Jump to content
Rozsko

High spec EJ257 misfire, cam correlation, cranking compression difference between two banks

Recommended Posts

Ahh yes, silly me. Potentially a good difference then.

 

I have the Hantek myself. Don't know whether it's me or the setup, but I cannot seem to get it to actually log. I can display in real time no problem...but cannot get it to record data.

I've a cheap USB Stingray too and it records ok. But generally the software is very poor and clumsy. I will get a Pico someday.

Share this post


Link to post
Share on other sites

Cylinder#2 cranking compression:

image.thumb.png.35c8652b80c635f0501b84c6feac835e.png

Cylinder#1 cranking comression:

image.thumb.png.5cf4656df880f0a7ee49b0d331a10d22.png

 

The standard spec in the shop manual is 142-171 psi and the minimal limit is 128psi, this tells me that there is certainly something wrong, especially considering that I have 8.5CR Mahle pistons instead of the OEM 8.2. Although I am not sure if this is casued by some cam timing issue. As you can see, there is not much happening at the bottom of the chart, and this on its own is very strange to me.

Idle compression graphs to follow.

 

 

oh and btw, the WPS signal is 100 times nicer then the cheap ebay sensor's.

I think where the 8bit scope suffers to show all the details is at the low pressure range where over time there is little change.

Share this post


Link to post
Share on other sites

Cyl#1 idle:

image.thumb.png.52a752389b88b790ab2fb80cd09b2d0e.png

Cyl#2 idle:

image.thumb.png.b1ca3265d0f7dfa4f2a1fbbc93628cd0.png

 

So, the red and blue sections are mareked based on the Cosworth spec, measured at 1mm lift.

Now to me it seems as if the exhaust would open roughly 9 degrees earlier then spec and the intake would close roughly 20 and 15 degrees earlier then spec. What I am not sure about if this could be due to the 1mm lift as obviously before the valves reach 1mm lift they already start flowing some volume as long as they leave their seat.

Otherwise I can't see anything obvious on these charts.

do you guys see anything suspicious?

 

Share this post


Link to post
Share on other sites

Its normal to see little activity during cranking. Camtiming looks mostly OK to me on the running captures and i think you are spot on that it is teh 1mm lift spec playing with your mind here. Looks like there are some  differences between sides though. Like the intake valve closes later.

Now your cranking captures... You can see that the little "dip" in the pressure after the powerstroke extends lower into "vacuum" on cylinder 1 commpared to cylinder 2. That indicates that MORE of the air trapped in the cylinder during the intakestroke has leaked out of the cylinder during the compression and powerstroke, leaving a vacuum condition in the void when the piston comes down to the BDC again. This also correlates to the lower compression numbers it reaches in that cylinder. Follow so far? Also it looks like the intake vacuum is less on average in cyl 1 than cyl 2. And there seems to be more turbulence.

Now to figure where the compression goes... use your nr 1 coil as a reference and for example stick your WPS in the intake somehow (like brakebooster outlet, or evap or wherever) and see if you can catch anything abnormaly through there. And also at the dipstick tube. and the tailpipe. See what you can see. That way you could see wheter the compression goes through the intakevalves, pistonrings or the exhaustvalves. Hooking it up to the coolant tank can also show you if its a headgasket issue.

Compression on my own built motor is quite low compared to OEM motor and specs. Its just how it is. Having them equal is more important than comparing them to stock specs. Thinks like camtiming is one thing that changes it

And yeah. Scope seems to be a bit low on resolution/samplerate

EDIT: Looking again at the cranking captures and what numbers you actually got there im not sure what they are. What compression numbers are you getting again?? Also reading your first post again i see that it was low comp on that BANK, not just one cylinder. Wonder what a capture on cyl 3 and 4 looks like now...

Also. If you havent seen this then watch it. About compression numbers captured with the WPS compared to reguular gauges

 

Share this post


Link to post
Share on other sites
4 hours ago, Steve said:

Now to figure where the compression goes... use your nr 1 coil as a reference and for example stick your WPS in the intake somehow (like brakebooster outlet, or evap or wherever) and see if you can catch anything abnormaly through there. And also at the dipstick tube. and the tailpipe. See what you can see. That way you could see wheter the compression goes through the intakevalves, pistonrings or the exhaustvalves. Hooking it up to the coolant tank can also show you if its a headgasket issue.

Thanks a lot Steve for the taking the time to take a look at this and reply.

