On PCV’s, check valves and catch cans
 

My turbo is in, and I’m blowing boost past the stock PCV and into the crankcase. This has prompted me to read the 7 million threads on PCV and catchcans here., none of which are conclusive. A plan is developing.

One of my options was to ditch PCV valves entirely, and go with a catch can and check valve a la the Deezums catch can thread. I always assumed that PCV valves were just check valves, so why not. Did some research, and figured out that PCV’s work in more complex ways. Good link here:

AGCO Automotive Repair Service - Baton Rouge, LA - Detailed Auto Topics - What are the Symptoms of a Bad PCV Valve

This link explains much of the strange behavior I’m seeing with the five or so different PCV valves I’ve tried. For cars without boost, they only need to keep pressure out of the crankcase during a backfire- a brief, forceful event. As it pertains to turbo cars: it explains why many of them don’t seal properly when the boost comes on. Unless the positive pressure comes on very suddenly, the valve doesn’t seal. Roll on the throttle slowly, blow boost into the crankcase.

They also seal in the other direction at high vaccuum (idle) to keep it from acting like a vaccuum leak, and keep it from drawing oil into the manifold. A simple check valve wouldn’t work at all in this direction, it would suck right through.

What I’m taking from all of this is:
- PCV valve good, check valve bad- at least as far as idle is concerned.
- a catch can is a good idea between the PCV vent and intake manifold, but using actual PCV valve instead of a check valve will probably keep oil out of it anyway.
- the way people are testing PCV valves here (blow in it, suck in it, hook it up to an air compressor) is bullshit. The speed with which you hit it with pressure is a critical factor that’s being overlooked.
- most PCV valves suck for turbo because they won’t hold positive manifold pressure unless they get it in a quick hit. I’m holding out high hopes that the GTX valve I have on order is designed to close more quickly.

Discussion?

Yep.

Ditch it. It is an emissions device. When you hit boost it closes and your crankcase pressure all goes out through the other vent. If you have more Blow by than stock because you're running more cylinder pressure than stock because you are running a turbo then you will need a greater flow path. If the stock flow path is good for 100 horsepower and you are running 200 horsepower then you will need twice the flow capacity. If you are running 300 horsepower then you will need three times the stock flow capacity.

Running a PCV valve is a fucked-up idea for any aftermarket turbo car. If it leaks then you will have a boost leak. If it holds then you are still providing significantly more crankcase pressure than the system is designed for (because turbo) and are going to continue to have problems. I'm baffled why people continue to have problems with this. Not you personally but everybody else that argues, "OMG, I must have a PCV because Mr. Mazda put it there because emissions."

And for the people who feel like when they're in a vacuum situations they must have crankcase draw through so they can have wonderful fresh air, there is a tremendous amount of blow by gas being introduced constantly to the crankcase and needing to be vented out and it completely overwhelms the pitiful amount of vacuumed gas that is going into the intake manifold when the PCV valve is operational. This is why the factory has whatever comes out the other side going to the intake tract in front of the throttle body. And for argument's sake even if it was a tremendously high flow into the intake manifold at idle, anytime you are in boost it's closed and then what? It stops working. And if the flow into the intake manifold is too great in vacuum it closes as was discussed by the OP.

So then- VTA through catch cans on both sides? Or perhaps back to the air intake pre-turbo. I hate the idea of feeding that shit back into my turbo and intercooler, catchcans notwithstanding.

VTA. The oily vapor that comes out of your engine crankcase will cause detonation if your engine ingests it. And it reduces the volume of oxygen you're getting into the engine by giving you something other than fresh air.

Tried the GTX PCV valve that everyone recommends here. In the stupid blow/suck test, it seemed like it might work better. It didn't. Spool data was basically the same as the stock valve (which I'll post in my build thread).

Now that I understand how stock PCV valves work, I also see more deficiencies with the GTX valve at idle. My idle was intermittently 300-400rpm high, because the valve didn't close completely at high vacuum and it was acting like a vacuum leak. Basically, like the simple check valve in the OP.

Sixshooter is right. PCV is stupid for an aftermarket turbo car. Tomorrow I cap the IM and try it with both vents VTA.

So from what I’m reading, the PCV valve (cold side outlet) is an emissions device, and the vacuum from the manifold does little to assist in crankcase ventilation.

Therefore its better to gut the PCV valve, run it to atmosphere?

