So as many of you know, I've got a fairly extensive build thread from this year which I haven't updated in a while though the car is mostly done. Part of the reason for that is that I'm still chasing a few things here and there, and one of the bigger things is an oil burning issue. The car has been burning quite a bit of oil, about a quart a week. The way it was burning lead me to believe it was piston rings (no smoke on cold starts but nice puffs when I dump on it, plus the oil is black). This is obviously very unpleasant given I just had the motor apart. I bought a leakdown tester to confirm the issue this weekend.
In researching this, I found a few anecdotal experiences of chemical soaks of the pistons overnight doing wonders for carbon stuck in piston rings, so just in case I decided to do this with seafoam. Wasn't really expecting anything from it, kinda of a might as well thing. I took 2/3 the bottle and poured it down the spark plug holes, the remaining 1/3 went in the gas tank which had about 1/8 of a tank. I let it sit overnight and for a couple hours in the morning while I wrenched on a few other things.
After getting it back together and doing what we do after seafoam I took it back apart for the leakdown. Fully expecting to see at least a few bad cylinders, I was shocked when the worst was only 11%, best was about 6%. Pics of results are below.
Now I'm having a little bit of a hard time believing the seafoam really fixed my issue but the numbers don't really lie. Plus the idle and running in general has been much better although I did also replace the intake cam position sensor. The exhaust smells less oily and I haven't spotted smoke when dumping on it. Obviously too soon to tell via consumption, but this really does seem resolved - for now.
Any other way I could be burning oil? I should note that I changed the PCV and associated hoses about a year ago so that system isn't even old.
Cylinders 1&2
3&4
5&6
@M3AMI
96 BG/Magma Lux
Mods. Lots of Mods.
I've been a believer in seafoam for years, and it looks like it did a good job cleaning out your piston rings. I use it all the time in my boat. If used properly, it's pretty good stuff, and your leak down numbers back that up. unless you have worn valve guides, the rings were more than likely your oil consumption culprit
"**if you suck at driving, it certainly could put you into a curb. Don't suck."
What were the numbers before sea foam?
Maybe I should do this! Great after numbers either way
Yea I've used it before, pouring it in the brake booster hose in the manifold and through the gas tank. We also used it to great effect to clean the pistons and block when the head was off. It always seemed to make the car run a little smoother but I always wondered if that was a placebo effect thing.
This points to it really working well, but only time will tell what consumption is like.
You know I didn't think doing a test before was necessary since I didn't think it would have that much of an effect. Definitely regret that now, before and after would've been great here lol.
And thanks! I was pretty stoked to see this. I certainly can't advise against doing it yourself, pretty cheap and easy. The only thing I'll note is that my engine was apart and clean about 3000 miles ago so the amount of carbon buildup I have is likely less than average (although maybe burning oil brought it back? dunno), just something to consider.
Sent from my SM-N920P using Tapatalk
Last edited by Foda420; 08-23-2016 at 12:21 PM.
@M3AMI
96 BG/Magma Lux
Mods. Lots of Mods.
You should be doing that test at 100psi. Your leakage is not accurate.
Who's holding the crank if both hands are on the tester?
Leakage isn't just at TDC. If you pull the cams and redo the test you can see significant leakage at the bottom of the stroke from poor machine work. You really have proven nothing having no before results to go off of. You'll find out though when you do your weekly oil level check.
It was at 100. The pressure came down once the cylinder was pressurized. Per the instructions that came with the tool, I maintained the tank at 120+ (usually closer to 140), put air to the tool and turned the set knob until the right gauge said SET. This would only happen when the left read 100psi. After that I'd connect the hose to the cylinder and get the reading you see there. Other instructions and articles I found online suggested similar methods, mostly varying with the tool used.
The car was in 5th gear with the e-brake on, crank wasn't moving, guaranteed. I can verify this because I left my wrench attached to the crank at 12:00, and it didn't move for any cylinder. More than one way to skin a cat, but I'm sure you know that.
As you basically paraphrased my previous posts, I agree that without a baseline this isn't *super* informative. But at the very least it does alleviate some of the concern I had beforehand with regard to the severity of any leak.
Regarding your comment about machine work, are you referring to the decking of the head, or the valve job/port? Because if it's the head, I would think other head gasket symptoms would have already begun to rear their head given the volume of oil I was burning. And if it's the porting I would think that would have been more apparent in the leak down results. Not really trying to debunk you, just genuinely curious what you meant.
Sent from my SM-N920P using Tapatalk
Last edited by Foda420; 08-24-2016 at 12:17 PM. Reason: Spelling
@M3AMI
96 BG/Magma Lux
Mods. Lots of Mods.
I'm quite sure the tool needs to be showing 100psi on the fill side to give an accurate leakdown side. That's what the adjustment knob on the tool is for. This is what my tool instructions say. The same when you use an air tool and pressure drops on activation. You set your psi based on the use, not a closed valve. At 100 psi you get an easy % eg; 100psi in 95psi held = 5% leakage.
I always thought the test was more accurate holding a breaker bar till peak sealing was seen. A degree or two in either direction can make more or less leakage. Doing it by hand gives a more accurate representation (again, as I understood it).
You seem to be focusing on the head leakage and not the cylinder (block) leakage. I was under the impression you had a full rebuild, but I may of read that wrong. Regardless, I'm referring to the cylinders being out of round. Leakage with the piston at the TDC is always going to be higher than at the bottom of the stroke (piston at the bottom) due to the way these motor wear over time... or poor machine work on the bores (no torque plate, incorrect measuring, etc). To test the leakage of the entire stroke, you pull the cams out closing the valves. Assuming you have good valve sealing, the results at varying positions of the stroke (bottom, middle, top) can paint a whole different picture about the health of your motor. Burning off a quart every week is significant, your TDC test doesn't equate to that type of gross loss.
