Tech:Engine/K Series/How to build a tough K motor

From Rollaclub
Jump to: navigation, search

How to build a tough K motor

by Jamie

In One Paragraph

Don't spend heaps of money, it's just a crappy old K motor. Balance everything if you can afford it. Use a 20-60 cam in a 3K, 25-65 in a 4K, 4K is better. Run 0.400" valvelift, fit heavy duty springs if you don't want the stock ones to wear out in a couple of years, use 4K valve gear. Shim the oil pump spring 3mm for more oil pressure. Run between stock and 10:1 compression. Get the distributor recurved. Fit extractors and an exhaust larger than 1.5" but not larger than 2". Fit a new timing chain and tensioner. Put something better than the stock carby on, but not too big. Don't rev it past 8 grand.


These are my opinions, observations and conclusions on how to build a tough K motor. Some people might tell you I'm full of crap, some people might agree wholeheartedly. If you think you can do better, and you want to prove me wrong with proof of back-to-back comparisons, then please by all means do, everyone likes to learn something new or how to do something better.

All prices discussed are in Australian dollars, all laws spoken of refer to Australian Design Rules. If you're overseas, check with your local registration/law guys before doing anything.

By no means do you have to do everything listed on this page. Your engine will "work" with just some things, in that it will motivate the car forward, and be better than stock. It might not be at its' total full potential, but cheap fun is what this guide is all about. Besides, it's kindof unsatisfying to be at the point where you can't make any more power, trust me. You do need to do some things, and I will tell you which as I progress.

Reasons to use a K motor

DO use a K motor if you are on a budget. These can be very cheap motors to modify for noticeable reward, and also to replace if you stuff up and kill it. That's the best thing about them, parts are cheap, and spare engines and parts can commonly be had for very little or free. However like anything, they can be expensive as well, if you let them. Keep it in perspective.

DO use a K motor if you can't be bothered engineering your car. Want 100hp for as cheap as possible? Don't want to spend the money on upgrading the brakes, wheels and tyres? Want to just bolt something in and do burnouts? Technically, there's a clause in the ADRs that says if you sufficiently increase the power of an engine to warrant it, you need to upgrade other relevant safety systems, but that doesn't stop your car passing rego. This was one motivation for me.

DO use a K motor for something different. At the time I did my motor, it seemed everyone was doing a 4age transplant, or rattling off numbers numbers letters of all these high powered engines which they see in car magazines and will never buy. Nobody remembered the poor old stock K engine. It's also pretty cool to have a noisy lumpy sounding engine, and you can really impress people when you tell them a naturally aspirated 1300cc did THAT!

DO use a K motor if you just want something that bolts in. I can't weld custom engine mounts either, so I went with the easy solution. I also couldn't be arsed converting my car to an efi fuel system, wiring in a computer and all that jazz. I also like the self-tuneability of carbs and a distributor.

DON'T use a K motor if you have plenty of money to spend. If you want the best engine for your corolla and you'll spend more than a grand on it, forget a K motor. Drop in a 4age or some other efi thing, possibly even force inducted if you want to, and you can drive it without crashing it. Take it from someone who's built a $2000 4K, I could have got the same thing (minus the lumpy idle) by buying a $850 4AGE, with alot of potential for more power as well.

DON'T use a K motor if you want to beat things on the street. About the best you'll manage is flogging crappy old Lasers and Magnas, and a stock Commodore up till about 100km/h. All your mates probably have really fast cars, and if you want one too, be prepared to either spend alot of money on your corolla, or just buy a car that's fast in the first place. Budget modified Corollas are quick, not fast.

DON'T use a K motor if you want to "drift". If you think you have the "mad skillz" like Fujiwara Takumi, or you are building a "drift pig" corolla because you're a P-plater driving professional, then forget a K motor. The best power you'll reliably get out of one cheaply is less than a bigport 4AGE makes, and nobody drifts with them.

Starting Point


Your main choices are usually either a 3K, 4K or 5K. The 3K is easy because it might be in the car now, but really they're too small to make much power, because of the different stroke. About the best you can do is have a little torquey 1200 which peaks at 6500rpm, or a revvy little engine which makes nothing down low. The extra couple of hundred dollars a 4K wrecker motor costs (and really, if you hang around Corolla people long enough you'll pick one up very cheap or free) is money well spent. However, if you do modify a 3K, about the only thing that matters is the camshaft selection, which I will cover later.

The increased stroke of a 4K and the fact that it's a very in-square motor (having 75*73 bore*stroke) makes it ideal for shifting a fairly decent amount of air. A 5K can basically be treated like an oversized 4K, which is handy because extra torque is also very nice and helps make that little bit more top end. However, 5Ks are still quite expensive ($550) from wreckers, and still a little hard to find second hand for a decent price (under $300). If you can get one, go for it, if not, a 4K is almost as good for a cheap thrash motor.