Those are great suggestions and I will certainly do those when I get back home in 2 weeks.

4 hours ago, Steve said:

What compression numbers are you getting again??

They are 120 and 137 psi.

4 hours ago, Steve said:

Wonder what a capture on cyl 3 and 4 looks like now...

Yeah, me too. And that was always in the plan to do, but unfortunately I had only a few hours to play before I had to leave. Also the thing that I did not mention yet but makes me more and more curious to check the other two cylinders is that cyl#3 is the nosiest cylinder of all 4.

image.png.bd902c2c4ed1fc3118e011471f6b549c.png

 

One more question in my mind regarding the late intake cam on cyl#1 is that how can that be if the right cam perfectly lines up with the crank signal and TDC indicated by the pressure peak perfectly lines up with the 5th spike of the crank signal which is TDC. So if everything is in sync, then how can it be late?

image.png.5d192d8c12b90e46fecd0ea2f64100b2.png

4 hours ago, Steve said:

You can see that the little "dip" in the pressure after the powerstroke extends lower into "vacuum" on cylinder 1 commpared to cylinder 2.

the vacuum is -3.2psi on cyl#1 and -2.8psi on cyl#2

4 hours ago, Steve said:

That indicates that MORE of the air trapped in the cylinder during the intakestroke has leaked out of the cylinder during the compression and powerstroke, leaving a vacuum condition in the void when the piston comes down to the BDC again. This also correlates to the lower compression numbers it reaches in that cylinder. Follow so far?

Yes, I follow your line of thought, but I would have thought that the compression is lower becasue of the late intake (opening - I think the little bump in the descending pressure after TDC is the IVO?) closing as in this case there will be less air trapped in the cylinder to compress.

Share this post


Link to post
Share on other sites

MAYBE its not perfectly ground(?) Just speculating of course, but if its both cylinders we would have to look at what they have in common. IE cams. You dialed in the cams with a degreewheel or something? Or is it a non adjustable system where you just tighten it down and it should be good? Could be that the lobes are a bit offset from what they should have been. The WPS isnt lying so it has to be something there.

The lower compression could be as you say because of later opening. Hard to tell as the cranking waveforms dont have rulers set up. And the running ones suffers a bit from slow sample rate/resolution. Can you bump up the samplerate?. However it is the EXHAUST valves that opens there (EVO). Also a perfectly sealed chamber (there is no such thing but just as an example) should have the same amount of air going up as it is going down. So pressure should stay the same. If that made sense. Probably not other than in my head...

Share this post


Link to post
Share on other sites
22 hours ago, Steve said:

You dialed in the cams with a degreewheel or something?

No, these OE Subaru cams and are not adjustable. Exhaust is fixed, Intake is variable.

 

22 hours ago, Steve said:

Can you bump up the samplerate?

I can, but what should be the ideal? I tried already with different settings like 100k, 200k, 500k, but did not really see any difference in the smoothness or detaildness of the curve.

 

22 hours ago, Steve said:

However it is the EXHAUST valves that opens there (EVO).

I am talking about the second TDC at 360 degree. if you see that ramp down from 0 psi to -10psi, that should be sort of straight or at least one smooth straightish (if that's a word :) ) curve.  But if you look at it there is a little hump which I think is the IVO as if I measure the duration from that to the IVC event, it is roughly 226 degrees which is the total duration of the cam (although that is with the 1mm lift) as per the spec.

Share this post


Link to post
Share on other sites

So as I look at these diagrams more and more, I just realized, that the advance/retard that we see on the pressure waveforms is about 5degree.I think earlier did mention that the LH and RH cams dont exactly line up. Guess what, the difference is about 5 degree. The LH cam signal is advanced compared to the RH and it seems both the intake and exhaust cams are advanced on cyl#2. So it is not cyl#1 that is late as the RH CAM signal perfectly lines up with the crank.

 

image.thumb.png.d5e8f29eedffb3433c92d6b4f9cb276e.png

Share this post


Link to post
Share on other sites

Do you have a "known good" that tells you the cams are supposed to line up like that with each other and that part of the crank trigger?

Since its not adjustable at all there is a question about what changed compared to stock. So.... Is the block skimmed? Are the heads skimmed? Is the headgasket thickness altered. Stuff like that. Heads that are moved further apart or closer to each other will change cam timing

Oh. You are in the K range. What scope do you have again? I usually just bump it up to about 10Ms on my 4425

Share this post


Link to post
Share on other sites
18 hours ago, Steve said:

Do you have a "known good" that tells you the cams are supposed to line up like that with each other and that part of the crank trigger?

it is not a known good waveform, but from the shop manual.

image.thumb.png.1121b7f62b7401a8fac0aa2ab37dc5a3.png

18 hours ago, Steve said:

Is the block skimmed? Are the heads skimmed? Is the headgasket thickness altered.