I wonder if this might be part of the p0300 issue I have...hmm. I've seen it at idle, I've seen it going up a hill after rolling off throttle and I've seen it wide open in fifth gear. Haven't been able to catch a freeze frame at anything other then idle though.

I consider ditching the PCV system the lazy way of going about the problem. Also, I need it in order to be road legal. So what I did is run -10 AN lines with welded bungs, a sturdy metal check valve (that black thing sandwiched between the blue&red AN fittings next to the catch can) AND a PCV. I did the tiny hole mod in the valve cover, too. The correct order is as follows:

intake manifold - check valve - catch can - 323 GTX PCV - valve cover.

This way the only section that sees boost pressure is between the intake and the check valve. The latter is rated up to 500kpa, so that should be more than enough LOL

https://cimg6.ibsrv.net/gimg/www.mia...f99ea77e5b.png
Mo' pictures in my build thread.

No need to gut a PCV valve. I'm not sure how you would do that anyway.

Remove the valve and put a "push in" barb fitting in the grommet or remove the grommet itself and thread the hole for a fitting. Or you can buy a threaded bung to have welded in if you want to get fancy like some of us. It doesn't hurt to increase the diameter if you're planning on running any kind of power. You don't want any restrictions.

All that work on the cold side, yet you’re still VTA through a breather on the hot side? And the added line resistance of the check valve AND the PCV, AND a couple extra feet of tubing on the cold side. And the turbo pushing more blowby into the crankcase than stock? I’m skeptical whether that’s helping emissions much, or vacuuming much of anything out of your crankcase. Does your catchcan accumulate oil?

re: PCV, do you have intermittent issues with idle going high?

Since I started this as a reference thread, perhaps you could post a link or a part number to the check valve you chose, in case future searchers are interested. Nice clean engine bay BTW. Thanks.

For posterity, here is the GTX PCV valve in question.

https://cimg5.ibsrv.net/gimg/www.mia...5d46c182f7.jpg

Here is some interesting reading explaining some of the different systems I've seen people run versions of on here. Take it for what it's worth......

http://www.shophemi.com/images/media..._ccv_bible.pdf

Forgive the noob question, but when you vent to atmosphere (assuming that is what VTA means) does it have to be to a catch can? Is there an easier solution?

Also, am I missing something or is allowing the intake manifold and crankcase to vent by removing the PCV going to lower your boost pressure and essentially create a boost leak?

Trying to wrap my head around this.

From what I’m understanding, you can put a breather on both sides of the valve cover. Just cover up the intake manifold side barb.

A catch can only acts as a air/oil filter.

Can someone please confirm this is a possible solution to not running a pcv?


https://cimg4.ibsrv.net/gimg/www.mia...03be9e2730.jpg

that's still not going to fix your spool problem.

if you really suspect a leaky pcv, add a check valve in front of it.



The GTX pcv is no better than the oe miata one. stop reading posts from 1995...

Brain is right- didn't help my spool data.

If only for idle purposes, I remain convinced that PCV < VTA for a turbo setup.

LukeG, that can work (see most recent post in my build thread). If you have leaky rings, the breather filter may get saturated with oil and crud. I will be putting a catch can in there with a breather on the other end of that.

Awesome, thanks! What catch can are you looking at using? I know a lot of them are garbage.

This one is well-made and appropriately baffled.

Allstar Performance ALL36109: Breather Tank Dimensions: 11-1/2" x 3" | JEGS

http://www.jegs.com/images/photos/0/...9-all36109.jpg

So, basically unless you care about emission, VTA to everything. Catch cans prevent underhood fumes on the way to the atmosphere.

I posted this in my build thread, but figured I'd stick it here too because it's relevant.

THE CAR:
The car in question (mine) is a stock 1.6L motor with MKTurbo setup, no boost control, MSPNP2. Last engine compression test was 3 months ago, 180 psi on all cylinders. I've done extensive boost leak testing, and mitigated all boost leaks except for the PCV setup.

THE DATA:
All of the lines below are 4th gear WOT pulls. X axis is RPM, Y axis is MAP PSI.

THE RESULTS:
- Yellow line is OEM PCV valve. Blue line is GTX PCV valve. Both runs were with the PCV hooked up to the IM. In this test, they performed identically. In both cases, I can hear boost leaking out of the driver side crankcase vent when I'm at WOT.

