would show no continuity. But that’s kind of a inaccurate test because a) a single strand of wire could show continuity and b) what happens under load may be different. I took measurements but they were of the resistance of the wire from one point to the next and could likely be similar even if there were a short.

So what I should likely do is check for short to ground with an ohm meter. I would do this by taking lose booth ends of the harness from the mdk to the dme. I would put my multimeter to my Highest setting, I would connect one end to a ground and the other to one end of each pin or wire , I would then look for a voltage drop? If I saw nothing, I would slowly drop settings.

Thus I would identify a short to ground within the wire?

I know I’ve looked at a mdk wiring diagram.but I’m diagram dyslexic. Where is the actual path from the potentiometers to ground? Is that just way too co plicated.? Do they share my big engine ground?I’m thinking that since the mdk is on the engine, that it’s grounded to some nearby ground to the engine. But if the engine itself has a ground that’s showing resistance, then that mdk which is operating likely in small amp gauging by the wire, takes a path of less resistance,, could be anywhere under a variety of conditions and it’s going somewhere else means one potentiometer measures slightly different than the other throwing the thing in to eml but not necessarily blowing fuses or creating other problems( maybe more Corosion)

So barring a fancy breakout box, I can simply check my wires for voltage drop. But the differences could be very small, so I need to have a reference?

This sounds crazy and hurts my head. Are we really all wavelengths?

Yes a simple continuity test is fairly inadequate unfortunately to find all the things that can be wrong. You rightly identify some of them (low current carrying capability so yes its 'connected' but it is impaired under load). Also you can have perfect continuity but you can have also shorts to ground and shorts to other wires. A wire that's shorted to another wire or ground will show perfect continuity but most likely be electrically useless. Therefore a simple standard wire check should be...
- first check for voltage on the wire (if its supposed to have voltage, does it, if it isn't supposed to have voltage, does it aka is it shorted to something with volts...)
- check end to end resistance - ohms... (hopefully you get a modern DMM with auto-ranging so you don't have to play w/ the range settings!)
- then check resistance to any other wires that its bundled up with

You still can have situations where even the ohms measured are super low but the wire does have problems carrying current...

I don't have the MDK wiring diagram handy so I can't answer your question there for sure... some sensors have self-contained wiring (I believe MAF is an example) where the return wire is hard-wired, others ground through the block.

Does anyone know if it’s normal for the mdk to go back in to eml as it relearns it’s adaptations?could this simply be part of a process, and then I make it worse by clearing codes?

Like if I put a new mdk on it, and it starts acting stupid, and I erase my memory errors... do I need to let that memory sit and stay put so the dme can learn it’s adaptations or whatever? How many miles does it take for the process to be done ?

Nope. Normally if a part like that fails a 'diagnosis', any adaptation either is 'frozen', or, its measurement calculations get shunted off to some 'limp' tables one way or another. Clearing codes may or may not clear adaptations and reset it all... depends on the ECU/DME. Dunno without googling for MS42. I don't know the MDK myself, some cabled-throttle-bodies require you to cycle the pedal along with some key-cycling routine (something like "turn ignition to ON but not start... wait 12.35 seconds, then press the pedal to the floor 3 times while singing 'I'm a little teapot'" etc.) ... I don't recall hearing that for M52TU but I've not owned one personally and not had to do a TB on one when I've worked on 'em.

You might be interested tho' in how BMW errors and SES lights get triggered tho'... Each error has various parameters defined for it. Whether it can trigger an SES at all or not, etc. The DME keeps track of both how many times the error has happened, as well as counting the number of starts that have happened since the last time the error was seen. So each error can be different in terms of how many occurrences before the light is triggered. For some errors this number is clearly "1" but for others it has to hit some threshold like "3 times" or whatever. In addition errors can differ in "how many cycles (typically starts/drives) without this error appearing before the error is cleared again". So for an O2 error maybe if the error is triggered once, then you'd need to drive the car 40 times without the error appearing again for it to be cleared by the DME. For something else minor - maybe tank evap or something (the famous gas cap error) - it could be if you drive the car 10 times without the DME seeing it again, the error will be cleared and the light and counter reset. This is all why I always rant about a BMW compatible scanner being better than a generic OBD because you can see these things - count and # of starts since it was last seen and volts and engine temp when occurred etc.

