4g63's are famous for hosing crankshaft thrust bearings. This video illustrates the process of how to check the thrust bearing clearance whether the motor's in the car or not. Of course in my case the motor's on a stand for this video. Lucky for me!
In cases where the engine is still in the car, the same procedures can be used so long as the indicator is attached to the engine block. The plunger can be set up touching either the inside of the crank pulley or by removing the clutch cover plate and contacting the flywheel.
What the thrust bearing does, is prevent the crankshaft from having lateral movement in the main bearings. If a crankshaft develops excessive movement here, clutch engagement and hydraulic problems will begin showing up, followed shortly thereafter by catastrophic failure of main bearings, rod bearings, connecting rod failures, oil pressure problems, or even broken blocks, crankshafts and rods in extreme cases. It's important that every 4g63 turbo engine is within spec on this measurement.
When the crankshaft aggressively wears through the thrust bearing developing lateral play, this is called "crankwalk". On some block castings, replacing the bearings will NOT fix the problem. An engine block that is prone to crankwalking can not be fixed. The only option in these cases is to replace the shortblock and rotating assembly with new or used parts that are stronger than the one you've unfortunately encountered. For the 2g guys, the best option for repairing this problem is to remove the 7-bolt turbo shortblock your car came with and replace it with a 6-bolt from a 89-92.5 production date turbo DSM. Non-turbo blocks CAN be used; however, the block will not have oil squirters that aim towards the back of the pistons. That stream of oil aides lubrication to the wrist pins, cylinder bores, and somewhat cools the pistons. All good things on a turbo setup. Aside from that difference, there are no other differences between the non-turbo and turbo blocks. The pistons and thus the compression ratios are different, but that's it. Oil squirters can be machined into the main galleries of a non-turbo block, but it's more trouble than it's worth unless you can't find a turbo block.
There are tons of differing theories about what causes crankwalk. Nearly all of them are plausible and logical arguments. I will not get into those debates in this video in order to focus on procedures for testing and replacement. Please feel free to google "crankwalk 4g63" and read the volumes of information available already. The arguments and gathered data are older than the Eclipse itself and in abundant supply on the internets. Magnus, RRE, VFAQ, and many other parts vendors have lengthy write-ups on their own research and development. The bottom line is that the 6-bolt shortblocks are LESS likely to suffer from this.
Next time you see someone with a video that looks like it was shot with a potato asking "does this sound like crankwalk", you can send them this video. There's a reason for every noise, rather than focus on the sound, focus on eliminating the real problem. KNOW if it's out of spec.
Blueprint 105 - Main Bearing Oil Clearances
In this episode we measure the bores for the crankshaft and calculate the
oil clearances based off of information gathered in the previous video. If
you subtract the diameter of the crankshaft from the bore diameter, you end
up with your oil clearances.
If this were an assembly with new parts, I would have also paid close
attention to bearing measurements 45° off-centerline just to make sure the
bearings aren't pinched. I would also have double-checked the clearances
using Plastigage. But what I'm doing here is just getting baselines prior
If you're doing a dry assembly like this, DO NOT ROTATE THE CRANKSHAFT.
Without oil, there is nothing preventing it from being damaged.
Calculate Your Compression Ratio
This is everything you need to do to calculate your compression ratio. No
foolin'. Every equation and process demonstrated. Find all your
variables. Know your exact compression ratio in every cylinder. This is
how you do it.
Just because your service manual says your car is 7.8:1 or 8.5:1
compression doesn't mean that it is. Whenever there are casting
irregularities, variations in piston height, parts that have been machined,
non-OE parts, or changes to your head gasket selection, your compression
ratio WILL change. It's highly probable that you're only CLOSE to spec if
you've never touched your engine at all since it was "born", and that it
doesn't MATCH spec. Even if it did, how would you know? This.
V1 Swept Volume
V2 Deck Volume
V3 Piston-to-deck clearance
V4 Piston dish cc's
V5 Head combustion chamber cc's
The ratio math:
V1+V2+V3+V4+V5 = volume of combustion chamber at BDC
V2+V3+V4+V5 = volume of combustion chamber at TDC
The ratio is...
