E39 catless headers on E46

[Breeze1]

Hello all,

So I've bought E39 (catless) stock headers for cheap and would like to heat wrap/port/install them instead of the ones that come stock on E46's, to gain something like 10-15 hp, but also because I'm highly suspecting that at least one of my cats is shot ("echo" sound coming from where they sit, pops and crackles on throttle release and last but not least: I recently found out that a piece of mesh had made its way back to the muffler ; since I got my 330i exhaust line used I can't tell if it came from one of the cats currently installed or was already there before, as a result of previously damaged cats).

Anyway, I intend to flash my DME to account for the lack of cats and SAP (already removed), but I probably won't be able to do it the same day I swap the headers, which brings up a question: apparently driving without cats and without an appropriate software will make the sensors think that the cats are overheating and enrich fuel delivery to "cool them down". Running too rich would cause cylinder bore wash, but basically how fast? Can I get say 200 miles worry free before flashing the DME or is it merely a matter of miles before causing serious damage ?

One of the alternative solutions I've been thinking of, since some adaptations are going to be needed for flanges anyway, is to solder two small hollow tubes that would act like housings for the sensors, which would trick them into "thinking" that emissions are fine. Does that sound feasible? I'll go with the flash no matter what, but that would buy me some time...

Any input appreciated.

Edit: I could have just "gutted" my own cats and left my headers on, but I want to get rid of that echo sound (likely coming from resonation inside the larger sections of the cats) and I'm not sure empty cats housings would be good for exhaust gases velocity.

[bmwdirtracer]

Um, sorry, those are not headers. Those are terribly ugly crude manifolds, which aren't going to do anything for your power output. They are just logs.

These are headers:headers.jpg

Mandrel bends, smooth transitions, flowing collector. That's where the potential power increases come from, not from deletion of the cat. IF you ran long tube headers like these, AND you ran a largely open exhaust, and you had the proper tune and complementary intake mods, you might find that 20 horsepower, although you'd lose low and mid range torque. Those logs are not power adders, sorry.

The sensors don't think the cats are overheating . Even the DME of your car isn't going to see it that way. Sure, it's going to set catalyst efficiency codes, maybe it will add or subtract a bit of fuel because of slightly different exhaust breathing, but nothing significant, unless your present cats are dramatically clogged.

[MauiM3Mania]

Any input appreciated.

Swap in a S54. Better yet, swap your car for a E46 M3, complete with the S54.

[Breeze1]

Um, sorry, those are not headers. Those are terribly ugly crude manifolds, which aren't going to do anything for your power output. They are just logs.

These are headers:headers.jpg

Mandrel bends, smooth transitions, flowing collector. That's where the potential power increases come from, not from deletion of the cat. IF you ran long tube headers like these, AND you ran a largely open exhaust, and you had the proper tune and complementary intake mods, you might find that 20 horsepower, although you'd lose low and mid range torque. Those logs are not power adders, sorry.

The sensors don't think the cats are overheating . Even the DME of your car isn't going to see it that way. Sure, it's going to set catalyst efficiency codes, maybe it will add or subtract a bit of fuel because of slightly different exhaust breathing, but nothing significant, unless your present cats are dramatically clogged.

Thanks for your reply.

I thought "headers" was pretty much a synonym of "exhaust manifold". I would gladly put one of those instead of the E39 manifolds, but even getting one of the "ebay headers", would cost (with shipping) about 10 times what I paid for them ($40). So not now...

It's true that they're ugly and I wonder what BMW were thinking when they decided to put them on one of their cars

I still think they are better shaped and seem less restrictive than the other, flat headed ugly manifolds, the ones currently in. But let's agree on the fact that there is little to no hp gain to be expected. I will still be satisfied if:

- they allow me to get rid of that echo coming from what I suspect to be one or more gutted/broken cat(s)
- the fact that they are catless will allow me to heat wrap them, as opposed to cats (because of overheating issues).

I have a mandrel tool so I was thinking I would at least polish away the rough patches on the ports.

[Breeze1]

Swap in a S54. Better yet, swap your car for a E46 M3, complete with the S54.

