Help with head gasket

[James Wieler]

HI guys,
Long time no posts from me as I've been working on my '88 XR head gasket for some years now. I ported and polished the head extensively (long story there) and had it then machined, pressure tested, magnafluxed and flow tested. It has oversized valves, springs, retainers, wide grove HLA's and roller rockers all from Racer Walsh. I also ported and polished the E6 exhaust manifold, and waste gate.
The car runs but it feels like the ignition or cam timing is wrong as it just won't rev past ~3K RPM (spits and backfires). I double checked both cam and ignition timing and the marks are lined up and ignition timing is set to 13 degrees BTDC (SPOUT disconnected) it reads more like 30 degrees with the SPOUT plugged in.
The engine takes ~10 minutes to warm up to 180 degrees and make good pressure in the cooling system so I don't suspect any kind of cracked head issues leading to pressurization of the cooling system. Plus I check the compression today and it's 145 PSI across all cylinders.
I've done the head gasket before many years ago which worked out fine so I should have the skills/knowledge to pull this off, but this issue has me stumped.
I also installed a wide band EGO sensor (from AEM) and have it set to deliver a narrow signal to the ECU, but I'm thinking about putting the narrow back in just to eliminate that variable.
I have not pulled codes yet (will do tomorrow)
Here's where I need some advice, I swapped cams too as I went with roller rockers I figured I needed a steel cam as the roller rockers would eat my stock cast iron cam. But this steel cam is small 1.180 base circle, 1.400 lobe height which gives 0.220 lift, not sure what that translates to at the valve. I could put the cast iron cam back in just to eliminate that variable.
I've spent a lot of time looking for vacuum leaks, missing ground wires (like the orange wire on the turbo compressor that serves as the ground for the O2 sensor) but I haven't found anything like that.
I'm using the 1035 head gasket and ARP head studs.
It idles fine, and has good bottom end torque as I have a steep driveway that it climbs no problem. So is the small cam the mistake?
Thanks to all and Happy New Year

[john keefe]

Don't recall what the stock cam specs are, but if you've got upgraded, higher energy valve springs working against a weak cam lift, then the springs might be too much for the lift, opening too little. And/or the HLA's in that combination could be the problem. Did you load each HLA up with oil before you assembled?

If you go through the archives at Turboford.org, you'll find that non-Ford HLA's were problematic, and can cause all sorts of issues. Even the FMS anti-pump-up high performance HLA's had problems. If your TFI, coil, etc. check out good, then I'd contact Racer Walsh and see what they advise on the cam, and possible culprit. I'm sure they've come across issues like yours before, and may have a ready answer that'll save you time.

And, check the distributor retaining pin to make sure its not bent.

[thesameguy]

There are definitely some potential issues with using simulated narrow band, and from what I've read those effects are worse on older fuel injection systems which may be more based in analog tech. There was a big article I read on using it with turbo BMWs running Bosch L-Jetronic, which is very similar to EEC-IV (to the point EEC uses L-Jet VAMs..). I'm surely gonna get this wrong, but since wide band sensors are fundamentally reporting on different data and then constructing a conventional O2 sensor's 0-1v signal based on an algorithm the output looks and functions differently... for some applications, it's close enough, but L-Jet doesn't respond well to these subtle variances. Given how sensitive EEC is even to the O2 sensor's ground point, this logic makes sense to me. I can tell you my Saab 900 ran like crap using AEM's simulated output!



(Admittedly, the ECM was pretty angry at that point, so I can't entirely attribute it to the AEM, but it "knew" it wasn't getting the right data)

That said, I doubt this is your problem if the AFR is good. Even when I tried a planar (vs. correct thimbal) O2 sensor on the XR it ran ok - only minor goofiness.

John's explanation makes sense to me, although with backfiring in the mix makes me wonder about valve sealing or timing.

[James Wieler]

Don't recall what the stock cam specs are, but if you've got upgraded, higher energy valve springs working against a weak cam lift, then the springs might be too much for the lift, opening too little. And/or the HLA's in that combination could be the problem. Did you load each HLA up with oil before you assembled?

If you go through the archives at Turboford.org, you'll find that non-Ford HLA's were problematic, and can cause all sorts of issues. Even the FMS anti-pump-up high performance HLA's had problems. If your TFI, coil, etc. check out good, then I'd contact Racer Walsh and see what they advise on the cam, and possible culprit. I'm sure they've come across issues like yours before, and may have a ready answer that'll save you time.

And, check the distributor retaining pin to make sure its not bent.

Thanks John, this is very helpful. By distributor retaining pin do you mean the pin that hold the dist gear on the shaft?
James

[James Wieler]

There are definitely some potential issues with using simulated narrow band, and from what I've read those effects are worse on older fuel injection systems which may be more based in analog tech. There was a big article I read on using it with turbo BMWs running Bosch L-Jetronic, which is very similar to EEC-IV (to the point EEC uses L-Jet VAMs..). I'm surely gonna get this wrong, but since wide band sensors are fundamentally reporting on different data and then constructing a conventional O2 sensor's 0-1v signal based on an algorithm the output looks and functions differently... for some applications, it's close enough, but L-Jet doesn't respond well to these subtle variances. Given how sensitive EEC is even to the O2 sensor's ground point, this logic makes sense to me. I can tell you my Saab 900 ran like crap using AEM's simulated output!



