The Merkur Encyclopedia

Component: Cylinder Head
Valves, Springs, and Gasket


photo curtisey of

OE: All the Turbo 2.3s used the same head. The OE replacement now supplied by FORD is not the same casting as the original. In fact it is inferior from a performance standpoint.If your existing head is not cracked, save it. They will be getting scarce. Thats a guarentee. If you cannot use it you can be sure that another lister will.

Note that the design of the "cam follower" is a variable ratio design, so there are no different ratio rockers for this engine.

History: All XR4's have the same basic head. 89's had valve seat inserts, but otherwise were pretty much the same. There were only a couple of thousand inserted heads made. Not very common. T Bird, Cougar, and SVO's since 85 will also work as is. It is believed that all '87 on heads had inconel exhaust valves.

The Turbo heads that had the "in line" intake manifolds (pre '85) lack 1 bolt hole to make the later manifold fit. You could machine this to work, but then the bolt hole goes into the inside of the cam cover area and can leak oil. This can be cured with loctite thread locker, so ultimately you could use it. You just need to know, and now you do.

In 1984, Ford had in the Mustang, a "D" port shrouded intake vavle (called fast burn) head that does NOT FLOW nearly the air that the "turbo head" flow. Stay away from all NON TURBO cylinder heads, they are much different in combustion chamber shape, volume, and size. This affects compression ratio (very important) and how much air flow the intake port will pass. (These heads can be used for really radical modifications only, ed)

There is a number cast in the spark plug area but they are seldom readable. If not look at the combustion chamber.The main difference is chamber volume, intake valve shouding, and compression ratio. The turbo head has 8.00:1 compression ratio with a chamber volume of 62cc. The NON turbo has a volume of the low 50's cc. All turbo heads, and many non turbo, have a "cTp" cast on the top near the #3 valve spring. In the combustion chamber, the turbo head is a "D" and the non-turbo is a "heart" (see photos below).

If you clean the spark plub holes really good, the casting number is in the rough cast portion. You can see it, but need really good light and a little luck.

The following prefixes are for Turbo heads:

E3ZE inline intake bolt pattern

E5RY new maching and slight casting changes

E7SE slight change in alloy and valve seat inserts

Another source lists the part numbers as E5ZE-EA and E6ZE-AB are turbo head casting numbers. the current part number for a reman head from ford is E7SZ-6049-AX

Differences in TURBO heads

Intake bolt holes: all 3 have the "extra hole" above #1 The extra hole began in 85 and is indicative of being efi with the new generation manifold.

cTp: all 3 have this mark, in different places. One above #3 ex, one above #3 intake. That casting ID means it was cast at Taubate Casting Plant (Ford) in Brasil. But not necessarily machined there. All Turbo heads were cast in Taubate and have a slightly larger chamber. That is the only obvious difference. When assembled of course they got different valves and springs etc. When Taubate started building the 87 Mustang N/A engine and it was transferred to Lima Engine Plant, Taubate cast and initially machined the head. It has a smaller chamber. Later when Lima machined and assembled the Taubate casting, the cast identifier was still there. I'll bet a beer though that they are in different locations. Casting plants are notorious for small changes that do not follow the print. The turbo is 59 to 62 cc in volume. The plants are pretty consistent with those volumes these days. I think the N/A was 54 to 57 cc, but I need to check.

Chamber: 2 are the same, 1 has more metal between in/exh valves opposite spark plug (kinda heart shaped)

Chamber volume: the different 1 above has a volume about 58 cc. Cracks: all are cracked at least on #2 exhaust

Aftermarket: There is only one aftermarket head for the 2.3L, it is strictly a racing part since it has different: cam, intake manifold, exhaust manifold, valves, springs, etc. It is made by Esslinger and sold by various companies including FORD Motorsport. Using this head, Rousch made 750 HP in the 1984 Trans Am Merkur (winning the title). This head is not the applicable for the street since it needs different intake and exhaust manifolds and of course is not smog legal.

Cylinder head bolts: Unless damaged you should NOT replace head bolts. The only time other than damage that replacement is necessary is when the bolt is torqued to yield condition. This is NOT the case with the XR4 engine. Torque angle is not necessairly torque to yield.

