OK OK I know there are loads of oil catch tank threads and loads of us have chipped in with our opinions. Suffice to say I've read them all and contributed to a few myself.

This is the thing. I want to make a totally clean, totally correct and as-green-as-can-be catch tank system. At the moment I have kept the crankcase-to-PCV line connected as I do a lot of everyday driving and think it's better to allow the inlet manifold to do its work and keep the crankcase clear when in vacuum which is 90% of the time. The other port on the crankcase breather which would normally go straight into the inlet pipe, now runs through a catch tank filled with fish tank foam which does a great job of keeping the return pipe back to the inlet totally clean. Since doing this I have had no issues with oily turbo outlet or TMIC. No part of the system is vented to atmosphere and I don't want it to be either, because I run a sports cat as well.

So you might be wondering what am I changing? Well I'm still not convinced that the rocker cover breathers only draw IN fresh air from the inlet pipe. I reckon they may well vent outwards back into the inlet pipe either from positive pressure in the rocker covers or from the high air speed in the inlet when in boost. That's easy enough to fix though. I'm just going to stick a spare PCV valve into that breather line so that it's all one-way traffic.

My other issue is that even though my TMIC is bone dry now, I still must be burning off a lot of oil vapour that gets drawn in through the PCV valve into the inlet manifold under part throttle or idle. Seeing as I have an intermittent problem of knock spikes under exactly these conditions, why not intercept the PCV pipe with a catch tank as well? Meaning two catch tanks, one for the crankcase-to-inlet pipe for boost conditions, and on for the crankcase-to-PCV pipe for vacuum conditions. This would do everything I want.

But then it struck me as duplicating things unneccessarily. Why not run ONE pipe out of the crankcase breather port to ONE catch tank. Then have ONE pipe coming back out to just under the throttle body. At this point tee it off to go to the PCV valve and the inlet pipe. Whether in boost or vacuum condition, there will be flow out of the crankcase breather, through the tank and out again. If it's in vacuum the PCV will be drawing it through and if it's in boost, the PCV shuts and diverts the flow directly into the inlet pipe.

Any reason why this won't work?

 

Do you not think perhaps your knock sensor is to blame for the spikes?

 

Possibly. Although they always happen under the same conditions. Full advance which is 46 degrees, very light throttle. If it were a fault with the knock sensor I'd expect it not to give a proper readout the rest of the time, and to spike at totally random intervals regardless of what the engine's doing. It does not do this.

One of these days I will get this sorted as it's annoying to have the CEL flash in my face when I'm cruising to work or tickling along in stop-start traffic, yet never ever when I'm giving it a proper caning.

 

Now you have done it

That answer means more work, More bits to check

Answered in true politician format & i don't me that in a nasty way just light hearted as Subaru drivers are Paranoid as it is !!!

Anyway back to serious stuff.

When i was running 1.5 bar on an Apexi Ecu, I started to suffer with oily residue. I purchased a catch Can, Added an extra port, Spent 3 weeks finding somewhere to mount it , Plumbed it all in back into the intake.

Instantly on full boost the knock increased.
So i vented the catch can to atmosphere, Still the knock was higher than previous (No catch can). I came to the conclusion that the catchcan was a bottleneck, a Fly in the ointment, It could not let the crankcase gases Etc escape quick enough & was building further back pressure.
I removed the catch can & vented all breathers to Atmos, Rockers Tee'd together & exiting ones Side, Crankcase Exiting the other.
Knock on the Commander unit dropped to an all time low & stayed there.

If you do think about it, The standard breather setup is helped by the scavaging effect of the turbo litterally sucking the air out. Once all these items are combined into 1 catchcan with a very small volume, no scavaging effect & very small exit port/Ports Pressure will build quite rapidly.
You blow down a big pipe it is easy, Blow down a small one & you get Hamster Cheeks !!

Just my thoughts / Experience.

Dean

 

I was hoping this was going to be the definitive oil-catch thread. But DeanF has introduced another can of worms! Interesting findings though.

I think the standard breather setup is very good, in the sense it’s a circuit which helps push and pull the oily air and sludge around. Disconnect it from the inlets and it can only push which could make the sludge build-up. (? All theory btw)
I think the trick is to catch the oil in such a way so it doesn’t block the pipe-work, whilst maintaining the original ‘circuit’ (?)