When the M52TUB25 blew up I asked around for one but couldn't find it used. New would obviously have been out of question because it probably costs 3 or 4 times what my whole car does (and used Bmw's are not as dirt cheap here as they are in the US). And even if I did find one, replacing any part on it (plus all the necessary upgrades around it to make a safe, coherent car) would have meant going to BMW because M parts can't be found used or OEM in Morocco. Too little M3's were put on the market to incentivise parts dealers to offer them.
Swapping the whole car = same thing + very expansive yearly emissions tax.

It might be hard to believe given the kind of mods I'm getting into, but I don't find my car to lack power at all; let's just call it a combination of exploiting "easy hp" opportunities and a weekend hobby/way to learn new things.

[Breeze1]

Ok so I have done what's described in my first post today, and actually I'm not getting any catalyst efficiency codes but rather random cylinders misfire codes and actual, heavy, misfiring. Whenever I shut it off and restart it, the SES light disappears and the engine runs like a clock for like 1 minute and starts misfiring, with the SES back on. On a side note, I'm sure the manifolds were properly and tightly bolted on, with the appropriate gaskets.

Anyway at this point I think it's due to the post catalysts sensors going nuts (the pre-cat sensors have been both put into the new manifolds and they fit perfectly, although located further down as seen on the pic above), and that I have to apply a software solution to this, to basically deactivate the sensors readings. I know of one way, which is flashing a catless "ZB" number software to the DME. 328Power04 had given me one of those numbers (thanks Abel); that calls for two questions:

- the ZB number I know of, which has Cats and SAP deleted, is for a Euro car: will it be fine for a US car? I'm thinking particularly of the speedometer and mileage readings, econometer units...
- Is there a way for me to make my life easier by avoiding the flash and basically "turning off" those readings in my current DME software, using Winkfp, NCS Expert or INPA?


The car runs like crap and I'm not even sure it's not hurting it so any input would be highly appreciated.

[MauiM3Mania]

Swap in a S54. Better yet, swap your car for a E46 M3, complete with the S54.

.... And even if I did find one, replacing any part on it (plus all the necessary upgrades around it to make a safe, coherent car) would have meant going to BMW because M parts can't be found used or OEM in Morocco. Too little M3's were put on the market to incentivise parts dealers to offer them.
....

..... The car runs like crap and I'm not even sure it's not hurting it so any input would be highly appreciated.

I understand about a minimal supply line. With all of 160K ppl on Maui, it's a struggle. With Casablanca's 3.5 million, you'd think that you would have options. LOL

I stand with my input of a few weeks back, but I get your point. It's fun just playing... but some real results are rewarding. Ask the BMWDr about the euro motor that was imported from 10 time zones away and he managed to install it into my 99 M3 over a weekend in a small shop in Kula, Maui . It's been 10 years, but it still rips.

[Breeze1]

Casablanca is closer to 7 million inhabitants, and it has more to do with supply and demand, which is very low as far as M parts are concerned, for various considerations .
Any input on how to disable the sensors without having to do a DME flash?

[Breeze1]

Okay so I went ahead and flashed it with ZB number 7528722 (euro without cats and SAP), and I thought that was it since it idled like a charm for at least 10 minutes, if not 15, revved like a charm several times, no SES, nothing.
Had the maf sensor unplugged for those 10 minutes or so, shut engine off, plugged maf's connector back in, started it up, still ok for a few minutes, and then it got rough again but without throwing any misfire codes.
I cleared the (ancient) error memory in INPA, turned it back on, idled for about 5 min with no issues. Shut it off again for 3 hours+. Went to start it up on a "stone cold engine", back to exaclty the same point I was at before even flashing it. Heavy misfiring and heavy backfiring on the slightest throttle lift. It does seem to get better as the engine warms up, but nothing to be satisified with. I'm getting 0 codes after half an hour of driving.