(Admittedly, the ECM was pretty angry at that point, so I can't entirely attribute it to the AEM, but it "knew" it wasn't getting the right data)

That said, I doubt this is your problem if the AFR is good. Even when I tried a planar (vs. correct thimbal) O2 sensor on the XR it ran ok - only minor goofiness.

John's explanation makes sense to me, although with backfiring in the mix makes me wonder about valve sealing or timing.

Thanks also very helpful. It's easy to swap over to the standard O2 as all the wiring is still there.
James

[john keefe]

James, yes the distributor gear pin, and the hole in the gear. That gear has to fit snugly. Might not seem like much, but if there's any wobble in the pin, or the gear slips a little, it'll throw off timing as your revs get high. Eventually it'll shear off, or the pinion gear hole will wear oblong. If you're unlucky, as happened to me a couple times, it'll shear off in the middle of nowhere while you decide to "open 'er up." After the first time, I learned to keep a spare dizzy in the trunk, with all the components ready to go, then kinda' ballpark what seems like 13BTDC just to make it home.

[James Wieler]

Great thanks John, much appreciated
James

[James Wieler]

Thought I'd provide an update. I did a leak down test and it's leaking everywhere, intake, exhaust and rings on all 4 cylinders. I removed the roller rockers from one cylinder and repeated the test and no change. I don't get it, the head has been machined, I lapped the valves and tested their mating surface using Prussian blue stain. All the valves and seats looked great (within spec).
I put a latex glove over the exhaust tip and over the throttle body inlet during the leak down test and I can see them blowing up like balloons and I can feel the air rushing out from the oil breather. If I put in 50psi, I get only 20psi in the cylinder so like 60% leakage.
Well I guess that explains the rough running, looks like the head needs to be revisited by a professional and likely the bottom end needs doing as well.
Thanks everybody
James

[andyofcolumbusmerkur]

What if just the PCV was bad?

[James Wieler]

Uncanny, as I thought of that last night as well. I'll disconnect, plug both ends and retest
Thanks!
James

[john keefe]

Is it the same on all cylinders? 3-5% is very good; anything more than 15% usually points to repairs.

Don't know if this is gonna' help that much... With your compression @ 145psi in each cylinder, and it stays there after several revs, you could rule out rings and cylinder wall, likely not a bad head gasket, either. I'd suspect PCV too, but don't see how any crankcase pressure would build up to the point it would leak past spring-loaded, seated valves. it sure sounds like leaking past the valve seats. 3-angle job? Did you get a good, wide band of gray around the seats when you lapped, or just a thin contact line?

Kinda' weird, because if the cam and rockers are in place, then one of the other intakes should at least be starting to get off its seat, and that should be enough leakage to prevent the intake balloon over the TB orifice. On the plus side, if you somehow got all intakes seated (or you removed all the rockers), then the intake balloon is indicating there's no leaking across the other intake valves.

If it's just that cylinder which gives the balloon test results, then you might have a crack extending between valves that bleeds to both ports. Hard to miss with magnafluxing, unless you were rushed...

Hard-cap (plug) the intake, and hard-cap the test cylinder's exhaust, and put your latex glove over the adjacent cylinder's exhaust. Test again. IF you get another balloon in that second exhaust, then you also have a head gasket problem, or a crack in the head between cylinders.

[James Wieler]

OK, so I repeated the leak down test, this time with the PVC valve disconnected and plugged. The glove over the throttle body did NOT inflate and I left it for 15-20 seconds so there's no air leaking past the valves. This was done on cyl#1, with the crank pulley at TDC, valve cover was removed so I can confirm cam position. There's a ton of air coming from the crank case (through the oil return passages in the head). If I put 35psi into the cylinder (left gauge) I get 15psi in the cylinder (right gauge).
I then squirted several shots of oil into the cylinder to seal the rings and repeated the leak down test. Now if I set the regulator to 35psi, I get 28psi in the cylinder so still not great but huge improvement, again suggesting it's the rings that are leaking.

Question, can this cause the engine to run poorly? I mean rough idle, hates to rev, stalls, dies at a stop sign. Maybe if there's huge blow by? I'll try disconnecting the PCV (let it vent to atmosphere) and see if it runs any better. Not that, that's a fix, it just would add evidence that blow by is the problem.
Thanks for letting me vent (pun intended )
James

[andyofcolumbusmerkur]

What about using a smoke machine to find leaks?

[thesameguy]

Generally speaking, leak down pressure should be about 80psi... 60-100 is acceptable, but 80 is the typical number.

Was the engine warm or cold?

Gut reaction is those rings are shot, BUT it's worth noting that you can have pretty high cylinder leakage but ok compression... At the end of the day, cylinders need hold pressure only for an instance. It's only after a compression test has shown an issue we really need to rely on a leak down test to show where the issue is.

[James Wieler]

Yes I like the smoke machine idea, not sure how often I'd use it. I might just resort to soapy water.

I wanted to try the leak down test because I noticed during the compression test that the pressure didn't hold. usually when I do a compression test it would hold at 140-145psi until I released the pressure. This time it didn't, it slowly leaked away so something is not right. I think I'll repeat the compression test but seal the rings with oil and see if it still leaks.
Thanks all
James