2.8 Roadster possible intake leak, symptoms?

**JKuhn** · 10-29-2017, 07:02 AM

I just noticed that the intake pipe on my 2.8 Z3 roadster is full of cracks. Since I can't buy one immediately, I decided to put some RTV on the cracks that I saw, however I'm aware that I probably didn't catch them all. Now my question is, what symptoms can I expect from such a leak (between the MAF and throttle body, no warning light)? I'm wondering if it might be the cause of the problems I've had with the car (intermittant RPM drop when I start to open the throttle, no power below 4k RPM, surging).

**MIKYZZ4** · 10-29-2017, 07:30 AM

Rubber intake boot cracks are a rather common source of intake leaks.
Replace boot and any associated cracked vacuum hoses.
Your symptoms are normal for an intake leak causing lean fuel mixtures.

**JKuhn** · 10-29-2017, 07:47 AM

Well, that's good to hear. I've been suspecting the VANOS all this time.

**MIKYZZ4** · 10-29-2017, 08:06 AM

A scan with a BMW specific scan tool can help identify if error codes are present.
A replacement of worn VANOS seals can improve low end torque.

**JKuhn** · 10-29-2017, 08:16 AM

I think I'll just go ahead and replace the boot(s) for now, I won't be suprised if it helps quite a bit. Maybe it will even eliminate the problems.

Besides, I still need to do a couple other urgent things, so the VANOS will have to wait anyway (assuming it's faulty).

**JKuhn** · 10-30-2017, 08:57 AM

UPDATE: I pulled the boots out, and patched them. I'm just giving the RTV a bit more time to cure and then I'll test it this evening. The lower boot had a very significant split on the bottom, and it's already been patched before by a previous owner (but the patch evidently didn't last). Also, I think I should ensure my GoPro is ready, so I can get the power difference on video (I already have one of how it pulls at the moment).

Regarding a replacement, I'll place an order tomorrow if I can, otherwise later this week. I can get both from BMW for R500 (about $35), on order.

**MIKYZZ4** · 10-30-2017, 09:07 AM

Truly doubt patches will work, as they will probably be ingested into intake,
I would place order for replacement ASAP!

**JKuhn** · 10-30-2017, 09:47 AM

As I said, I'm planning to order the replacement tomorrow.

And the patchwork is large enough (relative to the leaks) to not be ingested that easily.

**JKuhn** · 10-30-2017, 12:19 PM

UPDATE: I drove the car a bit. At first it did feel slightly better, but not much. Then after a couple stops it opened up a lot. I still think it can do better, but at least it's smooth, and pulls well.

Now I'm wondering, since the car previously also intermittantly got its power back. Should I expect an immediate return after fixing the leak, or must the ECU first get some time to adjust?

**LannVouivre** · 10-30-2017, 07:54 PM

I don't think it actually takes long to adapt and change the fuel trims, as it will respond to the data from the rear O2 sensors very quickly once they've heated up. My last car would only take a max of 15 minutes to finish "optimizing", regardless of whether the previous fuel maps were reset or not.

**rf900rkw** · 10-30-2017, 10:36 PM

This has nothing to do with the computer relearning. The stated symptoms simply do not match a minor vacuum leak. If I was diagnosing this, I would start with a leak down and compression tests. But then, I said the same thing when the engine was cooked in a different thread a few weeks back, only to be told it was running "great".

**JKuhn** · 10-31-2017, 02:52 AM

The top-end is fine though. It's only in the lower RPM range that it struggled. And it had a rather large leak.

Anyway, I'll see if it continues to run well today, because as I said it also got that missing power and smoothness back in the past, and back then it didn't last. That's why I'm careful to be too optimistic.

I suppose I could do a compression test, but on a car like this (compared to the old ones where everything was exposed, and you had normal HT leads) it seems like it will be more effort for somethign that doesn't look like it will be needed. I'll do one if I suspect a loss of compression though.

EDIT: I don't actually want to end up driving the Z3 so much since it's expensive to drive, but I'll give it another go today and see if the power is still there. In the past it always disappeared quickly.