So everything as far as engine assembly goes are brand new. Shortblock was ordered from IAG Performance, and since the block is converted to closed deck by them, I assume they skimmed it, but to be honest, I can't recall the specifics around that. Heads are also brand new. They are CNC ported Cosworth heads, so I don't think they were skimmed. Headgaskets are Cosworth 0.78mm.

Share this post


Link to post
Share on other sites

Ok, so I got home and spend some time with the car. (I will leave the interesting part last :) )

here is the cranking waveform from the crankcase through the dip stick tube, synced to cyl#4:

image.thumb.png.5ee127d7f5b3127281bce0bc07286eb1.png

Not sure if you can see anything on this, but I cant. I know it is noisy, but all vacuum pulls are roughly the same shape and size.

Cranking intake waveform, again synced to #4 cyl:

image.thumb.png.cf39bb54eceee8eb31024e03da43029c.png

not sure if got the piston char lined up right, but even if I did, I can't get my head around the information shown.

And finally #3 cylinder pressure waveform. Surprisingly, the compression is not low, but high compared to the other 3 cylinders, and the bottom does not go into vacuum.

cranking:

image.thumb.png.2c551252d0f55972dcbb4763f0ecaaec.png

Idle:

image.thumb.png.96a58bbfa7bebbf6531dfe93bfd1d159.png

Snap: I think the backpressure here quite high compared to the other three (~65psi vs ~25psi), not to mention the peak pressure which is ~450psi in #3 and ~250psi in the other three.

image.thumb.png.d7ba13f5c835bfc857e161fc436d9351.png

Now, on the backpressure: No cat, straight through 3.5" exhaust (with silencer though, so the neighbors don't scream on me)

#4 snap throttle: The reason why #4 for comparison is because it shares the uppipe tube with #3.

image.thumb.png.063e509c0e07d61539d2c0ee9b65b96b.png

 

 

image.png

Share this post


Link to post
Share on other sites

Measuring the 3 cylinders there definitely seems to be differences between cylinders when it comes to cam timing. Of course my measurements are approximations but still.

image.thumb.png.3c10c9de137026e6ea916dd3a31a9ffa.png

 

Your pressure measurements are flawed for some reason

The WPS self calibrates when you turn it on, so the issue is probably in the probecalibration in the software. Do you choose the wps in the probemenu? Do you use the automotive software? The pressure should read around 0 psi (or atmospheric pressure) at that plateau at the exhaust stroke. The shape of that plateau looks normal so i suspect its just the readings that are skewed. Or is the software actually reading 0 when the sensor is measuring atmospheric?

The manifold pressure reading... Two cylinders seem to be pulling more vacuum than the others. Also that hash on the top of them... Seem suspect.

Try setting up the WPS as a pulse sensor. Range 3 zoom 3 IIRC. That way it indicates when pressure is changing  and how fast it is changing) instead of pressure.

Share this post


Link to post
Share on other sites

I think the 4 arrows i attached in this picture points at the same point on each cylinder. From there on up there is something wrong on two of them. Also the distance between those points and the nearest TDC/BDC is different in pairs. Two of them are around 30-35 degrees (the two hashy ones) while the two others are around 20-25 degrees. Again this isnt necessarily 100% accurate but indicates differences

image.thumb.png.74dabf33d365e92fa057a6b4b87c268e.png

 

I bet all these differences equals pretty different AFR`s between cylinders

Share this post


Link to post
Share on other sites
1 hour ago, Steve said:

Your pressure measurements are flawed for some reason

Well, I will repeat the test on cyl#3 then and see again.

I am using a non-auto pico software, but I select the WPS in the probe setup, so that must not be a problem.

59 minutes ago, Steve said:

I bet all these differences equals pretty different AFR`s between cylinders

LOL, that's exactly how it all started. I tried to scale the MAF sensor as I had an APS cold air intake that is bigger diameter then stock, but the AFR readings were jumping all over the place (+/- 1.5AFR). But I actually thought that is due to the misfire, as in that case only air is passing through the engine momentarily and that is obviously read by the wideband.