Apparently the design of the blocks suffer from harmonics above 8000rpm and crack with stunning predictability. My 4K has hit 8500rpm a fair few times and it's all good every now and then, but don't hold it there or rev it that hard every day. There are also two different types of 4K block, one has three very noticeable ribs below the manifold side welsch plugs, just like a 3K if you can have a look at one. The other has small ribs and a large flat area. The unribbed block is far weaker, if you have one of them, scrap it.

3K and 4K engines usually have flat pistons, there are some rare engines, the 4K-E and 4K-U which have dished pistons and a flatter head, with longer reach spark plugs. Because of the less valve and spark shrouding a design like this offers, these are great motors for good quality combustion. However, replacement pistons for them would be expensive, Toyota probably being the only place that would sell them, if they still do.

I haven't really even given much thought to 7K engines, they're still very expensive to buy and I don't think a 40 yearold engine design, made for a van, with a massive 87.5mm stroke, is going to be very happy revving over 6000rpm for a long time.


People go on and on about how good 3K bigport heads are. Whatever. A bigport is also well over 30 years old, and probably has a fair bit of corrosion around the water jackets, it might even have developed a crack, will strip at least one manifold or rocker thread, and probably has a few helicoils in it. They also have crappy valve retainers and stem seals, which you need to change to later 4K gear when you make power.

With the amount of money you'll spend having an old late 1960s 3K head reconditioned to a working state, you could have taken a 20 yearold 4K-C head which works perfectly fine, is less prone to cracking, and had a port job done on it which is better than a bigport casting. Don't waste your time searching high and low for the elusive mystical 3K bigport, I have two of them and take it from me, they're not that bloody good.

Forced Induction

This guide is mainly about building a naturally aspirated motor. Obviously all the reliability things would apply even moreso on a forced engine, though all my cam recommendations are probably too big. Some general rules for forced induction on K motors are: don't go too big, and don't run lots of boost. A stock K motor can handle 6psi quite happily, at 8psi their life gets reduced noticeably, and they go bang pretty often at 10psi.

A T28 is way too big, a T25 is probably good for a 5K (cheaply had from CA18DET Silvias), if you had one of the smaller engines, you'd probably be best to go for a T2 (cheaply had from N12 Exas and ET Pulsars). I don't see any reason to put the turbo on the other side of the engine on a J-pipe, there's enough room between the block and the inner guard with clever manifold design. You'd have to move the alternator to the other side but that's just a bracket or two. As far as superchargers go, an SC12 from a 4AGZE would probably be ideal. You could use modified aircon brackets and an aircon crank pulley, or modify a 4AGZE crank pulley to suit the K motor.

You'll need to work out some sort of fuel supply. I'd use a drawthru SU or sidedraft, but that's just me. Considering the small amount of boost you can run, an intercooler would probably be pretty pointless. You shouldn't use intercoolers with drawthru carbs either, it makes the front of your car like a massive petrol bomb waiting to go off.

Bottom End Mods

Block Prep

These are pretty well put together little engines. 5 bolt mains and rod bolts as big as a 350 chev. While you've got the thing apart, change all the welsch plugs and take the studs out of the timing cover and rear main carrier and use bolts, it makes servicing the engine (if you ever need to change the timing chain, for instance) so much easier. Check the bearings if you like, but if the engine was working before and hasn't done a huge amount of mileage, you should be right to just slap it back together with a new bearing and seal kit. But if it needs new stuff, it needs new stuff. Running bearings one size over isn't that bad, but two sizes and I'd scrap the crank and find another one.


One thing I will suggest, if you're going to rev an engine high (8000rpm+), then get the internals balanced. Not only will it prevent the engine from fighting against its' own inertial mass, hence making more power and having less wear, it will also transfer less vibration to the rest of the drivetrain, so the weak little K gearbox will last longer, especially at high speed. Get anything that moves balanced - crank, pulley, flywheel, rods and pistons. They don't need to be balanced together. A balanced component has equal rotating mass on it's own, as well as with other components. Balancing should set you back between $160 and $250.


One good thing about 3Ks is that they have Japanese Toyota pistons in them. Aussie-assembled 4K motors have these crappy aftermarket Repco pistons, which will crack. You can tell by looking up the skirt of the piston, one side will always have the TEP logo, the other side will have REPCO embossed in it further up on Aussie pistons. If you want aftermarket stuff, standard ACL replacements are just fine, they're actually far better quality metal than the stock pistons ever were. Retail on my set of 060 3K/4K pistons was about $390, but I paid cost which was $160, so you've got a fair bit to bargain for discount if you buy new stuff.