[effduration]

Does anyone know if itÂ’s normal for the mdk to go back in to eml as it relearns itÂ’s adaptations?could this simply be part of a process, and then I make it worse by clearing codes?

Like if I put a new mdk on it, and it starts acting stupid, and I erase my memory errors... do I need to let that memory sit and stay put so the dme can learn itÂ’s adaptations or whatever? How many miles does it take for the process to be done ?

No, it is not normal to go immediately into EML when resetting adaptations. I have reset several times, and it will not go in to EML unless there is a persistent fault.

To be clear, here is the reset process I use on my '99 528i - It includes BOTH an INPA reset and manual setting of throttle range.

(found in post #4 in https://www.bimmerforums.com/forum/s...-Throttle-Body)

Followed procedure:

1999-2000 M52TU 2.8L motors / Siemens MS 42.0:

Procedure:

1. install part.
2. ignition off 30 seconds.
3. ignition on
4. clear codes via INPA and reset DME adaptions via INPA
5. press throttle to floor (WOT) 4 times.
6. ignition off.
7. Ignition on, start car.

Good to go.

To get to adaptions and reset them in ediabas/INPA...

F5 - E39
"engine"
"MS 42.0 ..."
F4 read codes
F2 clear codes
F7 reset adaption
F10 exit

[bjornk]

No, it is not normal to go immediately into EML when resetting adaptations. I have reset several times, and it will not go in to EML unless there is a persistent fault.

To be clear, here is the reset process I use on my '99 528i - It includes BOTH an INPA reset and manual setting of throttle range.

(found in post #4 in https://www.bimmerforums.com/forum/s...-Throttle-Body)

Followed procedure:

1999-2000 M52TU 2.8L motors / Siemens MS 42.0:

Procedure:

1. install part.
2. ignition off 30 seconds.
3. ignition on
4. clear codes via INPA and reset DME adaptions via INPA
5. press throttle to floor (WOT) 4 times.
6. ignition off.
7. Ignition on, start car.

Good to go.

To get to adaptions and reset them in ediabas/INPA...

F5 - E39
"engine"
"MS 42.0 ..."
F4 read codes
F2 clear codes
F7 reset adaption
F10 exit

Thanks I followed this procedure at install and it ran well a few weeks. I read somewhere it can take several hundred miles of driving for an ecu to adapt to new components but it was a general idea not bmw specific. I’m gonna look for short to ground next time I get a chance after i fix the major ground. I’m gonna measure the rounds resistance just for kicks. I need to do some dmm practice exercises anyway...


Sent from my iPad using Tapatalk

[bjornk]

That looks like a cool website, I'll have to look at it more later. I would def not say a scope is super high on my list of required diagnosis tools - software like BMW Tools / INPA (for cars of this era) would be way way way up there before that. Seldom in most civilians cars are you into something where a 'scope is really going to be required to diag it, although there's certainly places where it can help on occasions. Still using the scope even in those cases tends to be a rare event. I have a proper scope and never lug it out to the garage (though I've been meaning to get one of the mini-scopes and/or a PC-based one for ages...)

~14V running is perfect (normally its like 13.x but ~14 - great). 12.6 fully charged is weak - you may need a battery.

That ground might be your problem. Good investigation.



Yes a simple continuity test is fairly inadequate unfortunately to find all the things that can be wrong. You rightly identify some of them (low current carrying capability so yes its 'connected' but it is impaired under load). Also you can have perfect continuity but you can have also shorts to ground and shorts to other wires. A wire that's shorted to another wire or ground will show perfect continuity but most likely be electrically useless. Therefore a simple standard wire check should be...
- first check for voltage on the wire (if its supposed to have voltage, does it, if it isn't supposed to have voltage, does it aka is it shorted to something with volts...)
- check end to end resistance - ohms... (hopefully you get a modern DMM with auto-ranging so you don't have to play w/ the range settings!)
- then check resistance to any other wires that its bundled up with

You still can have situations where even the ohms measured are super low but the wire does have problems carrying current...