- When I disconnected the PCV valve entirely (orange line) and capped the IM, spool actually got worse- but only because the boost leak was gone, and it was sending a stronger boost signal to the wastegate actuator, which is activating the actuator sooner. I no longer hear the hissing boost leak.

- When I increased the actuator preload (grey), spool got much better.

THE CONCLUSIONS:

- Brain is right, the GTX valve is a $25 paperweight.

- Sixshooter is also right, All PCV valves leak boost without a check valve in line. At least the ones that I've tested- OEM, two different ones from NAPA, one from Autozone, and the GTX valve.

- Capping the IM and running both crankcase vents to atmosphere appears to be the best way to get fast, predictable spool and stable idle. In my case, it's narrowed things down to a single variable (wastegate preload).

https://cimg0.ibsrv.net/gimg/www.mia...527eb3798b.png

As an interesting sidenote: all of the plots start to break up above 4000 rpm. I'm speculating that this is where the airflow becomes less laminar and more turbulent. This may be where you would see an improvement from better flowing manifold and/or head.

..and prevent blowby in an engine with tired rings?

^ it would be hard for me to speculate on that. My motor has good rings and compression, edited my post to clarify.

Just to clarify...

Was #3 done after #2 and with no PCV valve?

imo, spool-up is more or less identical in each pull. only difference being the knee-point where the wastegate opens.

That is correct.

Brain- I agree with you in the linear regions. This leads us to a theory that I'd like to prove or disprove:

The stock PCV system leaks boost. It's a slow enough leak that the direct affect on the air entering the motor, which is constrained by flow and volumetric efficiency, is minor. The direct affect on wastegate actuation, which involves only pressure and not airflow, is profound.

I'm going to do a run with current (higher) wastegate preload, and connect the PCV valve back to the manifold. Which PCV valve shouldn't matter, but I'll use the GTX one. If this theory holds, the result should be very much like the grey plot. Will try to do this tomorrow.

If it ends up like the grey plot the only thing you've proven is the preload has the profound effect while the PCV or not has negligible effect on boost attainment.

Just trying to understand, not argue the point.

^ the only difference between the blue/yellow plots and the orange plot was whether the PCV was hooked up to the intake manifold. So it's clear that the PCV makes a profound difference to spool, although not in the manner that I expected.

The fact that the lines above all coincide in the linear region, before the wastegate opens, whether or not the boost was leaking through the PCV, would suggest that the boost leak is a small contributor to the dynamic flow of air entering the motor (at least at low RPM/boost levels). If true, this would lead us to the conclusion that PCV systems may still be a problem for aftermarket turbo setups, but not for the reasons we thought.

Thoughts up until now.
- PCV valves leak boost. Seems like they all leak about the same.
- In my simply managed turbo system, the leak has a profound affect on spool, for reasons that are unclear.
- We don't know whether that's because of less boost entering the engine (dynamic), or less boost hitting the wastegate actuator (static).
- The spool graph shows two regions: before the WG opens, and after.
- before the WG opens, the leak doesn't seem to affect spool much. This may or may not be true at higher RPM and/or boost levels, when the motor is operating at higher VE.
- the leak seems to matter a lot as far as when the WG opens, and what happens after it opens.
- my two variables are 1) PCV boost leaks and 2) WG actuator pre-load. These variables may not be independent; perhaps higher wastegate preload mitigates the affects of the PCV boost leak.

IATs need to be compared between the plots or we might be inadvertently measuring the wrong variable.

Sorry to derail again.
When I did a leak down, I had most of the Air coming out the dipstick tube. Meaning I have crappy rings. In addition, I smoke a decent amount under boost. Less so since I changed out an old UEGO sensor, but it’s definitely still smoking.
Im running a GTX PCV valve to a catch can, to the intake manifold. It’s an eBay baffled can (was like 50$, some funny brand name) full of copper wool, 3/8 rubber hose.
I’ll do a couple boost runs, look for smoke. Then cap off the IM, block off the outlet on the catch can and add a breather to it instead.

I need pics of this.

Also: If you don't have your IM attached to your PCV, then you need to not have the PCV in there.

I have that data in the logs. I expect that there will be significant differences between the runs. Let me noodle on how to figure out whether or not it matters; may be able to just do a statistical correlation.

As one observation, I'm pretty sure there was a big difference in ambient temperature between the run with the OEM valve and the GTX valve, yet the plots are identical.