**deni2s** · 10-31-2017, 04:56 AM

Originally Posted by LannVouivre

I don't think it actually takes long to adapt and change the fuel trims, as it will respond to the data from the rear O2 sensors very quickly once they've heated up. My last car would only take a max of 15 minutes to finish "optimizing", regardless of whether the previous fuel maps were reset or not.

Sorry, but this is BS.
1) Rear O2 sensors don't play any role in ECU adaptations.
2) It's adaptations which needs to be reset, not fuel maps.
3) Yes, adaptations require some time, will not finish in 15 minutes.

OP, start with reading ECU error codes, stop guessing and speculating around.

**JKuhn** · 10-31-2017, 05:37 AM

I'll see when I can have them read. Unfortunatlely the old car clubs I'm in won't be of use, so I'll have to go to a mechanic.

I drove the car again. It's not perfect, but it's definitely much better than it was. Now I'm going to order the new intake boots, and arrange to have the front shocks replaced.

**LannVouivre** · 10-31-2017, 08:20 AM

Originally Posted by deni2s

Sorry, but this is BS.
1) Rear O2 sensors don't play any role in ECU adaptations.
2) It's adaptations which needs to be reset, not fuel maps.
3) Yes, adaptations require some time, will not finish in 15 minutes.

OP, start with reading ECU error codes, stop guessing and speculating around.

I meant front, had to keep telling myself not to put "rear" and erase it like 3 times. Third time isn't the charm. The term I was trying to recall instead of fuel maps was fuel trims. Sorry, the show about supermassive black holes kept sucking me in, lololol, so I lost interest in proofreading.

**JKuhn** · 11-03-2017, 10:56 AM

UPDATE: I finally got the upper boot today, and installed both. The now reset ECU (as per BMW dealership advice) still needs to optimise but so far the car seems to run great, except for one issue. Now the ASC and ABS lights are on...

Hopefully they'll clear but otherwise I'll have to figure out what went wrong.

**JKuhn** · 11-04-2017, 01:03 PM

UPDATE: I went to a show today, and turns out that was a very good idea (see below). Anyway, the car doesn't run right. As it is now, it has lots of grunt for pulling away, and excellent top-end power. But from about 2k-3.5k RPM it doesn't pull right. It's almost as if there's a slight miss (not dead though), but it's hard to really say what's going on.

Anyway, I decided to take a chance and enter a lucky draw at the show, and ended up getting a ticket for a diagnostic test for effectively R50. If I went to an independant BMW mechanic which I don't trust anymore it would have been R350 or so, and over R1k at BMW. I'll take the car in after the weekend and find out what's showing.

Oh, and the ABS and ASC lights cleared.

**JKuhn** · 11-06-2017, 06:36 AM

I had the diagnostics done, all codes cleared except the cat oxygen sensors (decatted), and a DTC41 (P0340) intake cam sensor.

A local BMW mechanic (which I don't trust anymore) said it's likely not the sensor as the car starts fine. He said it's probably an issue on the cam timing, and not the VANOS either. He also pointed out that the engine is very sluggish to rev, and showed me on an E46 how a modern 6cyl should rev. Let's just say the difference was like night and day.

What do you people make of it?

EDIT: The other codes that cleared were DTC 154 (unknown), CA (fuel trim bank 1 control limit), 67 (VANOS inlet cam, end position not reached) and CB (fuel trim bank 2 control limit)

**LannVouivre** · 11-07-2017, 10:43 PM

If you're getting full trim codes, you probably still have a vacuum leak or basically any other issue that would cause it to add more fuel to the point it can't compensate. A problem relating to the intake cam circuit could probably cause that, but I'm tired and lazy and I no like use brain.

You have the DTC for the intake cam sensor circuit, so what does that tell you? If it is getting faulty readings on cam timing, would that possibly change fuel mixture and otherwise affect performance?

