16 replies to this topic

[RazerX]

Does it matter if if you install and torque the cams on the head when the head is off the car or on the car? I guess I am asking this an optimum best practice from the engine builders.

 

 

[chris haldeman]

Either is fine

[RazerX]

Okay, thanks.

[Steve Scheifler]

You might want to confirm that the cams turn freely after installing the head but before installing the HLAs/shims, even if you checked that with the head off.

[RazerX]

I actually have a lot of work to do on a particular head but was looking for best practices.  It has been a nightmare and something is still not right.  So I am going to pull and look at it.  And put on a different head... 

[Steve Scheifler]

Well, don't hesitate to ask for help from the experts here, they seem to have run out of things to talk about until the next big scandal.

[RazerX]

Well... on the invite for intrigue..

 

I have doing leak down tests for years.  I have a set methodology, engine temp, and same tools.  I am a fairly accomplished mechanic.  

 

This is the cliff notes version;

I received an used engine with claimed very little track time built by a known SM engine builder.  I did a leak / compression check and measured some very large leak down numbers(20% - 40%).  I verified intake and exhaust valve leaks, I also verified my leak down device on another Miata given the large numbers.  Engine hasn't even been to the track but I decide to sent the engine back to the builder for a refresh(head, rings, etc.).  Engine builder tells me several valves are tulip-ed and likely from over heating by the previous owner and they are replaced w/ new ones, supposedly.  

 

Engine returns and is installed.  Still very high leak down numbers(18% -30%), but made good HP 124 but not great for a pro engine.  After 5 weekends I notice increasing leak down and lower compression not surprising but the rate of degradation was.  No zings.  

 

I decided to pull the head and send to known local shop(that i trust) to see if the leaks can be sorted with valve job.  Local shop calls and asked who did the work because it is a mess (screwed up seats, valves not ground properly, two springs upside down, two valves have 45 degree notches at the top where the touch the HLA. He said someone slipped on the grinder, tried fix it and put it back but should have been binned.   He does a valve job with 2 new valves and then checks valve leakage with a vacuum tester and says it is factory tolerance or better.  I put it back on the car.  Leak downs are still insanely high (over 20% on all of them.  I take it to a test day, maybe they will seat more.  Hoping a mechanical issue will fix itself..knowing that is not going to happen  And guess what, no such luck.  So I check my leak down tester again on our street car w/ 100K and it has 9-11% so still not the tools.  

 

I call the local machinist and talk to him, could the head be wrapped?  "No, I checked it the bottom was dead flat and the cams sat in journals perfectly."  I asked if it could twist when i put it on the block.  He said very unlikely.  He can't come up with a rational reason and wants to see the head again because he is also confused and offers at no cost to make it right if it is his work.  But this is the 3rd time that block has done this, so...  

 

So what is the post about?  What the hell could be causing this?  Pro builder played with the cams?  Yes, that is an easy check the next time I have the valve cover off.  While I have now acquired another head given the problem head might be illegal given this was going on when HeadGate happened.  So I will swap the cursed head with used virgin head.  But, the engineer in me wants to know what could be causing these weird numbers and yet the car ran fine, not as much HP as I would liked but it starts, idles and sounds normal.    

 

So any ideas?  

[Steve Scheifler]

Sorry you are having problems. It is often possible to tell where the leakdown air is going, especially with such high numbers. Did you listen at the intake, exhaust and crankcase? Are you sure that most of the remaining leakdown is escaping past the valves and not the rings? At what pressure do you perform the test? (Some gauge sets use 100 psi and read directiy 1:1, others use much lower cylinder pressure).

I'm trying to visualize what you describe about the valves, especially the notches, did you take pictures?

What year engine?

[RazerX]

Yes good questions.  Most of the leak is on the intake side, minor on exhaust, and only on #4 do get any ring leakage.  

 

I run my test 100 psi, I have a picture attached of the top of the valve.  it was hard to get a good angle for the small picture.  It has nothing to do with the valve sealing, just basis for some of the lackluster quality that went into head from a pro SM engine builder.   

 

This is a 99.

 

 

 

 Valve stem 2.jpg   128.83KB   8 downloads

[chris haldeman]

Check the valve lash on the intake side. There are no hla's on the 99's and they require a valve lash around .007 too not leak
Btw there would be nothing cam shaft related and no " messing around " that would cause leak down

[Steve Scheifler]

Ugly valve. Did you say two were like that? Stock valves are stupid expensive for what they are, but putting that back in was dumb.

You initially asked about the cams so I thought maybe you had them out when doing one of the leakdowns, which eliminates any issue with valve lash etc. If you are losing 20% past the intakes then with the head off they should leak like a sieve when you turn the head on its side and fill the port with liquid.

Did you or the local shop test them with Prussian blue or a little lapping? I assume the shop confirmed the head is flat and is not being pulled flat when torqued to the block.

[FTodaro]

Interesting, but Steve had a good point do the next leak down with the cams out that will narrow the issues. Did they check if the valve guides are out of spec?