Share this post


Link to post
Share on other sites

Took a little while as I took samples with multiple options, but here you go.

Everything except cyl#3 pressure is cranking. As you can see, you were absolutely right, the measurement I took yesterday was not correct.

It is interesting that the hashes and those bumps present on yesterday intake capture is gone on today's zoom3 capture.

Cyl#3 running compression is ~50psi, cranking compression is ~117psi. Back pressure on snap test is ~35psi

I did not really have the time and peace yet to think over what I can see on this picture, but on the WOT intake capture there is a major difference between two and two cylinders.

image.thumb.png.fb97c6ec27a59e55dbbd877f54f2a4f0.png

Share this post


Link to post
Share on other sites

There seems to be different resolution/sample rate between this last cyl 3 idle capture and yesterdays capture? On first look it looks like the intake opens SOONER today than i did yesterday. But they seem to close at about the same time. That must be a capture thing i think...

There seems to be less exhaust energy from cylinder 2. It makes less pressure in the exhaust compared to the others.

Looking back at page on in this thread i see that cyl 2 and 4 got higher compression than the other bank based on your relative compression test.

This last intake snap throttle capture shows that those two (2 and 4) creates more pull on the intake during their intake strokes. Sot that probably agrees to more air drawn in on that bank resulting in better compression.

I dont know man. Either your Dual AVCS system (thats what you got, right?) is playing games with you. Or the distance between heads changed and skewed timing. Or you got a cam ground wrongly. Or maybe you got one wrong cam (with different specs). Either way your cam timing seems to be different between banks. As you stated on page 1 already.

I think we got evidence enough and its time to pull the engine and get yourself a degree wheel and a dial and start measuring actual numbers i am afraid. Unless its something with your AVCS and you can verify it by disabling it. But they should be zeroed at idle according to google (?)

On the bright side those engines are relatively easy to pull at least :) I think(?)

 

Edit: you put the WPS in ZOOM 3 but not RANGE 3. Am i right?

OR valve sealing issues for some reason. If an intake valve is leaking on one cylinder it will leak pressure into the intake when its companion cylinder is on its intake stroke, resulting in a lower intake pull (less vacuum)

Share this post


Link to post
Share on other sites

As said early on....a very common cause of idle misfires etc is poor valve sealing. Although usually older engines due to valve/seat recession.

A new build should not suffer this...unless somehow valve/cam clearances were set too tight.

 

Removing the valve covers is doable in the car, and IMO should be done to measure these clearances. Either to rule it out, and/or for peace of mind.

 

Although as for cam timing, this is just a single AVCS engine ? not dual ? ie, intake only. 

 

Share this post


Link to post
Share on other sites
6 hours ago, Stevieturbo said:

Although as for cam timing, this is just a single AVCS engine ? not dual ? ie, intake only.

Correct. This has only avcs on the intake.

And you are absolutely right. Removing the valve cover and checking those clearances will be the next step.

The only question is when, as I need to go back to Italy now, so in an other couple of weeks.

 

Thanks a lot to both of you for following up on my problem so many times. I really appreciate all your insights and hints!

Share this post


Link to post
Share on other sites

Yes, not that uncommon either as the covers tend to leak and need new seals. But totally doable. Probably not that easy to measure the clearances though, but should be able to get some feeler blades in ok to get a good idea of what's going on.

 

I cant comment as to whether the live compression tests would show poor sealing....they maybe would...they probably should. But I know on cars I've seen this in the past, even a leakdown test didnt show anything, but the clearances were too tight, almost nothing in some cases.

But I've never had the benefits of a scope/in cylinder to try on those faulty cars.

 

But the symptoms strangely were usually idle misfires, but strangely seemed to affect running less at higher rpm's

Share this post


Link to post
Share on other sites

Thanks a lot guys!

I am getting prepped for pulling the engine. I checked the bay the other day and while I could pull the covers for sure, it would be a pain in the back to do all checks.

And I have ordered a few bits. Adjustable cam gears for the exhaust and LIC adjustable idlers for the intake.

I just need to find the time to do it.

20191130_153635.jpg

Share this post


Link to post
Share on other sites

Sidenote. Here is a car i had yesterday. Problem is a hydraulic lifter resulting in a short EVO duration. About where my markers are set. Care to guess what a conventional compression tester would have showed :D

image.thumb.png.f982bc7e2afd6143ead43dfb61724f44.png

Share this post


Link to post
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

Loading...

×
×
  • Create New...