[mabbadgerbimmer]

Originally Posted by cntrvrsy

I hate to say it and I hope I am wrong but since you mentioned 3/4 shifts, it leads me to think it's the either the compression springs or synchros. Check this thread.

Springs were replaced as well.

[mabbadgerbimmer]

Originally Posted by SteveEvans

I suggest you get a diagnostic cable and run INPA to determine the fault codes and post them. They may require some interpretation as the temperature issue manifests itself as excessive motor on time.

Steve

Went to the shop on the way home. the file was shipped to the software diagnostic engineers in the U.K. to see if they could figure it out.

[CJG M3]

I have been on multiple forums researching this over the past few weeks and then something dawned on me...

when i went in for my inspection II around 60k miles, what i thought was a bad O2 sensor turned out to be a fault for a bad catalytic converter. Since the dealer didnt want to eat the bill for replacing my CAT under warranty, they insisted on a software "fix" and reprogrammed my emissions computer.

i have to assume the reprogramming of the emissions computer would change how hard the CAT was working - would it be safe to assume this would change the amount of heat being generated at the headers (more heat near the SMG Hydraulic pump temp sensor)?

i first saw my SMG temperature related fault right after my inspection II and i always assumed the dealer F'ed something up. i am going to do the resistor install this weekend. below is a post i had on e46F last week which details my symptoms:

---------------------------
Being from Orlando, FL, it sounds like my SMG is experiencing the prototypical summertime blues. it sounds like i have identical symptoms with my SMG dropping to Neutral during city traffic on hot days (and with the AC generating additional heat)...

Anyone who has performed this mod already: i just want to know how your SMG is holding up after the resistor mod - it looks like most of the traffic was from about a year ago.

i am comfortable working with electronics, but i am still a little apprehensive modifying wiring on such an expensive and intricate car. please let me know how this is holding up and if you all are still insisting that this is the saving grace a year later, i'll give it a shot. Thanks in advance


Car: 04 M3 vert SMG II

i am experiencing the same intermittent fault that was addressed above

this has happened about 4 times in the last 1500 miles. the SMG cog light would come on and the smg would eventually shift to neutral for about 3-5 sec before popping back in gear. i just completed a 200 mile trip non-stop mostly on the highway and did not have any issues until i was in city traffic (shifting frequently). Outside temp varies from about 80-100F

I had the computer scanned at the BMW dealer and saw the following codes which may be related:

0038 SMG: On period, hydraulic unit (OBD#P1717)
0069 DME: signal, engine, temperature, implausible
00B1 DME: MIL lights up as result of request by SMG
00E6 DME: Electric throttle: setpoint/ actual value deviation

other notes: i have replaced the salmon relay. I have ~64K miles. i recently received the inspection II service and i saw this light go on shortly thereafter. When the dealer scanned the computer they did not provide me any worthwhile insight to the problem since i'm sure their best interest would be to sell me a whole new hydraulic unit...

this is a very intermittent issue. i truly love this car, but if i am looking at a $3-4K tranny job i may just trade it in or sell it as-is.

Thanks for all replies.

[ghok]

From the other message, one of the option is to change the temp sensor. I changed the temp sensor but still the SMG cog light and Engine SES light "ON" but didn't drop to netrual.

I think I will have to go for the resistor mod and see what's next.

[Nyght]

Just did the resistor mod finally.. Been busy because of work.

The delay on start up is no longer there which is good.
I must say this is a good fix.. I installed a 3.3k Ohm resistor part 10W233.. Sooo far so good.

[CJG M3]

i agree with SteveEvans on the resistor wattage comment. this is a temp sensor wire, not a power wire. there isnt any significant current going through it and it is going to be low dc voltage, so a regular 1/2W resistor should do the trick. anyways, i used a 1.5k 10W resistor since i had already bought it, and i did a continuity test from the white-wire of the connector behind the driver side front wheel up to pin 21 of the connector in the electronics bay and it buzzed out, so i spliced my resistor in up top to prevent any water from getting to it. its been one week and no light so far. this is a better place to do the mod in comparison to under the car where it is exposed to the elements

[GreyMatter]

Looks like the original link posted to the temperature sensor specs is broken now. I managed to find the new link to the pdf here (will also attach to this message).

http://www.bosch-motorsport.de/en-US...2782569739.pdf

Based on this it seems that a 1K ohm resistor should be more than enough. The error seems to come on when the temperature readers 70C or above, at which point the resistance should be 436 ohms according to the chart. Adding 1K to it would be 1436 ohms which would give readings around 35C.

I was thinking that I might try installing a variable resistor here so that I can easily revert back to stock if necessary. Also then I could "fine tune" the bias so that the sensor actually still goes off in the event that it really is reading too hot.

The one piece of info that I'm missing is the max temperature range that the fluid should reach. Is 70C really too hot? According to the specs for Pentosin CHF11S it should be able to handle 70C no problem, but that doesn't mean the internals of the pump can handle that. Does anyone know what the max SMG fluid temperature is?

[daytonaviolet]

what would the motive to return the resistance back to stock if necessary? adding a variable resistor sounds like i cool idea if you could measure its resistance on the fly while in the circuit. perhaps switch to toggle in and out the 1k resistor.

[GreyMatter]

Originally Posted by daytonaviolet

what would the motive to return the resistance back to stock if necessary? adding a variable resistor sounds like i cool idea if you could measure its resistance on the fly while in the circuit. perhaps switch to toggle in and out the 1k resistor.

I was thinking that this would only be a temporary solution. If it does solve my issue then eventually I want to swap out the temp sensor and then return the wiring to stock. I could always desolder the wire and put it back but I figure it would be easier this way, plus I get the benefit of being able to tune my resistance.

However you are right, once it's in the circuit I unfortunately can't measure the resistance easily so I'm not sure how useful it would actually be.

[iceberq]

Originally Posted by SteveEvans

Why are folks (a) using a huge 10W resistor, and (b) why are they fitting it in an exposed location where corrosion will cause a failure?

The resistor will have 12V across it, perhaps 14V, to be on the safe side, so even without the temp sensor in series, the power dissipation of a 1K resistor will be 14*14/1000=0.2W. Therefore you can use a very small 1/3W to 1/2W resistor.



The best place to fit this is in the electronics compartment next to the SMG control unit. The white wire mentioned in this thread goes to pin 21 of X53004.



Such a small resistor is ideally fitted there by cutting the white wire, sliding some head shrink sleeving up the wire, soldering the resistor in place, and then sliding the sleeving over the resistor and joints and shrinking it.

This will be a much, much neater fix and totally weather resistant.

Has anybody done the resistor mod in this ECU box location? If so, is it working fine too? I would love to hear some feedback as I will be getting this done soon.