[Jim Drago]

We( as do many other top builders) do all the heads on a tq plate it is a real pain in the ass to do, feels like building a ship in a bottle sometimes.  But the leak down numbers are better if done that way.  IMO, you need to start over.. new valves, new springs with someone who knows what they are doing.  This is not rocket science.  Pull the head, bring back to the machine shop and have him vacuum test the head. If each cylinder does not completely bury the vacuum gauge you will have leak down issues.  Most of the gauges have a green area that is good /acceptable, you need to be well past that .

BTW the tulipping is caused when the any combination of these occur..When you push teh margin out too far, resurface the valve too many times ( we will only do once)  sink the valve too much and lose spring pressure.

Jim

[Steve Scheifler]

Or sink the valve within FSM spec, or have factory springs that are too short or too weak new, but are denied the FSM stipulated additional shims to correct.

Every chance I get.

Fix this stupid rule.

[RazerX]

Check the valve lash on the intake side. There are no hla's on the 99's and they require a valve lash around .007 too not leak
Btw there would be nothing cam shaft related and no " messing around " that would cause leak down

When I first got the engine IN range from (.006 - .010)  EX (.010 - .012)

After I got i back from the engine builder IN (.007 - .011) EX (.010 - .014) I was told that was fine (I am anal not happy at this point)

After local shop IN (.07-.09)  EX (.11 - .014)

 

 

Ugly valve. Did you say two were like that? Stock valves are stupid expensive for what they are, but putting that back in was dumb.

You initially asked about the cams so I thought maybe you had them out when doing one of the leakdowns, which eliminates any issue with valve lash etc. If you are losing 20% past the intakes then with the head off they should leak like a sieve when you turn the head on its side and fill the port with liquid.

Did you or the local shop test them with Prussian blue or a little lapping? I assume the shop confirmed the head is flat and is not being pulled flat when torqued to the block.

Yes, two were like that.  And a third valve was replaced as the local shop though the previous grinds/resurfacing made an edge too thin and could bend/break on the seat.  On the first two leak downs I eventually took the valve cover off and ensured there was a gap between the cam lobe and valve bucket/shim.   So i guess i just answered my self about the cam messing.  

  

I was "trusting" work in past, never did a liquid test.  Good idea for next time, is there a preferred liquid to use? 

 

I know he used a vacuum tester, not sure about the blue.  Again he said that his vacuum test was as good or better than OEM specs and i didn't ask him how he got there.  Again he is a great guy and is perplexed by this and want to see it.  He too is 'rattled' this is happening.   Yes, he said the head was dead flat when i gave it to him.  Which at that point had been 'redone' but the original engine builder and I had maybe 9 - 12 days.  

 

 

Interesting, but Steve had a good point do the next leak down with the cams out that will narrow the issues. Did they check if the valve guides are out of spec?

Valve guides were checked said to be in fine working order, valve seals replaced.

 

 

[RazerX]

We( as do many other top builders) do all the heads on a tq plate it is a real pain in the ass to do, feels like building a ship in a bottle sometimes.  But the leak down numbers are better if done that way.  IMO, you need to start over.. new valves, new springs with someone who knows what they are doing.  This is not rocket science.  Pull the head, bring back to the machine shop and have him vacuum test the head. If each cylinder does not completely bury the vacuum gauge you will have leak down issues.  Most of the gauges have a green area that is good /acceptable, you need to be well past that .

BTW the tulipping is caused when the any combination of these occur..When you push teh margin out too far, resurface the valve too many times ( we will only do once)  sink the valve too much and lose spring pressure.

Jim

Yes, this shouldn't be rocket science and there little secrets and technique engine builders have give you the little extra but in this case something is just wrong and weird. Yes it will go back and be vacuum tested again.  Thanks for the tulipping explanation.  I am always learning.  Part of what I love about racing.  

 

But it is just bugging me and I NEED to know why it is doing this.  

 

I doubt this head will be on one of my cars again and not want to put the expense into a hole.  It will likely go to someone who needs a head for a street car.  As I said starts, idle and runs find and does not make any seriously abnormal noises.  There are one or two noisy lifters that quiet when warm but still a little ticking.  But not abnormal for miatas to make these sounds.  

 

 

 

Or sink the valve within FSM spec, or have factory springs that are too short or too weak new, but are denied the FSM stipulated additional shims to correct.

Every chance I get.

Fix this stupid rule.

Steve, how often do you replace springs?  In a head that hits your shop it is likely to replace one, a few, none? Is it purely when they are out of your "spec" of what makes a good spring?  or do you replace them as a matter of practice?  

 

Thanks for your ideas.  Usually these things happen to be something simple and generally obvious and feel stupid when you realize it.  In this case I am still scratching my head.  And the last time I did the leak down, I swapped and added an extra gauge, swapped the hose, used a new o ring to seal against the head, and the results we consistent with previous findings.  So, while bad equipment would be the obvious and simple answer I keep ruling it out. 

[Steve Scheifler]

I go though a pile of them, measuring each for rate & "posture". New is no guarantee of good or within FSM spec, at least in the 1.6. I've got 25 yr old springs that are still in spec and closely matched, and newer ones that are not. I've got a set of original 02's but haven't bother to check them yet. Because the free length is different between the 1.6 original parts, those that came in crates, and what you get now, determining seat pressure and rate at max lift is more difficult than it would be. It's certainly more work than most people will do or want to pay to have done.