I don't have the MDK wiring diagram handy so I can't answer your question there for sure... some sensors have self-contained wiring (I believe MAF is an example) where the return wire is hard-wired, others ground through the block.



Nope. Normally if a part like that fails a 'diagnosis', any adaptation either is 'frozen', or, its measurement calculations get shunted off to some 'limp' tables one way or another. Clearing codes may or may not clear adaptations and reset it all... depends on the ECU/DME. Dunno without googling for MS42. I don't know the MDK myself, some cabled-throttle-bodies require you to cycle the pedal along with some key-cycling routine (something like "turn ignition to ON but not start... wait 12.35 seconds, then press the pedal to the floor 3 times while singing 'I'm a little teapot'" etc.) ... I don't recall hearing that for M52TU but I've not owned one personally and not had to do a TB on one when I've worked on 'em.

You might be interested tho' in how BMW errors and SES lights get triggered tho'... Each error has various parameters defined for it. Whether it can trigger an SES at all or not, etc. The DME keeps track of both how many times the error has happened, as well as counting the number of starts that have happened since the last time the error was seen. So each error can be different in terms of how many occurrences before the light is triggered. For some errors this number is clearly "1" but for others it has to hit some threshold like "3 times" or whatever. In addition errors can differ in "how many cycles (typically starts/drives) without this error appearing before the error is cleared again". So for an O2 error maybe if the error is triggered once, then you'd need to drive the car 40 times without the error appearing again for it to be cleared by the DME. For something else minor - maybe tank evap or something (the famous gas cap error) - it could be if you drive the car 10 times without the DME seeing it again, the error will be cleared and the light and counter reset. This is all why I always rant about a BMW compatible scanner being better than a generic OBD because you can see these things - count and # of starts since it was last seen and volts and engine temp when occurred etc.

I have a scanner and Inpa and some things that came on that software suite. I haven’t used any if it for anything other than this mdk yet. Doesn’t that inpa suite kind of trump the scanner ?i guess you use the scanner to connect to the link under the hood and for other systems like tranny and brakes? Haven’t gotten that far yet.

I’m waiting in some better back pinning probes and piercing probes to arrive before I do too much poking around. I don’t want to disturb too much wiring it does seem brittle .


I’m thinking of buying that scanner danner text as it’s pretty cheap and looks like a fast why to get a grasp on basic principles. Some of it gets over my head but the pacing seems fast,

I was reading this and noticing that sometimes when I let off the gas and decelerate, eml light kicks on. Guess I’ll check that out too


Sent from my iPad using Tapatalk

[effduration]

I was reading this and noticing that sometimes when I let off the gas and decelerate, eml light kicks on. Guess I’ll check that out too


Sent from my iPad using Tapatalk

Huh... I haven't seen much about the Idle Speed control valve causing an MDK-related EML mode. I presume (though not certain) that we would get codes related to the idle speed valve if it was failing. One more good reason to clean it periodically.

[bjornk]

Will the scanner connected in engine pull up different data then the OBD two connector connected to inpa relating to an MDK problem?

Will the idle air control valve have it own codes and pop up as an error in both places?

Inpa is basically telling me what the computer is thinking but the computer may be getting bad inputs... can it distinguish between an mdk fault and a iacv fault? Clearly they kind of work together —- have a relationship...


Sent from my iPad using Tapatalk

[bjornk]

Huh... I haven't seen much about the Idle Speed control valve causing an MDK-related EML mode. I presume (though not certain) that we would get codes related to the idle speed valve if it was failing. One more good reason to clean it periodically.