(V1+V2+V3+V4+V5) ÷ (V2+V3+V4+V5) : (V2+V3+V4+V5) ÷ (V2+V3+V4+V5)
BDC ÷ TDC : TDC ÷ TDC
First you fill in the variables, then you calculate volumes, then you add
the volumes, then you reduce the ratio (fraction). It's that easy.
Here are your magic numbers:
0.7854 = Pi quartered to the ten thousandth
16.387 = number of cc's in a cubic inch.
If you divide any number in cc's by 16.387 it gives you inches. If you
multiply any number in cubic inches by 16.387 it gives you cc's.
Quartering pi lets you use the calculation:
BORE x BORE x STROKE x .7854 = volume of a cylinder
π x (BORE ÷ 2) x (BORE ÷ 2) x STROKE = volume of a cylinder
Either way is right. You get the same result if you calculate pi to the
ten thousandth. While I apologize for all the math, no I don't. I'm
really not sorry. You actually clicked here for it whether you realize it
or not. This is ALL the math, the tests, and the whole process to
calculate your cylinder volumes and compression individually even if you
don't know any of your variables yet. All of my numbers are present for
those who want to calculate out the last 3 cylinders out of curiosity just
to see how it affects cylinder volumes and compression ratios from one
cylinder to the next. Why would I do that for you? Why would I deprive
you of that practice?
Just assume that all 4 of my combustion chambers are 41.75 ml if you do
Clicking like share and subscribe helps a channel grow. It also motivates
me. Don't sweat the camera. It's enough to know that so many of you care
about what I'm doing here. From the bottom of my atmospheric dump, I thank
you all! This gift horse's teeth are all over the place, but he sometimes
poops gold nuggets.
PS: Use ATF for your piston dish volume tests, not alcohol. Of course
it's better just to use the spec sheet included with your pistons... but
not everyone gets that luxury. Water is just fine for head combustion
chamber tests. Dry and re-oil all parts that water touches.
Cylinder Head 105 - Valve Job Basics
Valves not sealing? Valves not bent? This is how you fix that problem.
In this video I outline the basic valve job procedure. Cleaning the
valves, cleaning the seats, cleaning the combustion chamber and lapping the
valves in to make a better seal.
Here I cover the process start-to-finish. It's the same exact process for
pretty much all non-rotary combustion engines. It takes patience and
perseverance to do this job, but anyone can do it. Reference your service
manual for measurements and service limits. Everything else that's not in
your service manual is in this video.
I apologize for not having broken busted crap to work with in this video.
It's more beneficial to all of you when bad fortune falls on me because it
gets well documented, and many people watching these videos are looking for
answers. If you have bent valves, you will discover it quickly once you
chuck one up in the drill. You'll see the face of the valve wobble around
while it spins. You'll see evidence of this damage on the valve seat. If
it's bad, you may see damage on the valve guides in the form of cracks or
missing pieces where the valve guides protrude through the head ports.
Give all that stuff a good visual inspection. ...and if you doubt yourself,
never hesitate to get a second opinion or consult a machine shop. They
will have access to expensive tools that you wont find in your average
7-Bolt Shortblock Failure - Full Diagnosis
If you are your own mechanic, there is no more important character trait
worthy of development than the ability to own your mistakes. That's where
the line is drawn between good mechanics and bad mechanics. It's not the
failures but how they deal with them that measures their ability.
In short, it's not easy to admit you did something wrong or were negligent.
But if you don't own it and talk about it, it doesn't get fixed, and
nothing positive can come from it. It was my quest to overcome my clutch
issue that lead to the creation of a video. That video is the textbook
perfect guide for how to correctly install a DSM transmission.
Crankwalk as described is caused by a casting defect. This was not a
defect. This was preventable. A lot of people would find something like
this and not tell anyone out of embarrassment. I'm not ashamed. It's my
fault. I got good use out of this engine and it was tough enough to make
it 48K miles since the last rebuild despite my abuse. I'm here to tell you
if you bought a used car that's had its clutch replaced, or if you ever pay
someone else to do it... make sure it has this bolt. It's stashed away
between the starter and the transfer case, so it's hard to see. Make sure
all of your bell housing bolts are torqued properly because fastener
problems can destroy your shortblock, clutch and transmission. If your car
fails because of a mis-aligned transmission, you have no reason to blame
It wasn't until I bought my next AWD car that I discovered there was a
smaller bolt on the other side of the block. I destroyed 3 transmissions
in the GSX first. With the damage already done to my crankshaft, I then
lost a shortblock. It's an ounce of prevention that's worth metric tons on
your bank account.