May be easier to start fresh- consider everything above to be the pilot study. If I have time tomorrow morning, I'll try two consecutive runs with the current wastegate preload.

- GTX PCV valve, connected to IM.
- no PCV valve, IM plugged; crankcase VTA on both sides.

I mixed up words. I've got the traditional setup of VC to IM through a catch can.

Thinking this through further. I’m going to get my WG optimized first (max spool without boost creep), then test different PCV options while keeping wastegate the same. Best way to remove it as a variable.

why not just run both IM lines to a single catch can? The easy thing to do is just VTA both through a breather filter but the better way would be running said IM ports to a catch can. Having said that I still use the PCV and a single breather on the hotside of the IM. I never liked the idea of running the hotside port to the intake pipe as that will get the compressor oily. The only upside to running the hotside to the turbo intake would be that it would help suck out the blowby gasses much faster, since running two breathers wont really evacuate the blowby gasses well and it will contaminate your oil faster this way.

I finally took the time to read this through... it’s really good. Should be required reading for anyone planning out their crankcase ventilation. TL;DR version: both vents VTA through catch can for turbo.

I got the JEGS catch can referenced by Sixshooter above. It’s well made. It’s gigantic. I need to make some room in my engine bay for this thing; probably do the Suzuki washer bottle transplant to make some room on the cold side.

After reading that CCV bible, I have proposed two different way to plumb a 3 port catch can. A single 3 port is cheaper and cleaner looking than running dual cans, but should serve the same functionality. Both of these setups are with the PCV valve removed and "B" uses an intake slash cut. VTA still leaves the crankcase with positive pressure. So, which of these would work better?

https://cimg2.ibsrv.net/gimg/www.mia...5c90550466.png

https://cimg5.ibsrv.net/gimg/www.mia...23358c427c.png

"A" is a permanent vacuum leak and boost leak. "B" ingests crankcase vapors into a boosted engine, which I'm against for detonation reasons but is the better option of the two.

Better would be leaving that line to the turbo off entirely.

I'm actually doing the "B" install now with a 3-port mishimoto can. Very nice and compact for $130 btw!


PS - it fits perfectly at the driver side front of the engine if you make bracket using the hoist hook bolt.

I was under the impression having the turbo pull air from the manifold under boost was a plus. I considered going with a vented catch can like many have mentioned, but doesn't that defeat the purpose? My car is running an FMII and pcv in place with the driver side valve cover vent going to the intake before the turbo. Car has run amazingly well even with getting a little oil into the intake. So by adding the catch can I'm removing the oil issue.

So the way I see it, you can have better air flow from the valve cover by venting to intake, but also get possible vapors. Or vent to atmosphere, but move less air.

Does that sound right?

Edit: The factory car vents both crankcase hoses into the intake. This makes me think it is an even better option than VTA.

Just finished installing the Mishimoto 3-port can. Diggin this install, hope it works as planned. Will test it tomorrow.

https://cimg9.ibsrv.net/gimg/www.mia...87fa705d59.jpg

Reading that CCV link I was also wondering if you could just run one dual inlet can when you connect the can out pre compressor like what you are doing. To me it seems like the main reason you'd run 2 cans is so one of them can act like a fresh air intake if you left the PCV in place and that cans return is post throttle body so it sees higher vacuum at low RPM/load. With them both connected to the same place in the intake before the compressor other than doubling your catch can filter area I'm not sure of any advantage to having two cans. I'm also running on 2 hrs sleep so maybe missing something.

BTW - I think someone else mentioned this but I think on my MSM factory turbo they vented the PCV side through a catch can then return that post throttle body (high vacuum at low RPM). The hot side goes to the air box (pre compressor) so it probably functions as a fresh air in at low engine speed but probably sees vacuum at higher loads or in boost when your PVC is closing.

In the middle of an engine build EFR swap and I was planning on ditching the PVC but not going VTA like you. Just wasn't sure about 1 or 2 cans so curious how yours turns out.

If anyone knows where to get a 90 degree push-in fitting to replace the pcv, please tell me. I measured the pcv gasket ID to be .536". Making me think a 3/8" to 1/2" barbed reducer won't seal properly.

Will remove it and tap if I need to, but was hoping to do this without removing the valve cover.