So not software related. I can only think of 2 things that might be causing that roughness:

- One of the manifolds conducts was soiled with engine oil. I had put some oven cleaning solution in it and rinsed it to get rid of the oil about 2 weeks ago and it was bone dry when I put it on, but there were still traces of oil in that conduct. Possible O2 sensor contamination?
- Possible o2 sensors swapping? Would it actually cause that? If there is a way to swap the connectors around without unscrewing the actual sensors I'll try it, otherwise it'll have to wait next weekend.

Any ideas?

[Breeze1]

Bump. Another thing I was wondering about: is it possible that the E46's O2 sensors are simply not meant to function on the E39 manifold because of where they're placed (further down the manifold, whereas on the E46's they sit high, right after the engine head)?

Also, since I have flashed the DME with the appropriate software, I have disconnected the secondary O2 sensors. Should they be kept on the exhaust line regardless?

[dworthy]

Placement of oxygen sensor in the stream doesn't matter, as where they are in reference to the cat's. So if you don't want the secondary or post oxygen sensors, I think you have to flash it in.

Also the element in most oxygen sensors are fragile, so you cannot bang them around as they will break.

I wouldn't bother with the E-39 headers as it maybe more problems then they are worth, for you would have to modify the exhaust just to make them fit.

[Breeze1]

Placement of oxygen sensor in the stream doesn't matter, as where they are in reference to the cat's. So if you don't want the secondary or post oxygen sensors, I think you have to flash it in.

Also the element in most oxygen sensors are fragile, so you cannot bang them around as they will break.

I wouldn't bother with the E-39 headers as it maybe more problems then they are worth, for you would have to modify the exhaust just to make them fit.

Thanks for your reply and sorry for the relatively long post hereafter.

I deleted the post cats sensors (circled in red on this pic; I basically disconnected them from the connectors that sit on top of the engine), flashed the DME accordingly (with ZB number 7528722: no cats, no SAP) and put the pre-cat sensors (circled in green) on the E39's manifolds, where they bolted on perfectly. The necessary modifications have already been made and they now fit to the rest of the exhaust line.

At some point (before flashing the DME) I got misfire codes for cylinder 1,2,3, which makes me think that the culprit for the misfires is the lambda sensor down that bank of cylinders.
- A possible cause for this is that one of the conducts on one of the manifolds was dirty with engine oil. I cleaned it the best I could before putting them on but it definitely wasn't 100% clean. So maybe it soiled the sensor to the point where it's not reading correctly anymore.
- Another possible cause is that, while I was cleaning the sensors with a metal brush and a white spirit solution before putting them on the manifold, I accidentally dropped one of them on the ground (which brings me to your comment in bold above).

That brings up a few questions:

- Are parts 1 (regulating lambda probe) and 2 (lambda monitor sensor) shown here http://www.realoem.com/bmw/enUS/show...diagId=11_2181 interchangeable? The idea is to replace one of the pre cat sensors with one of the post cats, IF they are interchangeable of course... EDIT: please disregard that 1st question, since I had missed something obvious: while the sensors themselves look identical, the connectors are different; so I guess I'm in for a sensor replacement if it's confirmed that it's faulty.
- If the engine oil is really the culprit, I don't think it would be wise to replace the damaged sensor with a good one before that oil is burnt away by engine gasses (risk of damaging the replacement one as well...). So the idea here is to disconnect the pre cat sensors as well as the MAF sensor to force the DME to run in Open loop for a few hundred miles and then replace the lambda sensor(s). I don't know:
1) if that would get temporarily rid of the misfires
2) if that would cause any damage to the engine (cylinder bore wash...) to do that, even if it is possible. I intend to drive sensibly during those few hundred miles if that makes any difference lol

About "how worth it is", I had gotten rid of the rough patches on the manifolds' inlets and outlets and tried to "port them" as best as I could using the gasket as a guide, and heat wrapped them. When the car does NOT misfire, it definitely sounds like the gasses exit the engine much faster and also louder obviously. Basically it's the same sound but 20-30% louder than with the cats (probably even more at WOT). The rest of the exhaust line is fully "stock 330i", with both silencers and muffler.

[dworthy]

No worries mate, for this is always fun for me!