Grade 10 M8x60 bell housing bolt = MD706012. It gets 22-25'lbs of torque.
Owning my mistake permits me to learn from it through con$equence$, and
never repeat it. What good would it have done anyone else for me to learn
this lesson and not share it? That's why I'm providing this video to all
of you. Sharing it can perhaps help someone else avoid this costly
mistake. This is the final chapter for my 7-bolt, and this book is going
back on the shelf.
Here are some valuable resources if you're trying to read bearing damage:
And of course, now that I've covered the complete oil system, transmission
and driveshaft series of videos, you now have all the tools necessary to
ensure your 4g63 lasts a very long time. Whether the casting defect
exists?... or it's all caused by a bolt, or the harmonics, or whatever...
Sure, crankwalk exists and it's horrible. But with the small amount of
movement required for your crankshaft before it contacts the block isn't
far enough to make your clutch drop to the floor when you turn. You'd be
hearing woodpeckers and jackhammers on the crank long before that clutch
pedal would fall to the floor. Some people are going to hate on me for
saying that. That's fine. I believe all of the people who experienced the
clutch pedal issues had fastener problems on their bell housing.
DSMs get a bad reputation for this but we can change that. Crankwalk is
never the cause of your engine failure. Crankwalk is always a symptom of
the real problem. It's your disease that makes you deny it's your fault.
You've got the 'itis. DSM-itis.
Whenever you dig deeper, you'll discover what applied all of those thrust
loads to your crankshaft to begin with, and it's not going to be a casting
defect that moves your crank .101". Mine only went .014", but all of the
same parts failed.
PLEASE tell me in the comments if you find this bolt is missing from your
Block Preparation Part 1
Preparation for powder coating and Glyptal application. Audio track is an
original performance by Rojo Del Chocolate.
My block is being powder coated rather than painted. It's just something I
do. The GSX had it on the last block so it's getting it again.
Since the tools are so similar and the mess is the same, I'm going ahead
and preparing it for the Glyptal application as well. These 2 coatings
will require being baked separately. The powder coating is baked on at a
hotter temperature than the Glyptal, so it's going first.
The surface preparation instructions for Glyptal is as follows:
Surface to be painted should be dry and free from dirt, wax, grease, rust
and oil. Remove all grease and oil by washing surface with mineral
spirits. Wipe or scrape off all loose dirt, rust or scale.
The last sentence is what's covered in this video. The 2nd sentence
happens next (although it's already degreased), and I'll get it back from
powder coat with it in the state described in sentence #1 completed. If
following these instructions to the letter of the law.
Second and third opinions in... the main journal is fine.
You'll notice that I didn't coat the main caps, or "suitcase handles". I'm
not going to. You bang around on these installing and removing them, and I
don't want to risk chipping them once they're coated. They're below the
windage area, and there will also be an un-coated main bearing girdle down
This video covered 25 hours of actual work. Yes, I kept changing into the
same filthy clothes every shoot because I wanted it to look consistent.
You have to take your time doing this kind of work, and be VERY VERY
CAREFUL! If for some reason you're crazy enough to attempt what I do in
this video, you do so at your own risk. This is an elective treatment that
I've never done, but I am by no means the first person to do it. I'm
learning about it just like the rest of you.
6-bolt 4g63 Crankshaft Chamfer & Oil Clearances
These are some things you need to think about during your build. Some
engines don't have any chamfer on oiled journals whatsoever. All equipment
like that can benefit from at least a light chamfer like the one that's on
a stock Mitsubishi crank shown in this video.
When you Chamfer an oil passage, you create a low-pressure zone where the
edges of the oil passage lift away from the bearing as it passes over it.
The principles of fluid dynamics dictate that if there wasn't an available
substance to displace that low pressure zone (in this scenario, there is an
oil supply), cavitation might occur. If we were talking about
aerodynamics, the effect would be lift.
An extremely-advanced or leading chamfer is actually capable of sucking oil
off of a flat bearing, whereas a trailing chamfer vacuums oil out of a
gallery and does a better job of spreading it around.