For a little reference; My poorly rebuilt S52 never had great compression (165psi) and burned a quart every 1k miles (boosted 10psi). My issue was poor ring sealing due to bad machine work on the block. TDC leakage was around 20% (on one or two cyls). Cams out, bottom of the stroke it was like 60% leakage. This test was after 70k miles of dealing with it.
Last edited by Braymond141; 08-24-2016 at 02:05 PM.
I definitely get the concept of the tester. In my view this worked basically the same - it read 100 before putting pressure to the cylinder and the amount lost in the psi gauge corresponded to the percentage of leak shown. I should also note that upon disconnecting the cylinder hose the pressure gauge would snap back to 100, and I did try increasing the pressure by turning the knob while the cylinder was pressurized. All it did was make my results better. The two gauges seemed to work directly in line with each other.
Using a breaker bar very well may provide more accurate results but really we're probably splitting hairs and there's only so much I can do on my own. While my girlfriend has been a massive help many times, she was only tasked with taking pics this time.
I was definitely most concerned with leakage past the piston rings vs the head. The only reason I mentioned machine work on the head was responsive to you mentioning machine work in your post. At the time I didn't realize you were referring to BMW's unfortunate machining process when they put these together in the first place. Actually if you go back and look at one of the first paragraphs I specifically mentioned I thought it was rings since it's not exhibiting symptoms that would be indicative of a leak from the head or valves.
I did a half rebuild more or less. The top end was worked while the bottom stayed the same. A compression test is on my list of stuff to do when I get back from BMW oktoberfest. Hopefully that will give me a more complete picture combined with the oil consumption (or hopefully lack thereof).
Pulling these cams to test various cylinder heights isn't very high on my list right now, but I do appreciate that tip. Maybe one weekend when the GF is traveling and I'm bored... Question though, wouldn't those readings inherently be worse even if the rings were healthy? My understanding was that the rings provide the most complete seal at the top of the stroke, thus measurement anywhere else would inherently be different, regardless of any potential issues with the cylinder shape or rings. Knowing you went as long as you did is comforting as I'll most likely have to do the same.
Sent from my SM-N920P using Tapatalk
Last edited by Foda420; 08-24-2016 at 06:48 PM.
@M3AMI
96 BG/Magma Lux
Mods. Lots of Mods.
With the gauges you have, you must get to 100 psi *while hooked up to the cylinder* by adjusting the regulator. And yes, of course the right side will move around when the left side moves around--that's the whole point. I just borrowed a gauge from a friend and it was a bit more forgiving, because each gauge read in PSI (they appeared to just be duplicates of each other). That way if it were easier to achieve 75 psi on the left by using the regulator, you could do that by using the regulator to peg 75 exactly, and then if the right gauge read 70 psi you knew you had 7% leakage (1.0-70/75). But the gauge you have, in an attempt to save you math and do the calculation for you, must be at 100 psi on the left as you supply the air or it's going to read the incorrect percentage on the right. It's still measuring the leak-down, it's just not doing the math correctly because it's assuming (1.0-x/100), where x is whatever psi the right gauge is measuring, assuming a 100 psi input.
Current:
98 M3/4/5 Alpine/Magma
05 MazdaSpeed Miata
Sold:
00 Honda VFR
99 528iT M/T
98 M3/4/5 Arctic/Dove
94 R-package Miata
89 Honda NT650
87 325is turned SpecE30
Well then I guess I must've done it wrong, I appreciate the information from you and Braymond, always willing to learn. But like I said increasing the pressure would only serve to make my results look better which maybe isn't out of the realm of possibility...
I just drove a little over 400 miles from socal to Monterey for BMW Oktoberfest. On Wednesday night I changed the oil and filter and overfilled maybe half a quart in anticipation of burning some on the trip. After checking twice - once during a gas stop and again this morning, she has not lost a drop of oil. The dipstick still reads slightly overfilled. I thought maybe I wasn't getting a great reading when I got gas last night, but this morning kinda confirms that I didn't burn any.
At one point after cruising at 75 for about an hour I decided to test and see if it would smoke. I downshifted to 4th for the rev match and then 3rd and immediately punched it and ran out 3rd gear. In theory, this quick change in crankcase pressure after the car had been running constant for so long should've sucked oil into the chambers and left a nice puff of smoke hanging in the air from my tailpipe. This in fact happened frequently before and I was told about it by multiple friends and my girlfriend. I was watching specifically for that the entire time and saw nothing, no puff, no trail, just a disappearing highway.
I gotta say I'm feeling pretty good about this at this point. I will prob do another leakdown just to get accurate results the next time I change my spark plugs but based on the way the tool works, and how the car is running, I certainly can't see it getting much worse or suddenly showing a problem that already exists now.
@M3AMI
96 BG/Magma Lux
Mods. Lots of Mods.
Miracles do happen!
Back from the dead cause i hate wrong information.
OP, you had set the gauge correctly as per the instructions.
There's a restriction orifice in that tool, usually .04" (cheap tools tend to be .08" or none at all which is worthless).
The way the tool gauges work is based on pressure drop across that restriction based on the total allowed flow.
For a .04" diameter hole, you have an area of .00126".
If the combined "area" of leaks is equivalent to that, then you get 50% on the gauge as the flow thru the orifice and the leaks are the same.
The smaller the orifice, the more sensitive the % gauge, and thus your readings will be higher.
REAL LIFE EXAMPLES:
With a .08" orifice, i've seen 8% leak
With a .03" orifice, i've seen 20%
No longer working there just using the account for posting.
**ALL POST MADE ARE IN NO WAY AFFILIATED WITH R.A.I. MOTORSPORT OR IT'S PARTNERS.**
Bookmarks