SR, I like your idea. So, in short, you’ve basically pulled away the T breather on the crankcase and inserted your tank in between crank breather and T, providing oil-free air into inlet and PCV? Thus reducing the sludge?

I’m thinking the air from the rocker breathers is being sucked out due to the inlet, but could possibly draw air back in off throttle. A one-way valve could be handy IMO (All in my un-educated opinion btw).

- I await my re-education

 

Dean,

Just re-read your reply, you did have it all plumbed back into the inlet - my mistake.

 

Very interesting Dean. The thing is with mine I don't have any issues with knock when on boost at all. In fact it's the one time I can be sure I won't have any problems; it's mapped spot on. The problem is this very occasional big spike that I get on light throttle which of course will never repeat come mapping time. But that's getting off the subject.

Here's an idea. Could your knock problems have disappeared once you totally vented the breathers out, because you were no longer drawing back ANYTHING into the inlet pipe? I.e. the oil vapour is in the crankcase gas, but once the oil drops out in the catch tank, it just leaves...well I don't know, but it's not a vacuum and it's not air with its full complement of oxygen. Whatever it is, when it gets drawn back into the inlet, even though it's clean of oil vapour it's 'diluting' the oxygen content of the inlet air. That must effect knock in some way when you're really working the engine hard?

I did think about the fact that the pipes could be a restriction to flow through the catch tank, which was the reason why I initially thought about using two separate tanks and pipe runs. But after I mulled it over, it's not as if I'm expecting twice as much flow to be returned to the inlet...it's just the same pipe used for two jobs which never really happen at the same time.

 

I have recently worked on a car that had random periods of 45-60 on the knock, only on light throttle and sustained overun, nothing could be hear on the det cans at all. This is in contrast to cars that have light det on liftoff and perhaps a few clicks on initial overrun too, all of which is clearly heard through the det cans.

Excessive carbon deposits in the head or on the piston crown could be to blame, but you really need to listen to the engine and confirm with your ears what the PowerFC is seeing.

The crankcase should breath as freely as possible. I always try to get at least a 20mm hose on the outlet of a catch can.

Paul

 

Gingerbread man - that's the thing I'm uneasy about, the fact that as a closed circuit it's a combination of push and pull that keeps the crankcase clean, and once you vent them all to atmosphere, you've taken away the 'pull'. And yes you've said what I want to do in a much simpler way than I could have put it. ALthough I haven't done it yet, I'm planning to very soon though. At the moment, the PCV pipe is still connected normally and the only place where a catch tank intercepts is between crankcase and inlet pipe. This on its own keeps my turbo and FMIC totally clean. It used to be full of oil before I did this. In fact it even got oil in it when I used a plain catch tank without filter foam in it. You could see the oil fouling up my clear return pipe to the inlet, although not as much as went into the tank. Soon as I cut some fish tank filter foam to size and put it in, bingo, a totally clean return pipe and no apparent change in flow.

 

But is it better in your opinion, Paul, to just let the crankcase vent out to atmosphere, or to have vacuum in the inlet pipe or manifold pull the fumes out as well?

 

plumbed mine like u have 12 months ago and have never had any problems so I don't think your problems are down to the breather system.
ken

 

Do you mean how I've got it right now (catch tank in crankcase-to-inlet-pipe line), or how I'm planning to have it (intercepting both the inlet pipe and the PCV line)?

 

plumbed it from crankcase pcv valve to catch can then to atmostphere. (blocked off hole in inlet pipe). I then used a second pcv valve between inlet pipe and rocker covers tee piece.

 

Interesting thread, gone off the original topic a bit but maybe explains what I am seeing?

When I am on say wide open throttle in fifth gear and revs rise from 3,000rpm to 6,500rpm at about 5,000rpm I will see a quick flash up the knocklink to red and then it disappears. I only get this flash to red on 100% throttle and it is a quick flash.

Do you think this could be related to the catch can I have fitted? The hoses are quite large throughout and from what Paul stated above I would estimate them to be 20-25mm.

 

That sound like something else to me although I'm no expert. It's giving a knock on full load and sounds like it's repeatable? Maybe it's a weird spot in your map that needs fixing?