Cylinder head gaskets: (Rick Byrnes) Do not use the Ford part (gasket) for a turbocharged engine. It is not designed for the high boost of the XR4 engine and is substandard by all standards for high performance, as is the "Detroit" gasket for anything other than a bone stock turbo. Go to your local parts store and order the Fel Pro 1035 "wire ring" gasket which is being made for our engine. It has a much higher unit load at the combustion seal (fire ring or armour) and is a solid 304 SS core (body) with the latest non asbestos facing material. This is the only gasket on the market that was designed specifically for the 2.3L (XR4) turbocharged engine, and is the only gasket I would use. The gasket is more than $50 but you will be much further ahead. Clean the threads in the block, lube the bolt with engine oil and torque to 90 lb ft and performance will be fine. If you have a modified engine upgrading to studs may be in the future, but for most of us "on the street" particularly with good charge air cooling, you can get away without them. The ridgid structure of the head and production block contribute greatly to good sealing.

Fel Pro 1035 is the correct part number. I have one sitting on my desk right now. It is the wire ring encapsulated in the armour (fire ring)(bore grommet)..... This gasket is now commercially available, and works with no music wire o ring in the block to support the armour. I have used this in production based turbo engines up to 400 HP with NO failures. I think the only thing that will cause failure with this part is detonation and/or overheating, I also believe that most of you have had detonation at one time or another due to high boost pressures and questionable octane of the normal everyday fuel used.

After going thru the process and trying to use music wire o ring with the 1035 Fel Pro gasket, I need to report to everyone that I am mistaken. A music wire "O" ring WILL NOT WORK with this gasket. I made my recommendations based on the work I had done on the prototype versions of the 1035 gasket when I was working with Fel Pro, several years ago. The production gasket has an armour finished diameter of 3.930 inches with a diametric tolerance of .020". This dia provides no space between the edges of the armour for the wire. To work properly, the wire must be just outboard of the edge of the armour flange. In other words the wire grove ID must be the same dimension as the armour flange OD. So, I am sorry to have misled anyone, and I hope that no one has done machining that is not reversable. Now, coincidently, I had a queery at the same time as Kevins, that asked how to move the o ring groove, or eliminate it because someone had installed it under the armour. (BIG MISTAKE) On Aluminum heads thats easy, but most of us have iron components and it is more difficult. There are some shops that can weld iron and predict the outcome, and that would be the best method. (most expensive). Due to the heat at this location of the deck, I would not use a filler like Devcon, but perhaps silver solder or brazing might fill the grove and a skin cut to clean up and provide flatness and surface finish might be a reasonable fix. So, you cannot use the Fel Pro 1035 gasket with an "O" ring. But all is not lost. Nick Manarino has reported good success with the Detroit solid core gasket, and there is a McCord solid core gasket also on the market. These gaskets are a smaller bore, designed to cover an overbore of only .040" (1.02mm) and have clearance for wire as the gaskets I was working with originally. Esslinger also reports success with the "Marine" gasket on their 155 CID midget motor. I will investigate this gasket and report later. In fact I will try to collect a number of different head gaskets and look at alternatives that we may have. Now, the reason the Fel Pro part is such a big bore is that it was of course designed as a race part that will fit the most broad range of applications. During this time, many engine builders were building 2.8 and 3.0 Litre big bore engines for desert racers. This was the biggest market for the SVO block that I did in 1988/89. I don't know if the desert guys use it any longer, but in the southeast the NASCAR racers also use big bore 4 cylinder engines, and the SVO block could be successfully taken out to 3.9370" finished bore. (and live). Fel Pro was also working with Ed Pink, Billy Talley, Esslinger, Don Vesco, and others whom I do not know, to satisfy everyone. So, you cannot use the "O" ring, but as my original note indicated, with the steel wire ring, you do not necessarily need to support the armour. But, you do need really good clamp load. Nick has indicated to me that he has seen failures of gaskets leaking fluids. This is ususally the result of the bore being loaded very heavily and the body less so. It is a function of design and manufacture. Sometimes (like most times) use of this type of gasket requires a retorque procedure. This isnt a big deal to racers that regularily pull the cam cover to check valve lash, but may be a problem to many of you. While we really want a very high unit load for combustion seal, if it results in fluid leaks the programme is a failure. One more thing. The 1035 has a copper grommet similar to the one that I released on the last good ford gasket. It is designed to help the long term oil leaks thru the paper facing and the "wet" back half of the block. The testing we did revealed that very little load was lost due to the clamp load on the grommet, but there was "a little" (don't remember the value and due to QS 9000 the data is no more.) (can't keep any knowledge too long you know) My point is, it unloads the left rear of the gasket and #4 armour very very slightly. For my racer, I will remove the copper before installation. Actually, my oil flow to the head is external, so sealing is unimportant in that area. You will notice though that no other gasket has that copper o ring. I hope this clears up some misconseptions and misinformation on my part. Thanks, Rick Byrnes