Nobody makes forgies specifically for a K motor, but there ARE motorbike pistons that fit, but you'll have to look all the specs up you need in the freely available ACL catalog, then trawl a bike piston catalog till you find a match. Serco, the Australian Wiseco distributor, offered to have me custom billet slugs made up for the bargain price of US$2000. No thanks.

There are also TOSCO (1960s TRD) 3K/4K pistons available, super duper rare and impossible to find, which are oversize up to 0.160". You might hear Bill Sherwood talk about the set he's got. Apart from the fact you'll have little bore wall left and will probably crack the block the first time you rev it hard, these are full race pistons with one compression ring, one oil ring and no thrust skirts, so they're near useless in a street motor which you want to last more than the run-in period. Don't bother.

Lightened Flywheel

Of all the things you can do to a motor besides air supply and camshaft, this probably changes its' characteristics the most. I had my stock 4K flywheel machined down to 7kg (I have no idea what they are stock) and I like the way the engine builds revs quickly. People say you lose torque, like the ability to climb hills, with a lightened flywheel. But the engine also has less mass to spin, so you transmit more power to the wheels. You can make a K motor pretty torquey anyway, I say it's worth it. Doing this will also make your idle more lumpy. Machining of my flywheel cost me $80.

Companies will also make you a custom alloy flywheel, you supply them a stock one, and they make an alloy center with a clutch face, and press the stock ring gear onto it from your flywheel. These things can get down to 4kg, and you'd probably need a fairly high idle (1200rpm+). Expect about $600, though we got down to $400 with a 5 or more group buy. Jun make two billet alloy flywheels for 4K motors, one is 3.7kg, one is 3.4kg. They're $990 and $1100 respectively through Speedworks Auto in Perth.


This is possibly the most important choice of parts to make. You modify the cam by having it ground, the diameter of the base circle is reduced to increase lift at the lobe. The only thing which really affects the cam profile you choose is the stroke of the crank. A cam grind plus lifter facing (so the lifters wear evenly with the new lobes) should cost about $130.

The stock cam profile is 16-50 (246 degrees adv dur), and Sprinter (3K-B) cams are 18-58 (256 degrees adv dur). The stock cam runs 0.338" inlet valve lift, and 0.356" exhaust valve lift. I can't find anything to say a Sprinter cam has more lift. Ideally you should be running at least 0.400" lift (the limit of stock springs), I've seen cams which run up to 0.432" lift without having to counterbore the springs into the head so they don't bind.

As I'll mention later, if you change to a cam any bigger than about 260 degrees advertised duration, you're going to need extractors and an exhaust, as the extra air the engine is moving is too much for the stock exhaust manifold, and you get inlet reversion, where the air-fuel fixture gets pulsed back and comes out the throat of the carby in a fine mist.


Because the 3K is so oversquare (bore is larger than stroke), there's a limit to the amount of power it can make and still be driveable down low. As I've already said, a 3K is always going to be a compromise of top end or bottom end, when compared to the potential of the larger stroke 4K crank.

Personally, I think the best cam for a street driven 3K is a Tighe 112 (20-60, 260 degrees adv dur), it gives not too much loss of torque down low, and peak power at 6500rpm, just above stock and great for unbalanced bottom ends and stock valve springs. The sprinter cam isn't bad, but it's not that good either.


This includes 5K as well, because there's no such thing as a 5K crank, they all have 4K stamped on them. A 4K has more stroke than a 3K, and can shift a noticeably larger amount of air. As an engine is basically an air pump which uses fuel to keep moving, the more air you can move, the more power you make. It's a very in-square motor (75*73 bore*stroke) so you can make torque and power at the same time. A 5K is basically a 4K with even more potential to shift air because of the larger 80.5mm bore, and it does make a difference.

Alot of people have favorite cams for the 4K, debate can get pretty intense, so I'll list all the common ones. The 20-60 described above is still fairly good, makes more torque than any other (useful) cam profile I can find, and still peaks at 6500rpm. My favorite is my CE701 (25-65 270 degrees adv dur, 0.398" lift) by Camshaft Engineering in Coorparoo, I think this gives a nice balance of LOTS of torque and good strong top end, with peak power at 7500rpm. My KE25 with this cam does over 200km/h and can drive up most hills in 5th. Tighe Cams do this profile too, the Tighe 113 has 0.400" lift, the Tighe 104 has 0.423" lift.