So you cannot swap around the sensors, as the "regulating lambda probe" will be a wide band sensor, and the "lambda monitor sensor" will be a narrow band sensor so it would throw the DME into a tizzy. It is the same as just taking any ole oxygen sensor and plugging it in, for there are specific parameter programmed into the DME based on conditions.

If there is or was oil on the manifolds, after 5 minutes of running, should be burnt off. Also the oxygen sensor doesn't use the body as a ground, so that shouldn't have any impact on it's function. Now, dropping it can cause the probe to be damaged, so I would start there.

Running open loop will of course cause the car to run rich, but with not cat's all it will do is use a little bit more fuel, and run the car hotter. I don't think it will change much of the BPH with the cat's gone, just don't go straight pipe all the way for that will really compromise power in the low end.

[Breeze1]

Thanks! Always fun for me as well, although not so much when the results are disappointing, as is the case here (hopefully just because of that O2 sensor). My comments in blue below, to make for an easier read.

No worries mate, for this is always fun for me!

So you cannot swap around the sensors, as the "regulating lambda probe" will be a wide band sensor, and the "lambda monitor sensor" will be a narrow band sensor so it would throw the DME into a tizzy. It is the same as just taking any ole oxygen sensor and plugging it in, for there are specific parameter programmed into the DME based on conditions.

Got it. I thought the regulating lambda probes were narrow-band sensors, since loggers such as Rom Raider don't log AFR's whereas they do log all other necessary parameters from the DME. The reason as to why is beyond my understanding... I guess I'll still have to order an aftermarket WB sensor after all...

If there is or was oil on the manifolds, after 5 minutes of running, should be burnt off. Also the oxygen sensor doesn't use the body as a ground, so that shouldn't have any impact on it's function. Now, dropping it can cause the probe to be damaged, so I would start there.

Got it. I thought whatever part located inside the sensor wouldn't be able to read O2 levels properly if soiled by engine oil, but as you say after about 50 miles it should have been cleaned by now. So I'll just replace that sensor downstream of bank one (if that's the bank matching cylinders 1, 2, 3).

Running open loop will of course cause the car to run rich, but with not cat's all it will do is use a little bit more fuel, and run the car hotter. I don't think it will change much of the BPH with the cat's gone, just don't go straight pipe all the way for that will really compromise power in the low end.

Hard to tell with the current issue, but maybe 30s on my way back from the mechanic I was able to go WOT without it misfiring, and I did feel a slight torque loss at low rpm and a and healthier behavior in the upper revs. After the flash as well, for the 15min described above, there was no misfiring and it was revving (in neutral) like the devil was on its tail lol.

[dworthy]

The cat's make back pressure which is good for low end torque, while at hi end you will see a little bit more as it can breathe easier.

The key is to make sure the exhaust is fully closed, that way you will get proper flow. Just think of the intake and how bust vacuum line mess with the power.

[Breeze1]

The cat's make back pressure which is good for low end torque, while at hi end you will see a little bit more as it can breathe easier.

The key is to make sure the exhaust is fully closed, that way you will get proper flow. Just think of the intake and how bust vacuum line mess with the power.

I don't believe there are any exhaust leaks but I will take a look just in case, when I replace the sensor on saturday. I get your point about the intake/vacuum. Basically you lose pressure, "negative" (sort of) in case of the vacuum and "positive" with the exhaust.
Once I fix this I'll have fred up the whole "air system" of the car with the exception of the cams, but given what mods I have upstream (CAI+ M54 Intake manifold) they shouldn't really be restrictive I guess...and they can still be tuned before they need to be replaced anyway (but not any time soon)

[Breeze1]

So we looked into this yesterday, and replacing successively both o2 sensors didn't resolve the misfiring issue, past the few minutes the DME seemed to need to make its measurements and make the car runs like crap lol.

The DME being flashed to run without cats and the o2's being good, the only possible cause left seemed to be that we found out the 1st (largest) resonator had gotten hollow although stock, with time. Therefore it's not providing any backpressure (messing up the o2's readings?) and apparently it's also causing the backfiring, which seems odd on an otherwise full stock exhaust (I did find two very minor leaks on the muffler, but they're probably too far away from the engine to have any effect on how it runs).