Jeff brings up an interesting point.

Not break in, but retourqing after sitting for some time or several heat cycles. There are many different opinions relative to how and when to retorque cylinder head gaskets, and actually one scenario that precludes retorque, but most of us agree retorque is sometimes required. The reason for retorquing is that the paper facing material of the gasket does relax a very small amount that can and will affect the clamp load of the fasteners. The more susceptible the material is to reacting to heat, or having what gasket engineers call creep relaxation, the more reason there is for re torque of the fasteners. Creep relaxation increases with more heat and really is exaggerated when the engine overheats and the gasket is charred. (black areas on the exh side and/or between bores.)

PRODUCTION FASTENERS PRODUCTION TYPE GASKET Initial torque to 60 - 90 lbs ft. The torque angle was for different fasteners. Originally the fasteners were material class 12.9. Actually I think they were marked 11.9, but the fastener engineers say they were actually 12.9. The M12 bolts were changed to 10.9 when Ford went to torque angle. Torque angle typically does only one thing. It reduces the variability between fasteners and/or cylinders. It does not produce more clamp force unless it was developed by the fastener engineer to do that. (seldom) It just reduces variability and puts the fastener into an elastic deformation so it actually has a little more recovery. This is NOT TTY or Torque to Yield. The process if followed does not have a stretch value that puts the fastener in yield, It is positioned on the "knee of the stress/strain curve to avoid that. Additionally, the bolts are not long enough to perform adequately in maximum stretch. When using the 12.9/11.9 fasteners, and a regular type of torque wrench I am inclined to retorque after running the engine for two heat cycles. There are many ways of doing it and almost none are wrong.

STUDS AND NUTS (not undercut) PRODUCTION TYPE GASKET. This is pretty much the same as production bolts, but the fine threads of the stud and nut provide more clamp load for the same torque. Retorquing is still recommended. I have tested ARP studs and nuts, and found that clamp load can be dramatically increased with more torque, but would not suggest doing so. On the head side, it is possible to deform or collapse the structural "column" This is not likely with an Iron head, but I had great concerns with the aluminum alloy head that I use now. Historically I use this type of fastener gasket combination (1035 gasket) and retorque at 100 lbs ft. while using ARP's lube. (iron or alum). The same parts when used on the bottom end can be torqued to 115 lbs ft and provide more clamp load. An interesting point is that when using ARP studs, the clamp load increases for the first 5 torquings. Each time you stretch the stud, it work hardens just a little and transmits more energy. This is probably true of most fasteners, but I have only tested ARP hardware. Also the fastener engineers recommend an increase of 15 lbs ft for each incremental step in initial tightening of the head fasteners...actually for any fastener. Torque wrench needs about that much to get the bolt moving because of something called sticktion.

STUDS AND NUTS (undercut or necked down) PRODUCTION TYPE GASKET. When I say production type, I mean any solid core paper faced with a stainless (usually 321) bore grommet. Graphite gaskets perform a little differently, but the big difference is that they are an expanded metal or perforated metal core and lack the stability of the solid core. The core is SS in Fel Pro, but I dont know about others. The undercut portion allows the fastener to stretch more and have additional recovery, so retorquing is probably not necessary. I have NOT tested or used these fasteners on a 2.3 engine so I cannot recommend directly, but do know they tend to provide a more uniform clamp load and much better recovery.

