When you click on links to various merchants on this site and make a purchase, this can result in this site earning a commission. Affiliate programs and affiliations include, but are not limited to, the eBay Partner Network.
After reassembling everything I pushed it out of the garage and proceeded to start it up. Well first I referenced the Bentley manual to make sure the coils went to the correct cylinders. They did, so I started it.
And it sputtered and coughed, and ran like crap, let it run for about 10 seconds. Then I shut it down.
Got a sinking feeling in my gut and muttered f*** about 46 times. Pulled the Bentley back out and said to myself, "I must have messed up the plug wires."
I hadn't.
Well, I figured, maybe some small, really small, fine aluminium particles fell into the intake, were 'missed' by the vacuum - but they'll burn and flush through, it must only be a couple, and some lube oil, that drained in there and it's not helping combustion either. So I've go to run it again, or what's the alternative.
Sinking feeling in my gut returns...
Started it up again, same thing: cough, sputter, mis-fire, putt-putt, definitely not firing on all 4.
Oh and now what's this? I knew you'd be coming to this party sooner or later - Herr CEL and his badder bigger brother. The one in the speedo and the nasty one in the tach (if you haven't ever seen that one, ya don't want to).
So I shut it down and had to laugh at my foolhardy recklessness. Way to go, do some preventative maintenance that destroys the engine, that's logical, ya bonehead!!!
So, if the motor was f*cked, than WTH, I might as well run it and see if it blows the shards through and is: not so bad, totally blown or "OK." It had to be one of the 3.
I started it up again, and more of the same, the lights-a-flashing on the gauges, horrible sputtering from the engine and now what's that I hear? Oh a valve isn't sealing, maybe two or three. Yep, now I am certainly "up s*it's creek and I sure don't have a paddle."
Internal chastising: dude, you are such a dumb-***.
Oh well, at this point what the hell do I have to lose, right? If the valve isn't 'burned,' then it's got a piece of aluminium on it that isn't letting it seal, keep running the ***** and burn it off. So I kept it running, kept opening the throttle, and it would reluctantly and weakly raise the RPM, but it kept running. And then it started to smooth out. There was no smoke from the tailpipe (this whole time). It started getting smoother still, and all the bad noises subsided and it idled just fine. But for how long...
Put the CANtool to work, cleared the codes, they'd come back if they meant something, right? I mean if it's f*cked royally they are going to make sure you know. It's a German car, schadenfreude, right Herr CEL?
Once the codes were cleared and the car was purring smoothly I ventured around the neighborhood, then further still, ran it for about 30 minutes straight and put 20ish miles on it. All seemed fine, no more CELs, fine idle... no complaints.
This is all true, I wish it wasn't to be honest... if you're still inclined to do this mod, please, please be careful and meticulous.
So I did some more 'test' runs the next day, and the next day. Then wrote about the big 'final' exam I put it through (earlier in this thread). I am pleased and immensely relieved to say: (so far) I have dodged another bullet. For how long?, time will tell, but sh*t, I could get in a head-on collision tomorrow too, live a little. Not that I recommend this foolhardy recklessness, not at all. Be safe, be judicious and weight the possible outcomes.
I've been lucky twice. I'm not pushing it for awhile
Now that the carbon build-up is 'fixed' time to get back into the ethanol blending and JB+ waxing.
Last edited by minniehaha; 04-25-2014 at 01:43 PM.
Reason: spelling, grammar, ssdd...
Short update. It's not hard to figure out why our OCC's are not very effective. Recently shut down after my daily commute and went to check the OCC level. Upon opening, the OCC was venting vapors out the top and the can was hot as hell.
The vaporized water, oil, unburned fuel etc. that exits the PCV system only partially condenses before it's ducted back to the inlet. Pulled the inlet tube and verified oily residue on the inside. What we need is a coalescer to cool the vented gases or a cyclonic system to spin it out, but the standard OCC is not cutting the mustard.
As an FYI I built a short ram intake system with a box similar to the DNA Filters Stage 3 intake system. It included provisions for the MAF sensor which the DNA unit does not since it's for EU Mini's. While designing it I decided to vent the PCV to atmosphere which was a mistake. If you vent to atmosphere on the N18 you will get a BMW code 2774 which is MAF sensor too high at idle. The DME looks at the MAF value as being too high because the PCV exhaust is no longer be added to the intake tract so more air must pass thru the MAF sensor. This value is out of range at idle and triggers the error code.