Decision was taken to simply disconnect both sensors and put bolts instead in their housings. The car nows runs in open loop all the time, with the MAF sensor still plugged in though.

It has resolved the misfiring issue and lessened the backfiring, although still present.

The car's behaviour has changed, I believe for the best. Although -slightly- rough on cold startup, once warmed up it idles very smoothly. The throttle response has changed in behaviour: much more linear and "analog" in its feeling. Basically the scope on which it influences acceleration has widened and gotten into a "you get precisely what you ask for" mode, as if there is less "thinking" involved from the DME and more control from the driver. Gas mileage is hard to assess after only a couple dozen miles.

And the sound... oh well, too much to say really

[dworthy]

True, take the oxygens sensor out as the flash isn't suppose to have them installed anyway. Now it will run under a stock fuel map and it's only references will be the MAF and temp sensors to control how much fuel goes into the motor.

I still think there are other problems, but as you said, it runs good now.

[Breeze1]

Thanks for your feedback mate. I replied to some of it in blue.

True, take the oxygens sensor out as the flash isn't suppose to have them installed anyway. Now it will run under a stock fuel map and it's only references will be the MAF and temp sensors to control how much fuel goes into the motor.

I thought the pre-cat sensors would still be active even with no cats and the matching DME flash. I guess closed loop operation only works as a whole, with all 4 sensors...

I still think there are other problems, but as you said, it runs good now.

I don't know, really... Known issues at this point are this hollow resonator I mentioned (and two minor leaks on the muffler) and a really small leak of the cylinder head cover's gasket, which is due for replacement. As far as air inlet and fuel systems, they both have been thorougly checked and overhauled recently and seem to be working fine. Before taking the cats off, the engine was as smooth as anything. It just seems to have somewhat improved since and to be "breathing" better. The slight roughness I mentioned on a cold -morning- start is very relative. The rpm needle itself is dead steady, but you can hear the exhaust sound "changing frequencies", if that makes any sense, for a minute or two, during the cold start enrichment phase.

[dworthy]

The pre-cat oxygen sensors would really be to ensure the mixture is as close to stoic as possible 14.4:1, thus not running too rich or too lean. Remember lean means pinging with a high-output motor can be deadly.

Not sure if the hollowed out resonator would cause cold running issues, just that you would lose low end torque mostly. The cold start program is just the ECU adding a little bit of extra fuel for the first 60-90 seconds of the start-up.

[Breeze1]

As far as I know, pinging/detonation/pre ignition only happen with a combination of high temps/high loads/ high revs/not enough fuel to cool down the mixture (well there is a case where it can happen at high load/low revs/, but I don't remember exactly how or why, and it's less dangerous than at high revs), which the DME normally prevents by switching to open loop past a certain load/revs points, and relying on fuel tables voluntarily increasingly richer than stoich. Unless fuel starving occurs for a given reason (injectors/fuel pump/unmetered air...), the risk is minimal.

But constantly running open loop like I am now results in lower mileage, which can be compensated with tuning up to a certain point only, since under no circumstances you can target stoich in base tables (dangerous). Stoich is, as you said, the job of the o2 sensors when they're running closed loop and giving feedback to the DME. On the other hand, according to what I read on some online articles, running without cats makes the engine's life easier (less backpressure = less "breathing effort" involved and cooler temps) and should partly compensate the gas mileage decrease due to OL.

The 60 seconds you mentioned is exactly the time it takes for the exhaust "rumble" to stop after a cold start!

[dworthy]

The placement of the cats on a M52TU/M54 engine is just horrid IMO. I understand the reason why it is so close to the heads, but it does nothing for performance. As a matter of a fact it chokes about 15hp on the M-3 for being in that location in comparison to the Euro car.

Yes exhaust flow is important, for each "Pulse" of the exhaust has to be able to move easily before encounter too much resistance. That is why on some cars swapping to a "straight pipe" exhaust kills the car. Thus the market for mandrel bent pipes and tuned/tested exhaust system is pretty big. The Manufactory has to pass a Government mandate initially, so after the warranty period, most owners mod the engine to get it to breath better.