[MYKIDZM3]

JP, did VAC take the pump out, if so do you want them to send it to me so I can have my guys look at it? We deal with pumps all the time and have a lot of expertise in the area. We might have someone who could rebuilt it!

[Ghozt]

Here's what I'll do to really add to this post, and make it as informative as possible, and I will double post this in the SMG troubleshooting thread... I'm going to list everything I tried, with the help of Armand at VAC motorsports. My car has been there for more than 2 months now. I wanted to be one of the few who found a fix outside of full pump replacement for those of us who have similar issues. While I made the attempt to save myself some money on a full pump, I also spent a lot of money on diagnostics. So here we go....


My issues: (Short of the exact codes because I forget all of them)
Code 056 (I believe)
Car will randomly pop out of gear during any random time during a drive... I could driving gingerly for 10 minutes, or very spirited for 15, and it could happen. It doesn't necessarily happen on a shift. It doesn't necessarily happen during a stand-still. It can happen at anytime. It is more prominent in warm weather. Only once (in a very hot climate) was I unable to get the car back into '1' (out of neutral) after parking and turning off the car. I don't drive my car much in the winter, but as I mentioned, the car has done this in cold weather as well.

The result: The car will literally drop all gears and go into 'N'... After coasting or sitting for anywhere from 5-30 seconds, the car will re-engage gears and driving will continue. This can happen numerous times while driving. The car never disengages all gears to the point where I'm "stranded." As I mentioned... Only once, in a hot climate, was I unable to re-engage gears upon startup after experiencing problems during the drive. I get the Orange COG light. Upon turning off the car, and then restart, if the COG light remains lit, I get an orange SES light. The SES light will remain until the car has been driven XX amount of miles without a similar fault...

The attempted fixes and diagnostics: Codes were pulled first, aside from 056, I forget the other 1 or 2...
1. Salmon Relay Replaced - No change
2. Topped off fluid - No change
3. Updated to latest SMG software - No change
4. Replaced the Transmission fluid - No change
5. Replaced the SMG computer, reprogrammed - No change
6. Armand monitored pressure while driving the car... Noticed pressure was dropping below 50 bar (this is the number if I'm remembering correctly), a point at which the SMG pressure sensor would typically tell the pump to engage and "refill" pressure. Armand suggested a unit replacement.
7. Before succumbing to the full unit replacement, I asked him to try one more thing... Disabling the HCU temp sensor. Armand did so, and drove the car over the weekend. Although he notice a little improvement, he still had issues with pressure as he monitored the car.
8. Armand was able to install a bypass switch. When he noticed the car had dropped below the necessary pressure, he could hit the switch and the pump would engage immediately, bypassing the "sleeping" pressure sensor. The use of this switch basically alleviated my issues.


After all of the above, Armand is diagnosing my issue as a faulty pressure sensor within the Hydraulic unit. A part that does not seem readily available to us, but I spoke with a BMW tech, and it is something that is replaceable. It's just another one of those scenarios where you can't justify spending $500+ dollars to replace a little part when you may run into more failures shortly down the road... I am opting to replace the entire pump at this point....


I tried guys, I really did. I like to think of myself as a determined person. I absolutely refused to just throw my hands up and replace this entire unit. I wanted to at least try to accomplish something that would help myself and others save money in the future when experiencing problems with these units. It's a shame we're seeing more and more of these threads day by day, even coming from folks with relatively low mileage M3's. I can only hope that with all of this information that is gathered, a suit is really brought against BMW for these cheap parts that continue to fail on these very expensive units, requiring a full unit replacement, rather than just the cheap part...

I hope this has been informative. Please feel free to ask any questions...



JP

[cntrvrsy]

Ghost, What did Armand use to monitor the pressure in the HCU? The fact that he was able to use a bypass switch would make me want to try another relay. It's a possibility the one you got was faulty or even been dropped by accident (never know). Perhaps something is up with your Accumulator or non-return valve? Taken from the SMG training manual that's been posted online...

Pressure Accumulator and Non-Return Valve:
The Pressure Accumulator in the hydraulic unit ensures that pressure generated by the pump is stored in the system for a certain amount of time. The accumulator is divided into two chambers by a piston, with nitrogen at 39 bar filling one chamber and fluid delivered by the hydraulic pump filling the other chamber. The volume of the accumulator is 150ccm. The Non-Return Valve at the pump outlet prevents the hydraulic oil pressure from dropping when the pump is not running.

Pressure Relief Valve:
The Pressure Relief Valve opens if the hydraulic oil pressure reaches 100 bar, creating a circuit between the suction and pressure sides of the pump, thus preventing further increase in pressure.

Hydraulic Oil Pressure Sensor:
Mounted on the hydraulic unit the Oil Pressure Sensor informs the SMG II control unit of the current hydraulic pressure. At ) bar pressure a voltage signal of .5 volts is sent to the SMG II with the voltage increasing linearly to 4,58 volts at 100 bar of pressure. Failure of the pressure sensor whether shorted or open will cause results in the hydraulic unit being activated for a fixed amount of time at predetermined intervals and after each gearshift to maintain system pressure.

Good info although I wish it was more technical with schematics even.

[Ghozt]

Originally Posted by cntrvrsy

Ghost, What did Armand use to monitor the pressure in the HCU? The fact that he was able to use a bypass switch would make me want to try another relay. It's a possibility the one you got was faulty or even been dropped by accident (never know). Perhaps something is up with your Accumulator or non-return valve? Taken from the SMG training manual that's been posted online...



Good info although I wish it was more technical with schematics even.

The codes seemed to rule out the accumulator. I even purchased a brand new accumulator because I really thought that was going to be it.

I'm not sure exactly which tool Armand was using, but you know VAC, they have everything...