Back to the drawing board to add a PCV inlet to the short ram intake and a better OCC.
Last edited by Tigger2011; 04-04-2018 at 06:19 PM.
Agreed, the designs are not what I would consider impressive. Metal cans that heat soak from the engine compartment, limited baffling if any, no centrifugal designs whatsoever, blah blah etc. It just seems as if very little ingenuity has been applied to the design of the can itself.
Agreed, the designs are not what I would consider impressive. Metal cans that heat soak from the engine compartment, limited baffling if any, no centrifugal designs whatsoever, blah blah etc. It just seems as if very little ingenuity has been applied to the design of the can itself.
Some have argued that's it's better to have oil in your intercooler and intake plumbing than have it go into the engine, it's stupidity!
Have you thought about making a coil of copper tubing as a condenser to go in-line before the OCC. Get some 3/8" copper flex and wrap it around a convenient form in a spiral, kind of like a still. ("Moonshiners" meets Minis!) You'd have to mount it above the OCC inlet so gravity would drain it for you, someplace with airflow so it wouldn't heat soak. Or use an oil cooler in-line as a condenser.
What is the velocity of the vapor coming out the PCV port? Is it fast enough to make a cyclonic separator effective?
If the OCC were more effective you would be pulling out as much oil as you would have burned. So if you have to add 1.5 quarts every 1500-2000 miles you would have about that much in your OCC.
With all the copper tubing I might have to take it to my Uncles Jed's place and park it by the cement pond
I'm thinking more along the lines of a four stage separator. A central venturi shaped spiral inlet surrounded by a nylon mesh and a metal mesh around that in a PVDF can. That would have the advantage of cooling the gas via the venturi effect with condensation being assisted by the centrifugal forces applied. Then surrounding that with two separate materials of different wettability to further enhance droplet coalescence in a can that will not heat soak, is chemical resistant and can withstand 300 degrees continuously.
I'm going to need to fire up autocad and sell a kidney to prototype it but what the hell.
Last edited by Tigger2011; 05-08-2014 at 08:20 PM.
I have got a LOT of good/useful info from my expirience (post port-plugging) to add here: my "homemade" system, trials with it, coalescing, plumbing, etc. Been really busy with other stuff to post it all here, but I will try to get it posted up this weekend with pictures; hopefully it will be helpful to everybody using an OCC with the N18.
Airwolf has the best air/oil separator out there ... they are the supplier for Cummins, Detroit, CAT, & Volvo engine factories and used on many private piston power airplanes
it is also huge, 4 inch diameter x 6 inch tall, they also make a smaller size, 3 inch diameter x 4 inch tall but i can tell you from practical experience the smaller one is not as effective
another point ... they use a pressurized aux air source, the outflow from a vacuum pump
follow the link, take a read
i do not work for or have any affilliation to airwolf other than being a customer
Additional & Important notes about my experience - essentially a strong word of caution to anyone else that might go this route.
I was not as fastidious as Tigger had related and strongly emphasized in his post about this modification; it you don't think you're at his level of meticulousness - DO NOT TRY THIS.
Full Disclosure: I did the plugs as reported and everything is totally hunky-dory with Haha; I am confident in its full capabilities and will continue to mod and drive it hard, and report.
But - (and I relate this to 'warn' others) - the task of plugging the ports was demanding and fraught with multiple instances for substantial 'error.'
And I experienced some. I am not even going to chronicle the other miss-hap I encountered when I plugged the round port below the orange diaphragm valve, suffice it to say: DO NOT UNDER ANY CIRCUMSTANCES DO THAT.
So, this is how my plugging went down:
I used two shop vacuums, one at the respective intake port that was being modified. It has a flexible fitting that fully conformed to the intakes elongated oval shape and formed an excellent seal, this vacuum ran continuous during the threading operation. The other vacuum was placed alongside the tap as it cut it's way into the port. I didn't "seal" the combustion chambers in any other manner, I supposed that the vacuum was sufficient to collect all debris as the threads were cut. It was highly effective but not completely (more on this later).