Some time back I thought a couple of the gasket suppliers were working on a MLS (Multi layer steel) gasket like the 4.6 L part, but find now that its not true. It sure would be interesting to use this kind of part. They perform like the old steel shim gaskets, only WAY better. Essentially you have metal to metal contact and no creep, so clamp load stays constant for the life of the engine. The only negative side to them is they require very smooth flat and non wavy surface finish. It is difficult to do. After hunting for a while I found a shop in Detroit that did a fine job on my turbo head, when I was using gas filled "O" rings that require the same kind of finish.


If you have a pure stock engine with stock type gasket, you can torque production fasteners to 90 lbs ft and not worry about retorque. (Unless you overheat it) You do NOT need to replace fasteners. Even torque angle. Actually the 4.6 fasteners are good for at least 5 reuses (maybe more).

Modified engine. Retorque, and not just once. When we drive hard, and you guys with modified cars do drive harder, we induce more heat, particularly under boost. Typically when I tear down a race motor, the exhaust side fasteners have LESS torque. Once I measured only 50 lbs ft. SO, retorquing should be a regular maintenance item when the car is being worked out really hard.

Removing the cam cover and retorquing is cheap insurance as compared to failing the head gasket.

As for the 1035 Fel-Pro gasket, we feel it is not the best way to go with an in the car replacement. This gasket does not give much. We find a lot more cylinder distortion with this gasket. If you are rebuilding the block and hone it with the gasket and torque plate you will be fine. You should also use head studs with this gasket, and the good ones only.
Replacement gaskets are made to deal with more then just sealing. They need to work with heads and blocks that may not be flat. They need to work with less then smooth surface finishes and not perfectly clean. The 1035 Fel-Pro leave you no room for error, everything must be 100% right or it will fail. We have seen them fail on the intake side under the manifold.
If you are just doing the cylinder head I would go with the Detroit gasket. We have had better luck with this gasket then any other gasket you can buy. It will not pull the block out of round as bad as the Fel-Pro 1035 and you do not need to replace your bolts if they are good. Later.....Nick

Head Gasket Mofdifications Drilling additionacoolant holes in the head, gasket, and block will increase coolant flow in the head. My dimension appears to be .500 from the edge of the bolt hole to the c/l of the water hole. I used 1/8" at the front, 3/16 between 2 and 3, and 1/4" in therear. I chose these as as a guess. My logic being that 4 almost always runs hotter and we need as much flow back there aspossible.

Modifications: The traditional head modifications work well on the 2.3; porting and larger valves. On the subject of compression ratio, you always want to drive it lower. Original 8.0:1 downward to 7.00:1 work extremely well, so when rebuilding heads and blocks, remove as little surface material as possible, and when having a head modified, enlarge the combustion chamber, particularly on the spark plug side. Making the chamber go out to the edge of the bore in that area causes the plug to be more centrally located and is an improvement in combustion effeciency. (so the experts say).

Valve Sizing


 Dia. Circumference  % Increase Dia. Circumference  % Increase
 Stock 1.74 2.38 1.50 1.77  
 Large 1.89 2.81  8.6 % 1.59 1.99  6.0%

Part Weights

Stock Turbo Motorsport
 Retainer  32  28
 Spring  80  90
 Exhaust Valve  100  115
 Total Moving (gm)  172  188
 Retainer  32  28
 Spring  80  90
 Intake Valve  100  115
 Total Moving
 172  188

Part Numbers from Mid-West Motorsports
 Item  Qty Size Part # Price
 Valve Int 4 1.890 SIV23I $40
 Valve Exh 4 1.590 SIV23E $40
 Sprint, Retainer, Locks 16 - MWM935  $40
 Lifters 8 - FMPHT2012  $35