The modification that was performed here is intended to increase oil flow
to the mains and the rods. It's mentioned in the video that I'm setting up
my rod oil clearances on the looser side of spec. This will decrease block
oil pressure because more oil will be able to leak past the fillets of the
crankshaft and back to the pan.
But there's another modification being performed. A balance shaft
elimination. There will be lots of debate about this in the coming videos
as that transpires, but one of the side-affects of doing a BSE is increased
oil pressure. With several internal oil holes plugged off inside the
block, I will have a spike in oil pressure. I had my chamfers cut straight
in order to offer the largest practical surface area to apply oil to the
mains and rods. My intention is to relieve some of this oil flow that
doesn't have anywhere else to go. With the added flow, the straight
chamfer is actually beneficial to the mains, allowing them to intake more
oil as well as to spread more of it on the flats below the grooved upper
The animations illustrate this completely. They were created by
yours-truly. I know the oil hole on the mains is on the wrong side. It
was too much work to fix, but they get the point across. Don't laugh at
them any harder than I did.
Turbo Elantra Bearing Failure Diagnosis
I had time to look at this thing up close. Go through the oil system, and
check out all the bearings. Looks like another good study for my oil
system series because it's the opposite problem that my GSX experienced.
High oil pressure can be remedied a number of ways, but left unchecked can
actually take a toll on your bearings. The way your engine bearings work,
the parts they suspend are supported only by an oil film layer, and flow
needs to be right in order for it to work as an actual bearing. If the oil
supply is insufficient, then it loses the ability to suspend the part
causing it to crash into the bearing surface. If oil flow is too great,
friction is increased, the flow becomes turbulent, and the oil film doesn't
form properly. High oil pressure can float and spin rod bearings, and
that's worst-case scenario.
I had several un-favorable conditions going on inside this engine and that
makes it a little bit difficult to link what my engine experienced to any
one singular thing. I think it's easier to look at it like some sort of
From sub-standard parts for how the engine components would be used, to oil
pressure, to part fatigue, to part history to abuse... this thing's got a
little bit of everything working against it and that's why it's such a
hilarious car. It was given to me with one condition. "See what this
thing will do, and see how long it goes before it breaks." My take on it
is, the parts are still less than ideal, and they've still got life left in
them. It's worth fixing. These parts are worthless as a race motor, and
normally I'd have junked 'em, but it's the Hyundai.
6&7-Bolt 4g63 Front Case & Oil Pump Rebuild
Here we disassemble, clean, inspect and rebuild both popular 4g63 front
cases. This is not difficult, you just need to know what to look for.
Something else that happens in this video is the analysis of one of the
factors that caused my 7-bolt engine to fail. It wasn't the only cause,
and we'll talk about that later, but left to its own devices and without
the other contributing factors, it would have been the only cause.
Blueprint 108 - inspect the deck
There's a reason why there are no subtitled specifications in this video
for the block. It's because they don't exist in either service manual, 1g
or 2g. You're not supposed to remove material from a block on the deck
surface because it has ill effects on parts of the combustion chamber
geometry, and alters your compression ratio. It can be done intentionally
in some cases for a desired side-affect, but if you have to deck a 4g63
head, it would be advised to use a thicker head gasket. The Mitsubishi
Multi-Layered-Steel or MLS gasket is slightly thicker than the OEM
composite gasket. Also, HKS, Power Enterprise, Cometic, and other
performance brands all make MLS gaskets that are .065 and thicker.
THERE IS ONE ERROR IN THE VIDEO. I said a block with .002" warpage is
junk. I was completely and totally wrong. While I don't wish to spread
misinformation, I don't think it's a big enough error to warrant re-editing
this video. I just wasn't paying attention. .002" warpage on a cylinder
head is the service limit before it needs machining. I meant to say
.02"... or two HUNDREDTHS (not thousandths) of an inch.
...and here's my justification...
A warped block to me is junk either way even if its minimal because your
MLS gasket will never seal unless both the head and the block are perfectly
flat. Trust your machine shop to get the values for how much is taken off,
and buy the correct thickness gasket for your machine work.
A factory head gasket (composite) is .051"
The MLS Mitsubishi gasket is available in the stock .051 and a .062"
Cometic makes gaskets up to .072"
There are some brands that go as high as .127", but I'd have thrown both
the block and head away long before then.