If you read the entirety of that link ....it’s explains the relationship between the iacv and mdk... I’m still getting my head around it,,,, but it does say the iacv ca trigger eml and controls engine deceleration which is something that I’ve noticed triggers eml.

I think. Cleaning Iacv is next step for sure...



Sent from my iPad using Tapatalk

[geargrinder]

Doesn’t that inpa suite kind of trump the scanner ?

Overthinking it - when I said "BMW scanner" I meant "any scanning software or device that is properly capable of collecting the BMW proprietary info", and by that definition INPA would absolutely be qualifying.

Will the scanner connected in engine pull up different data then the OBD two connector connected to inpa relating to an MDK problem?

Will the idle air control valve have it own codes and pop up as an error in both places?

Inpa is basically telling me what the computer is thinking but the computer may be getting bad inputs... can it distinguish between an mdk fault and a iacv fault? Clearly they kind of work together —- have a relationship...

1. No. Engine data coming from either connector is using the same wire to the same pin on the DME (ECU). They aren't 2 different places, its just that on an older car like yours the OBD2 connector only has the DME and EGS (engine and trans computers) while the other car modules are all only wired to the engine bay connector. In fact a common mod is to bridge those pins in the engine bay so that anything plugged into the OBD2 connector can talk to all the modules as well.

2. Yes ICV will have its own codes. There aren't 2 places as just answered.

3. Yes it will distinguish between faults. It does so by watching for and testing for specific response to its output, i.e. "trigger ony ICV, does engine respond as expected while not changing anything on the MDK"... Yes it can get bad inputs. That is what "Read Status" is for in INPA - go into those screens and you can read the inputs that the DME is getting. If one of them looks out of whack then you have to determine is the measurement correct because something is wrong? Or is the measurement incorrect and its the sensor/input that's wrong.

[bjornk]

Overthinking it - when I said "BMW scanner" I meant "any scanning software or device that is properly capable of collecting the BMW proprietary info", and by that definition INPA would absolutely be qualifying.



1. No. Engine data coming from either connector is using the same wire to the same pin on the DME (ECU). They aren't 2 different places, its just that on an older car like yours the OBD2 connector only has the DME and EGS (engine and trans computers) while the other car modules are all only wired to the engine bay connector. In fact a common mod is to bridge those pins in the engine bay so that anything plugged into the OBD2 connector can talk to all the modules as well.

2. Yes ICV will have its own codes. There aren't 2 places as just answered.

3. Yes it will distinguish between faults. It does so by watching for and testing for specific response to its output, i.e. "trigger ony ICV, does engine respond as expected while not changing anything on the MDK"... Yes it can get bad inputs. That is what "Read Status" is for in INPA - go into those screens and you can read the inputs that the DME is getting. If one of them looks out of whack then you have to determine is the measurement correct because something is wrong? Or is the measurement incorrect and its the sensor/input that's wrong.

10-4 and thanks makes sense...,..


Sent from my iPhone using Tapatalk

[bjornk]

So i'm taking apart this wonderful vehicle with the intention of checking short to ground on my mdk wires(damn those things are the smallest gauge. while im in there, I thought id look at my iacv. Im kicking myself in ass not paying closer attention to "buzzing sounds" and reaction to idle when air is kicked on but....

I went ahead an looked at the resistance, Bentley says I should see 20 ohms plus/minus 2 if I jump pin 1-pin2, 20 if I jump 2 and 3 , and 40 if I jump pins 1 and 3.

Im getting, 13.6,12.3 and 25.6 respectively. Which leaves me wondering, could the iacv be bad? I cleaned the thing which im guessing means flood with señsor safe carb cleaner and compressed air, if anything the measurements are sightly lower cold and clean....


im about to jump in to my mdk wiring( I dielectric greased all those connectors a few weeks and I can still see residue...The connector seemed snug, I have another connector, and a new one on the way, but I think redoing the terminals would be at least as important as the connector at this point...

[bjornk]

isn't lower resistance better and the reading in Bentley are the max readings?