**JKuhn** · 11-08-2017, 02:21 AM

So that's what the trim codes mean. Like I said, they cleared, so hopefully that's not a problem anymore.

Regarding the cam sensor code, I'm trying to figure out if that mechanic was right. In the past I didn't have problems with his advice, but he tends to either charge for work without doing it, or he'll sabotage the engine (no coolant, loose radiator hose, replaced gearbox seizing without oil, etc). Also, on the E12/8, he rigged a compression test to show a bad cylinder. He did however once fix an issue on my Mazda's brakes that nobody else could find, and I didn't even ask him about it.

This time though, his info on the car sounded a bit strange. He said that the car has a M54 engine, because of the duel VANOS. He denied the existance of a later double VANOS M52.

**colb** · 11-08-2017, 05:01 AM

Hearing your experiences of the BMW mechanic I would give him a wide berth in the future.
If you have a cam shaft sensor code best get a replacement and fit it, get an oem make such as Bosch or Seimens dont use cheap pattern parts they seldom work correctly or last long. Air leaks are one of the main reasons for engine problems and given the age of our cars its not surprising that the rubber components are prone to cracking and developing leaks into the system. Replace any suspect ones you find. The Maf sensor is also a possible failure point and dosn't always set a EML light however if its dodgy it may well upset the car and cause other sensors to trigger the EML. If you disconnect the Maf sensor and drive the car it should default to default settings and run better than when a faulty sensor is connected. It may give you an idea if yours if duff and needs replacing by doing this. My experience of a bad Maf was bad running and sluggish performance, when I fitted an oem replacement it cured the problem in an instant.
If you havn't got a code reader suggest you get one, cheap enough on ebay and will recoup the cost first time you need to use it.

**JKuhn** · 11-08-2017, 06:07 AM

Don't worry, I'm not going to him to have work done on the car, unless I can stand there and see what he's doing. And even then I'll be hesitant (I did go to him for a second diagnostic test, because I got issues with the car after the first free one, but I stood there, watched him and also discussed the car's problems).

Anyway, can a bad MAF cause issues that mimic bad cam timing (not very willing to rev, lots of rev hang, low-end power coming in very noticeable "steps" and overall very weak on the bottom end, lots of power above 4k, but starts very easily)?

As I said, that mechanic said it probably wasn't the cam sensor itself as the car is such a quick starter. He said in his experiance a bad cam sensor causes trouble starting.

**LannVouivre** · 11-08-2017, 07:47 AM

Leave the MAF alone unless you are absolutely one million percent positive that the MAF has gone rogue. They can fail, but not usually. You have some DTCs to guide you, so work from that starting point. And get a code reader of your own.

If you have a code for the cam sensor CIRCUIT, the sensor circuit absolutely has a problem. It can cause many different symptoms. Just because he's experienced doesn't mean he's experienced everything and knows everything. He sounds like a complete wanker anyway, might be contagious. The last thing you need is a case of wankitis.

I don't think you would have a code for the circuit without a circuit problem. If you have damaged wires, they may lose connectivity while you are driving because of the vibrations and stuff.

**JKuhn** · 11-08-2017, 08:41 AM

I just spoke to the mechanic who does the work for me when I'm not comfortable to do it myself (or if my tools fail me). He said the first step will be to clean the sensor, then to check the circuit, and then to replace the sensor. What do you make of it?

By the way, clenaing the sensor (R460 max, depending on how long it takes) will definitely be cheaper than a new sensor, or even stripping to reach the rest of the circuit (which is aparrantly buried quite deep). The BMW mechanic quoted R1200 to check and set the base cam timing.

EDIT: He also said that a failed sensor should result in a non-starter.

**rf900rkw** · 11-08-2017, 09:01 AM

First step is to read the codes. You finally have done that. The second step is UNDERSTAND them and the potential causes. 340 is not necessarily a signal problem. It can also set if the cam timing is off. The VANOS limit code kind of points the same direction. My next step would be to check cam timing. Even dealer hacks get it right on occasion. I believe INPA will show the cam timing as read in realtime.