Cylinder Head 203 - Valve & Spring Installation
There are 2 critical values in getting your valvetrain geometry correct.
Valve install height and spring install height. On some models of cylinder
heads, getting these values is easier than it is on a 4g63 cylinder head.
On the first Glyptal video, you heard me complain about the complexity of
the casting and how hard it was to reach all the nooks and crannies while
applying that coating. The casting is very complex on a 4g63 head. There
are hydraulic galleries for the lifters elevated above the valvetrain
surface which make accessing each valve bore with precision measurement
tools very difficult. It's because of this that you need to do some math
to get these values correct.
Stuart is going to show you the process for obtaining the stem height and
spring height values on a 4g63 head. Using these numbers you can determine
other work necessary to correct the spring height value to correct seat
pressure, and ensure you have adequate valve travel for your springs to
It looks like rocket surgery, but really it's pretty simple. The ultimate
goal is to get every valve spring in as close proximity to one another as
you can, while doing your best to nail the recommended specification
PROVIDED BY THE VALVE SPRING MANUFACTURER.
Loose valve springs can result in leaky valve seats, valve bounce and
deflection that will drastically shorten the life of the valvetrain. If
valve bounce is severe, it can cause engine-killing interference with the
Tight valve springs can cause excessive valvetrain vibration generated by
the force necessary for the camshafts to push them open. On the narrow
side of the spectrum this can increase friction on the cams which can wipe
lobes and shorten their lifespan, and on the severe end in not only
increases the likelihood of wiping a cam lobe, it can lead to binding valve
springs and crashing the valvetrain.
You have to hit the sweet spot.
Valve springs specifications include several variables that help you
achieve these goals. The manufacturer rates their springs for their
installed pressure and height. They have a compression limit referred to
as valve spring bind which tells you how far you can compress them from
their installed height before the coils begin to bind and the spring stops
The valve springs used in this video are rated at 97lbs @ 1.440" installed,
and .500" lift. This means they should bind at .940", but my cams will
only generate .433" lift, giving me plenty of head room at the top (.067")
to prevent binding if they are installed correctly. One thing we found
which I wasn't expecting is they're a little on the stiff side of spec. We
measured 100lbs at 1.452", so rather than risk setting them up too tight,
that's where we set our tight specification. This decision was made
because if the rated pressure is lower than our actual measurements, this
would in theory decrease the lift specification and increase the
possibility of binding. Our install pressure ended up still higher than
spec with a barely-larger-than-spec spring installed height.
I don't consider this a defect. It is close enough within the margin of
error that it shouldn't cause any problems, and anyone doing this job right
will measure and check all of these specifications to ensure these parts
are what they say they are. That's what you watched us do. I'm confident
that this will work because the 4g63 utilizes a hydraulic self-adjusting
If the stem height is too high, it can be reduced by grinding the ends of
the valve stems to shorten them. This will have no affect on spring
installed height when the parts are assembled, however; it will change the
amount on paper that you'd need to subtract from the stem height in order
to accurately calculate spring installed height. If any of the valves have
been ground to shorten their stem height, all of the valves should be
measured separately with their retainers and keepers assembled, and that
new value subtracted from stem height individually to obtain each spring
installed height. You can't reduce this value any other way short of
replacing the valve seat.
If the valve stem height is too low, you can modify the valve seat or
machine the valve spring perches (seat or retainer) to increase the size of
the spring installed height. Another method would be to cut the valve seat
deeper to recess the valve.
In my video, we show this whole process on a brand new set of Supertech
valves. All of them are identical, and all of the retainers are new and
identical. Because of this (and yes we checked it), and because no valves
required any grinding, we only needed to use one value in our math for all
Hopefully this video clears up the process and covers the options available
for making changes if they're necessary. If you land within 3% of spec,
you've done your diligence in achieving correct valvetrain geometry.