Valve Size Theory

Rick Byrnes
I am an advacote of large exhaust valves and standard intakes. I have not done this yet, but it makes sense if you are prepared to blow on it harder. My full tilt iron head had the bowls opened up a lot, with 1.590 exhaust and 1.890 intakes. Not too much in the runners. Just cleanup. I have lost heads by grinding too much near the manifold face. On the theory of using large exhaust valves, I look at designs of older performance engines. I saw a Rolls Royce V16 P-51 Mustang (yeah, the real one) in Jack Roush's engine shop, (yeah he ownes a war bird) and was struck by the size of the intake and exhaust valve sizes. Intake valves in this 4 valve pent roof chamber head were of approx the same size. Exhaust were about 10% smaller. At any rate, I have no real engineering data that tells me to do this, but it makes sense that if we do a good job of porting the intake, and pressurize several times above atmospheric, the port will flow. Conversely the exhaust uses only atmospheric pressure for flow (except overlap times, but turbo engines don't use much overlap. (my cam is 19 degrees). So, I would really like to do that some day, but cannot afford to be that brave. Money, time, wasted parts etc etc..... With the good performance of the typical Nick big valve head, my theories are sort of redundant. Its an interesting theory that assumes that a port will flow differently under boosted conditions rather than just atmospheric pressure. As you saw last time I wrote about this at IMON, not all agree, and feel that flow is linear, but since air is not an incompresable fluid, I think there is no linearity. But the achedemics can debate that. I have seen no data where air flow has been done by pressurizing rather than sucking.

David Godfry
The above motors all had the intake valves pretty well shrouded by the cylinder and due to combustion chamber shape. The 3.8 heads underwent serious combustion chamber modification and I used the stock valves in it. It did not make sense to me to put oversize valves in a cylinder head that was already shrouded. Low lift flow is important since the valve passes through that region twice each time it is opened. True the combustion chamber can be modified to help low lift flow with a large valve, but these same modifications will help the stock diameter valve flow even better at low valve lift.

Another important reason (for me anyway) is that even turbocharged street motors spend the majority of their live running on the N/A side of the boost gauge. High intake velocities help throttle response and drivability and well help the motor get on the turbo better. Larger valves will lower the intake charge velocity. If the motor is kept wound up, on the cam, and on the turbo this is not as important and bigger intake valves can help make more top end power, but on a street car the motor has to reach the power band every time one takes off. I do not like to push motors to make power and on a street motor I will try to optimize for throttle and boost response and drivability over maximum WOT power.

I was very pleased with the performance of the 2.3T in my XR. I had oversize exhaust valves with stock diameter stainless steel intake valves (better shape than the stockers). Of course the head and manifolds were ported. I ran a 65mm throttle body on a upper intake that had been cut apart and had the runner junction knife edged then welded back together. I did not gut the upper intake to keep the throttle response sharp. I also ran a stock cam that was advanced 5 degrees.

This motor would indicate 5 psi of boost at light throttle by 1500 rpm and would reach full boost (buzzer level) by 22-2300. There was a clutch slipping torque spike at 2800. It was a blast to drive (for a 2.3 anyway) and was very streetable. Power fell off by 5300 or so but I did not care because the car was so much fun to drive. The torque curve was very fat for a 2.3 so I typically short shifted at 5K rpm or so.

True there were probably 2.3t's that made better top end power but I doubt there were any more streetable or easier to drive.

Allan Slocum
I used to believe the "stock intake, BIG exhaust" theory. I tired it both ways and the BIG INTAKES SEEMED to be better off boost. (which makes sense) Since I do drive quite often at less than 20psi boost, the bigger intakes make a difference. My next head will have big intakes AND exhausts.

Thread Cleaning: When the head is off, or especally if it is new to YOU, cleaning the threads is pretty important. You don't want to have to take it off again if you find a broken bolt in the valve cover bolt !! Don't forget the spark plugs either.

 Hole Tap
 Intake Manifold 8 M 12.5
 Exhaust Manifold 10 M 15
 Valve Cover 6 M 10

 Front of Head

8 M 12.5 and 10 M 15

Exhaust Port Modifications

Modify at your own risk, DO NOT PORT THRU TO WATER !!

photos curtisey of

photos curtisey of

Turbo head with polished chamber

photos curtisey of IdeaFactory

photos curtisey of IdeaFactory

Intake port modifications

photos curtisey of IdeaFactory

photos curtisey of IdeaFactory