[bjornk]

Ok

I disconnected mdk from dme. Found a known good ground. I’ve got a cheap ass dmm —- not sure if it’s auto ranging.. gonna break down and read the manual .

I started at 2000k ohm and dropped to 200 ohm. For a while I was getting nothing so I made up some mock tests with wires various ground etc to make sure I could test resistance v infinite ohms. Once confidence in ground, meter and self, I tested mdk leads w dme disconnected. On all pins all I ever got was 1 which I am reading to mean infinite ohms. Does this conclude wire harness test or could a short appear under load ?

I have also tested continuity of wires and all is good

If my iacv is good and my harness and mdk is good, I’m left scratching head,

I’ve been reading in a few spots about the throttle cable procedure and the importance of it being adjusted and even a dis procedure for that.supposedldy to ensure proper base plate angle / position. Bentley says the same. Others on forum have suggested not it, that throttle cable is not enough to put car in eml.

Also read some junk somewhere about possibility of brake sensor triggering eml. Like if you step on gas and brake then eml can be triggered, I know my brakes are a little fubar. I have : trifecta lights on, Bad brake bulb ( autozone brand new one on way ), Yellow brake light etc . But no codes being picked up on Inpa from connector under dash except those related to mdk ( they could be found on 2o pin scanner right.) Visual inspection told me lots of sensors disconnected on brakes . I figured I d get car running good before getting too involved in abs and traction control . Brakes work great imho but yes my dash looks like Christmas, could this be an issue?

Crazy thing is, car may start up in am and be just fine.....




Sent from my iPad using Tapatalk

[bjornk]

Overthinking it - when I said "BMW scanner" I meant "any scanning software or device that is properly capable of collecting the BMW proprietary info", and by that definition INPA would absolutely be qualifying.



1. No. Engine data coming from either connector is using the same wire to the same pin on the DME (ECU). They aren't 2 different places, its just that on an older car like yours the OBD2 connector only has the DME and EGS (engine and trans computers) while the other car modules are all only wired to the engine bay connector. In fact a common mod is to bridge those pins in the engine bay so that anything plugged into the OBD2 connector can talk to all the modules as well.

2. Yes ICV will have its own codes. There aren't 2 places as just answered.

3. Yes it will distinguish between faults. It does so by watching for and testing for specific response to its output, i.e. "trigger ony ICV, does engine respond as expected while not changing anything on the MDK"... Yes it can get bad inputs. That is what "Read Status" is for in INPA - go into those screens and you can read the inputs that the DME is getting. If one of them looks out of whack then you have to determine is the measurement correct because something is wrong? Or is the measurement incorrect and its the sensor/input that's wrong.

I’m gonna look in to that bridging task. So are you saying if bridged, Inpa under dash can also check on abs? But presently it won’t? I vaguely remembered some talk about cable bridging when buying a cable.... didn’t know you had to bridge the actual car too,,,,


Sent from my iPad using Tapatalk

[bjornk]

Huh... I haven't seen much about the Idle Speed control valve causing an MDK-related EML mode. I presume (though not certain) that we would get codes related to the idle speed valve if it was failing. One more good reason to clean it periodically.

My concern and hopefully it’s not imagined, is that throughout this process, I have tried to be acutely aware of anything I am doing : driving patterns etc, that seem to throw the car into eml. In general, noting I tend to do seems to matter except one thing: I have noticed fairly clearly, that I pf I am driving down the highway doing about 75 , and kind of ease off the gas slowly, the eml light will indeed come on. As I ease back on the gas, the light then goes out. Which from what I have read, doesn’t even sound possible. But I’ve done it dozens of times under that same conditions, I kind of feel like I’m talking about ghosts because it doesn’t makes no sense to me. But maybe it makes sense somehow..

From what I’m reading, the iacv can trigger eml. And works together with the mdk primarily under deceleration/ acceleration, that’s what has me thinking maybe the iacv could have something t’ do with it,


Sent from my iPad using Tapatalk