Cylinder Head 101 - Remove Cams Rockers & Lifters
Going through a 4g63 Cylinder Head? You're on the right channel. I don't
know how many parts there will be to this series, I suppose it could go on
forever... We all have different ways of doing it, and I'm going to show
This video won't put a dent in most people's grey matter outside of
entertainment value... but I tried to keep it fun. I'd like to leave at
least a small dent. But no music in these because they're going to cover a
lot of ground and fast. I'm not doing anything difficult. This part of
this job really isn't. Most everybody pays someone else to do the
complicated stuff for them, and those people do it every day so it's still
easy unless you earn your money to pay for it from difficult and degrading
But this video's about getting started. There's some safety and
organization tips to be aware of before diving in. Stuff that could help
people whether or not they even own a Mitsubishi. Everything else is in
Hyundai Elantra 4g63 Shortblock Assembly
HOLD ON TIGHT! HERE WE GO!
We begin the blueprint and assembly on my 1992 Hyundai Elantra's
bastardized 4g63. The parts used in this are from a mash of different
brands and models outside of the typical 2.0L 4g63, but the specs and
standards I am following for its assembly are for the 2.0L DOHC.
If you want to follow along in your service manual to verify what I've done
here in this video, the processes we cover here detail pages 11C-95 through
11C-105 of the 1g Overhaul manual. I would prefer you not rip them from
the binding and throw them away, relying only on this video for
instruction... but rather use this video as a motivational guide, and as a
demonstration of the techniques involved in those sections.
You gotta do the cooking by the book.
I never had any intention of making instructional videos on this particular
car, but after it blew up I slowly realized it's actually a better case
study for how a 4g63 ticks than anything else in my driveway. There are
several reasons for this. One being that it's a mix of parts that
shouldn't be bolted together, and the other is that many of you watching my
videos aren't trying to build a 600hp engine out of aftermarket parts.
You're trying to put back together what used to be your daily driver. This
car covers those bases. Don't think for a second I won't go through this
same trouble and level of detail for the GSX. I will. When I do, having
this information in this video will give you a better understanding on how
and why I do things the way I do when I get there.
This was the shortest I could condense this video. I've never uploaded a
video this long, and I hope I never have to do it again. It took a month
to create on cutting-edge equipment, 16 hours to export, and 9 hours for
YouTube to process. My script for the voiceover is 6 times longer than the
whole script for the movie Pootie Tang. 6 times. Longer. Than a
4g63 Balance Shaft Elimination - bearing modification
This is the first part of a two part series about balance shaft elimination
on 4g series engines. This video details the bearings, the other video
will cover the front case modifications. I've already got a low-def video
of the front case mods, and I plan to re-shoot that one in HD when I'm in
the assembly phase. It's linked in the video.
The balance shafts are designed to cancel out harmonic vibrations caused by
combustion and the spinning rotating assembly. They may offer a greater
degree of comfort to the driver and passengers, but with that comfort comes
Often, when a 4g63 timing belt gives up, it's because the balance shaft
belt breaks or comes loose and takes the timing belt out with it. When
that happens, it can total your pistons, valves, damage the crankshaft,
wrist pins, timing belt tensioner and crank angle sensor. Basically, it
can total your motor. The balance shafts also have a combined weigh over
10 lbs and both are driven off the timing belt making them additional and
heavy rotating mass. If you've got a lightweight flywheel but still have
balance shafts, you have your priorities mixed up.
So here's what you do with the bearings. It's easy. You can do this at
home. You CAN do it with the motor in the car, BUT DON'T. You must enjoy
punishment to do this like that.
The end result will slightly increase your oil pressure, but usually not
enough to cause concern unless you have a full-circumference bearing turbo, ball bearing turbo--with your oil feed coming off the oil
filter housing. The head feed would be better in that case because it's
regulated at 15 PSI.
Blueprint 107 - piston-to-cylinder wall clearance
This video covers how easy it is to calculate piston-to-cylinder wall
clearance. It's too easy. This is important because too loose of a gap
and the rings won't seal properly. Too tight and the pistons will scuff
the cylinder walls, ruining the bores.
We've touched on thermal expansion several times now, and the reason it
keeps coming up is because turbo engines
achieve much higher cylinder pressures, and therefore generate more heat
than a normally aspirated combustion chamber experiences. This affects the
growth of the metal parts when they're at operating temperatures, so turbo pistons need more cylinder wall
clearance to account for this.
I will cover the ring grooves, compression and oil rings in a 200-series
video while assembling this engine with new pistons. For now, these will
just be saved for a rainy day. After all, I have a stock bore stock 4g63
engine in the Colt.