After completing my first thread cutting (on cylinder #4, the furthest one to the right of the engine when standing in front of the car) successfully and without a hitch, I moved unto cylinder #2, with much confidence. A little too much. But first: as Tigger has related - the spring perch bolts must be removed, for access to the port at the right angle, and a 'handle' to turn the thread-cutter is necessary here, the typical "T" used in this type of operation will not fit. So, as I moved unto #2 I began to cut the threads and I proceeded at a rate that was not properly judicious, it was on the other hand a bit too overzealous and I ended up breaking the tap off. OMG! I was rather dejected, to say the least. I kept a positive outlook though and figured I'd just drill and tap it out. That didn't work. 1. because the doggone tap was so small, 2. because the metal is hardened for its purpose: to cut threads.... SO - if a thread-cutter breaks in your cylinder head and you are set on removing it, you WILL need to pull the head off. Well, I wasn't going to pull my head off for that, I mean I was plugging it, right? The tap was pretty much doing that now...
So, and I relate this with some trepidation, as it's not a practice I endorse. I removed a small bit of the tap that was still protruding above the port and moved onto cylinder #3. [This one (#3) and the port at #1 that followed were done with extreme caution and went fine. (Another word - if you've never tapped before this IS NOT the job to 'learn' it on, and if, like me, it's been a few years, TAKE YOUR TIME!!! Cut a very little, back it up, cut some more, do this over and over and keep it lubricated)] After finishing #3, with a proper threaded plug, and having the tap still broke off in #2, I applied this to the top of the joint port to complete the sealing: http://www.jbweld.com/product/j-b-highheat/ I know, shame, shame, JB Weld, WTF!, is this grandpa's tractor?!?
Well - it was that or pull the head... and given the conditions: the ports were "plugged", the 'load' on the epoxy would be a negative pressure, mostly, (or a boost pressure otherwise) and those pressures would only reach the epoxy through the small 'openings' of the tap. And these small openings, plus the taps rough break surface, plus the port itself's rough finish made ideal conditions for the adherence of the epoxy. This epoxy is robust and tenacious, btw. And it has withstood a 'top speed' WOT run of more than 1.5 miles (that is essentially the most load the engine can or will experience) as documented earlier in this thread. I am confident in it.
So, I encountered a broken tap, but soldiered on and finished up the job. I will post the crappiest photos ever (my phone sucks for photos, though it does make good calls, go figure, lol...) - the photos are horrible and the setting (in a garage with shoplighting) only exacerbated it's crappy quality, but I'll get them added for documentation.
Then what came next almost made me vomit. The start-up....
Gotta take a break, I will return to finish this up later, hopefully today.
Added the pix, of note they are "upside-down" so to speak, their view point is as if you were standing through the windshield. Also - #7605: that's the port for #4 or #1, they're pretty much the same and you can make out the threaded plug's inside hex top. #7606, that's the "pièce de résistance" the joint port, flushly filled with JB epoxy.
Just read through this wonderful thread...
Would plugging the ports with JB weld be sufficient for sealing, as opposed to threading and tapping?
I don't know if this would work or not, but would putting an N14 valve cover on the N18 mitigate any of these issues? Or is tapping the block to seal those ports the actual part thats necessary?
The valve cover swap is a no go as they are not interchangeable. There is always more than one way to skin a cat but I would have been uncomfortable trying something else that could possibly become dislodged and destroy a valve or piston.
Short update. It's not hard to figure out why our OCC's are not very effective. Recently shut down after my daily commute and went to check the OCC level. Upon opening, the OCC was venting vapors out the top and the can was hot as hell.
The vaporized water, oil, unburned fuel etc. that exits the PCV system only partially condenses before it's ducted back to the inlet. Pulled the inlet tube and verified oily residue on the inside. What we need is a coalescer to cool the vented gases or a cyclonic system to spin it out, but the standard OCC is not cutting the mustard.
As an FYI I built a short ram intake system with a box similar to the DNA Filters Stage 3 intake system. It included provisions for the MAF sensor which the DNA unit does not since it's for EU Mini's. While designing it I decided to vent the PCV to atmosphere which was a mistake. If you vent to atmosphere on the N18 you will get a BMW code 2774 which is MAF sensor too high at idle. The DME looks at the MAF value as being too high because the PCV exhaust is no longer be added to the intake tract so more air must pass thru the MAF sensor. This value is out of range at idle and triggers the error code.
Back to the drawing board to add a PCV inlet to the short ram intake and a better OCC.
It worked really well. Once I added the PCV port the error code on the N18 went away. Did some further development by adding a clear acrylic cover to further isolate the intake from engine bay hot air. Started working on a carbon fiber version but put it on hold. Currently planning to pick it up again in conjunction with another project. Haven't decided if I'll put that in the build thread for Vlad or start a separate one for it. I'll have to do some dyno comparisons in stock and tuned condition.