Edit: Going to try one of these tomorrow if Home Depot has them.
https://www.homedepot.com/p/SharkBit...2LFA/202721863

1/2” hose barb fits fine, that’s what I’m using right now.

Awesome, thanks!

I would think you would want to mount the can to the body, not the shaky-shaky engine.

The engine is full of oil, seems like that would be the best place for it. Thick steel mount has thread locker and lock washers on both ends, it isn't going anywhere. I also really like that that spot allows me to route the hoses forward and away from the hot side. There is plenty of room there for it.

I'll set the over/under at 10k miles before that bracket fatigue fractures. Bets anyone?

Really? Huh... how could I make it better then? Thicker steel, Billet?

My concern is not only fatigue. The objective is to settle the oil out from the airstream. Shaking the hell out of a container of liquid is usually done to get the liquid back into the air.

Shorter moment arm, less vibration. Put it on the chassis rather than the motor. It may not break, who knows. But I agree with DNM, I’m not a fan of running it back pre-turbo to begin with, and the vibration of the can may aerosolize the funk in there and send even more into the turbo.

Gotcha, thanks guys!

After seeing what I drained from my catch can, I decided that I will definitely NOT use the turbo to pull the vacuum. That stuff is nasty! All vents to catch can then VTA.

I'm going to catch can VTA for the following reasons:
1) Six said so. He has never steered me wrong.
2) Airplane gas engines are VTA, and they deal with more blow-by, run constantly at about 80% max power, use leaded fuel, and change oil at 50 hours. That is similar to a track car and 3000 miles, but worse due to lead.

I also change oil at 3K intervals because... Six said so. I sent my last change to Blackstone and will report results. This is all street, and with Check valve and PCV (VTA on hot side). I will get another Blackstone after 3K on the fully VTA (no air intake, only tied together output) and report that as well.

I ordered the Allstar can that Six mentioned above.

What do I not do that Six recommends.... I use Mobil 1 0W40 European. So I'm not a total groupie.

Sixshooter does not approve of the oil choice, lol. Too light on protective ep lubricants.

Speaking of oil... I've been running VTA through breather filters. The oil sure does get dirty fast. It used to be 2-3k before the oil darkened, I've run maybe 1k on this batch and it's pretty brown. The oil on the dipstick had a faint gasoline smell.

I'm not overly concerned at this point, I'll just change oil more often. Since we're on the subject, Sixshooter how we feeling about Rotella T6?

I like it so much I put it on my cereal.

My Tundra has 423,000 miles on T6 and uses no oil and the seals are still supple. My last truck had 330,000 and wasn't using oil and was clean as a whistle under the valve cover when I looked.

I would highly recommend it for a daily driver turbo Miata. I also think it's probably very good oil for a lower horsepower turbo track Miata. For a dedicated track car that does not see regular street use there are some more expensive choices that may be more specifically suited to that task.

Damn. I wish I had read this before purchasing that expensive shit...

Currently I have the hot side baffles opened up to 3/8" and the vent is drilled and tapped for a 3/8" NPT to AN-8 fitting and runs to a DIY catch can made from an old CDI enclosure. From the can it goes to another AN-8 fitting pre-turbo and it collects a fair amount of oil over the course of a track day or a long drive. I have left the collected residue out for several days to see if it separates but it does not and it seems to be 100% oil. Not sure of this is a good or bad sign. The turbo intake is sludge free.

Cold side vent has a GTX valve plumbed in the factory fashion.

^ me too. To summarize what I think we've learned:

- PCV valves, even the fancy GTX one, will leak some boost into the crankcase. It's inherent in their design.
- It's unknown how much the crankcase will be pressurized, or how much pressurizing the crankcase will cause harm in the long term.
- The PCV system will be less effective in general with turbo because it's closed at idle and closed under boost. It rarely pulls vacuum.
- When the PCV system is closed under boost, the increased crankcase pressure has to leave the hotside breather vent (only). This may exceed it's design capacity.
- A check valve is a poor substitute for a PCV valve, because it acts as a vacuum leak at idle.
- Eliminating the PCV system all together and letting both vents breathe to atmosphere is proven here at MT.net, and eliminates the concern over pressurizing the crankcase.
- Anecdotal observation of 1 indicates that the oil gets dirty quicker with this arrangement.

Excellent summary.

I would add:

If you do keep PCV because a street car spends lots of time out of boost: use the standard PCV valve and add an appropriate check valve.