We also played with the relay quite a bit... I would have loved for it to be a relay issue, but I really don't think that was it...

[Ghozt]

Update.....

Well, I want to let you guys know that I tried, I really did... With the help of Armand at VAC we went through all the possibilities, and I spent many dollars on labor and diagnostics, trying the HCU temp sensor, trying the accumulator, trying the HCU ECU, etc, etc, and none of it worked... It was really a bummer after going through all that, and having to settle with the complete replacement of the unit. I am happy to report that my car is running fine now.

The good news... With all of that time spent on diagnostics and really attempting any fix besides a full unit replacement, we found some important things out.

1. The HCU temp sensor absolutely is prone to failure over a long period of time. Long can be loosely defined as well. It seems like this sensor could begin to fail much earlier on some units and last longer on others (no big surprise there). On my unit, cutting the temp sensor wire actually did help a little, but did not rid my unit of all issues..
2. There is a pressure sensor on the unit that is readily accessible. This seemed to be the center of my units' failure. It seems like it could certainly be a replaceable part. My unit was losing pressure, and it was dropping to far below the needed psi before the unit would kick in and pump it back up. This was due to the failing sensor not recognizing how much pressure was left in the unit before telling it to fire up...

Just one more thing to add to the list of potential smaller fixes for these units... I urge anyone who is having similar SMG issues to feel free to ask me any questions about my experience. I wanted nothing more than to walk away from this with a "cheaper" fix replacing a cheaper part that is on the unit, rather than the whole unit itself. The problem seems to be is that these sensors are not readily available, as they are not BMW parts, and it will require a little bit of time and research to really get this fix right.

In my honest opinion, BMW really dropped the ball on this unit. It's nice to see some folks with 125k miles on their original pump, still shifting strong as ever, while on the other side, folks with merely 50k miles or less are experiencing issues... Believe it or not, I've never used launch control, I'm too big of a pussy to drive this car aggressively on any sort of regular basis, and I all around baby my car. At 75k miles, I wouldn't have expected a failure, and I wouldn't have been expecting to spend $3k on replacing the entire unit...

I can only hope that others continue to progress and push the more inexpensive fixes, cause they are out there. They are slowly popping up because more and more people are seeing failure. It's sad to see daily threads about SMG issues. I would love to see folks really bring it up with BMWNA. Cars that have just gone past warranty and have to replace the entire unit is horrendous.

Moral of the story, buy a 6MT...

[MPOWERD]

How about this... Could you buy the new pump, but swap put the pressure sensor from it onto you old pump and determine once and for all if its just this sensor?

You might get another 75K out of it, and be able to sell the new pump to someone else that hasnt suffered a pressure switch failure on theirs and they could swap the pressure sensor form their old pump to make use of your pump...

[Mobbin_PCH]

Ghost I'm sorry you had to finally bite the bullet on this one, it's a disappointing feeling to say the least. I finally had to give in last month and replace the whole unit too. Ralph at Terry Sayther Automotive here spent 2+ weeks driving and running diagnostics on my car (and simultaneously comparing results to another M with the same problems) but could not come up with a viable alternative. He and I had turned this into a personal quest and it sounds that you did too.

It's sad that BMW chose to cut corners with the SMG units and I agree that something should be done about it. Until then, the knowledge that you (and others) have provided this forum will hopefully lead to a cheaper solution to this problem.

[RedM3/4]

This thread is really informative. I've disassembled one pump and found it to be in immaculate condition, and the owner later confirmed a replacement pump did NOT fix the problem. In the case above, it does appear there was a pressure problem. It is unknown if it is the pressure sensor or the pump for this reason: we do not know what interlocks in the hcu exist to interpret the pressure sensor versus turning the pump back on. That is, there may be some min/max timeor intermediate pressure that the HCU uses to turn on the pump, regardless or pressure sensor reading. This would seem odd, but I've seen MANY servo systems with more sensors and interlocks than are practical, or have sub-optimal thresholds to under report, or over-report, errors. E36 lights (brake, tail, many) come to mind, coming on when things are working 99% of the time, as but one example.
What we're missing here is a temperature and pressure profile, or baseline, of a system that is operating normally. I think this would answer most questions on the wellness or over/under reporting of SMG errors that can cause problems when the pump is well OR in trouble.
One thing I have wondered about that this particular saga solves, is whether pressure loss in the actuators (cylinders moving the clutch and shifter forks) is losing pressure. In this case or any case, where a new pump unit solves the problem, that is clearly not a factor.
Thanks much ghost, a lot of learning here. If you'd like to send me your pump, I can tell you if the pump parts look ok and (I think - have yet to use my 1500 psi gage on the unit I have) probably tell you if the pump is ok. I have no reasonable way to test the accumulator, as it is a nitrogen-charged, apparently closed (welded) unit. If anyone knows more/differently about the accumulator, please PM me.

ADS

[johnm_oz]

From SMG training manual:

Hydraulic Oil Pressure Sensor
Mounted on the hydraulic unit the Oil Pressure Sensor informs the SMG II control unit of the
current hydraulic pressure. At 0 bar pressure a voltage signal of .5 volts is sent to the SMG
II with the voltage increasing linearly to 4.58 volts at 100 bar of pressure.
Failure of the pressure sensor whether shorted or open will result in the hydraulic unit being
activated for a fixed amount of time at predetermined intervals and after each gearshift to
maintain system pressure.

Which means that a quick fix for a sensor that is telling lies- as the one in this instance seems to be- would be to completely disconnect it and, on the basis that an open circuit pair may introduce noise into the system, short the connector. No need for a manual switch.

Cheers
JM

[IHatePosers]

sux u had to go through this.

this is a horror story that will def scare me away from an smg.