It worked really well. Once I added the PCV port the error code on the N18 went away. Did some further development by adding a clear acrylic cover to further isolate the intake from engine bay hot air. Started working on a carbon fiber version but put it on hold. Currently planning to pick it up again in conjunction with another project. Haven't decided if I'll put that in the build thread for Vlad or start a separate one for it. I'll have to do some dyno comparisons in stock and tuned condition.
CAI you made, that is on an N18? On mine, that is where the radiator overflow tank is. Did you relocate yours and how?
Very cool! Glad to hear that. I was buying the parts to basically do the same thing then saw your post about it. That would be very cool to read, once you write up that thread. All I really need to do is relocated the overflow/reservoir tank and I'll be in business! Then of course... make an actual box at some point. I suppose, still better than at the back of the engine.
Do you recommend the intake pipe diameter the size of the MAF (2.75" I believe)? Or the size of the turbo inlet?
Originally Posted by Tigger2011
It worked really well. Once I added the PCV port the error code on the N18 went away. Did some further development by adding a clear acrylic cover to further isolate the intake from engine bay hot air. Started working on a carbon fiber version but put it on hold. Currently planning to pick it up again in conjunction with another project. Haven't decided if I'll put that in the build thread for Vlad or start a separate one for it. I'll have to do some dyno comparisons in stock and tuned condition.
I used 70mm for the MAF, same as 2.75. I'll have to go back and check my notes on the reducer size for the S and JCW turbo inlet diameters. My X51 turbo inlet is a tiny bit larger at 60mm
I know this is a very old thread. I'm trying to get as much info as I can about how the PCV system works on the N18 engine.
I have the engine as shown in the picture below (not my engine but same type) with the 'simple' pcv sytem as I call it. It's just a single hose from the valve cover to the inlet tube. So not with the sensors at the Y insert near the turbo and to the side of the inlet pipe. Which I highlighted in blue.
The 4 little holes shown in the cilinderhead that are blocked/plugged in this thread. These actually connect to the inlet ports? So if I theoretically would poor water in those tiny ports that would flow onto the intake valves? Is that correct?
So in other words these ports allow crank case vapor and oil mist to be sucked up through the valve cover directly into the intake and doing so partially cause carbon build up on the valves. Or is this considered the main cause?
So blocking off these ports solves carbon build up by a big extend? Looking at the valve cover you could also block the ports in there instead of the cylinder head and save a lot of hassle. maybe glue a tiny blocking plate onto the valve cover that fits nicely in the gasket.
This won't hurt the engine by causing an increase in crank case pressure or other issues?
I've read in an other thread that for the PCV system to do its work (any PCV for that matter) you need a vacuum to suck out those vapors because these are very bad for the oil quality and engine life. So you can not just vent to air unfortunately. So what do these little ports do compared to the tube that goes in to the inlet pipe just before the turbo?
Sorry, lots of questions. Just trying the wrap my head around these direct injected engines and PCV systems and how to mitigate carbon build up.
Interesting... What year and model is your car? What country are you located?
The ports in the head do go down to the intake valves. The purpose for blocking those off would be to prevent the oil residue from getting to the intake valves.
Yes, the crankcase needs to be under vacuum to function best. Those small ports provide the vacuum when you are not in boost. When you go into boost, the tube to the turbo intake provides the vacuum.
04-25-2014, 01:27 PM
The one in the speedo and the nasty one in the tach (if you haven't ever seen that one, ya don't want to).
Yep, now I am certainly "up s*it's creek and I sure don't have a paddle." 
If the valve isn't 'burned,' then it's got a piece of aluminium on it that isn't letting it seal, keep running the ***** and burn it off. So I kept it running, kept opening the throttle, and it would reluctantly and weakly raise the RPM, but it kept running. And then it started to smooth out. There was no smoke from the tailpipe (this whole time). It started getting smoother still, and all the bad noises subsided and it idled just fine. But for how long...
OMG! I was rather dejected, to say the least. I kept a positive outlook though and figured I'd just drill and tap it out. That didn't work. 1. because the doggone tap was so small, 2. because the metal is hardened for its purpose: to cut threads...
lol...) - the photos are horrible and the setting (in a garage with shoplighting) only exacerbated it's crappy quality, but I'll get them added for documentation.