Rob Dixon's favorite cam is his custom Wade 169x (30-70, 280 degrees adv dur, 0.405" lift). This gives about the same horsepower as the above cam, but loses a bit of torque due to different valve timing. Stewart Ford's favorite cam is his Camtech 609 (33-64 72-32, 277/284 degrees adv dur, 0.435" lift) which is close to the other cams, but from what I can find, there's a small drop in both power and torque compared to the 25-65.

I really think a cam with anything over 270 degrees advertised duration is too big for a 4K crank. Sure, you can use a bigger cam to get them to rev harder and make more power, but you need to run higher compression and rev them past 8000rpm, which places alot more strain on the rods and the block and the motor becomes unreliable. Not to mention the fact you drop so much torque the motor is horrible to drive at midrange rpm, which is where most street driving is done, so it becomes a not-enjoyable car to drive.


If you get a cam grind, get the lifters faced too. The lobes of the cam are always ground on an angle and the end of the lifter is a slight dome, so the lifter rotates whilst it is in contact with the cam, this promotes even wear of the cam and lifter together.

Only 5Ks run hydro lifters. All the V8 boys think solid lifters are the way to go for big power. Hydraulic lifters can "pump up" at high rpm, as they're not given time to drain properly, especially if you're running a thick oil. Hydro lifters are quieter, but run zero valve clearance so you can't tune the cooling of the exhaust valves with extra valve clearance, which may be a problem in a high-heat environment (such as a high compression motor). I'd change to solids.

To change to solids in a 5K, you can use Holden red motor lifters and Datsun 1200 pushrods, and you should have just enough adjustment thread in the rockers to be able to set a useable valve clearance.

Oil Pressure

Oil both lubricates and cools the engine, so it's pretty important stuff. An engine which is drawing more air into the cylinder (different carb or cam) or whose components are under more stress (increased compression & rpm) is placing an increased load on the oil to do both of its' tasks. Increasing oil pressure is the way to put reliability back into a modified motor.

Oil pressure in a K motor is increased by shimming the relief spring on the oil pump. Take the split pin end of the relief, and shim the spring about 2-3mm (use washers or a little nut, something metal) into the spring cup (not into the pump) and put it all back together.

Shimming the spring in the oil filter block does nothing but increase the pressure at which the oil filter bypass opens, hence possibly starving the engine of oil under extreme conditions. Early KE10 engines had no oil filter bypass, so everything went through the oil filter. Stewart uses one of these.

It has been said in the past that the limit of efficiency of the stock oil pump is about 60psi, anything over that is just needlessly fatiguing the pump. Mine's about 50psi with a 2mm nut on the relief spring. It's normal for it to fall down to 15-20psi at idle. You'll notice your oil pressure will drop 5-10 psi when your oil is a bit worn out.


I'm not going to discuss brand preferences here, rather viscosity and composition.

I've found that straight-weight, mineral based oils with no synthetic friction modifiers last longer in K motors. These are not a modern, tightly built engine, and they have alot of piston blowby which pollutes the oil pretty quickly. The rings also seem to float at about 6600rpm (factory redline), so your engine will be extra hard on oil if you're revving it to 8 grand. 30W is good for an old trusty motor, it lets the components move free enough but provides enough lubrication to allow the tappets to be whisper quiet. 50W or 60W if you have a tired old rattly donk that you want to get a couple more years out of.

I've found friction modified oils (eg: 20w50) to be of limited use in a K motor. Sure, they might provide a bit extra protection when cold, but they don't last anywhere NEAR as long as a straight mineral oil when you punish them.

If you have a tightly put together engine, you can run a thinner, synthetic friction modified oil if you like. I run semi-synthetic 15w40 in my engine. I've run straight 30W and a fully synthetic 5w50 as well. I find the 15w50 gives a good balance of lubrication (judging from the noise the tappets make) and less frictional losses. The engine revs noticeably slower with the straight 30W but you can't even hear the tappets. With the 5w50 in the sump, it literally can't wait to jump off the tacho, which is great but I couldn't stand the rattle of the valvetrain.

I don't really think additives do anything to help your engine, and in some cases can harm it if you put in the wrong thing, or mixture of things, so I don't use them.

Oil Filter

Me, I have an expensive engine, so I use a genuine oil filter. I recommend you do the same, even if your motor is stock. These are the only filter which has the non-flowback valve, so your engine makes some effort to retain oil pressure after the flow stops, which is good for cold starts.

As far as aftermarket filters go, my preference is with Ryco. They seem to have good flow and filtration qualities. I'm not a big fan of anything Valvoline, and they're what's commonly on the shelf at Supercheap. Fram filters are big in America, and have started to come out here in Australia as well. From what I've heard, they used to be great back in the 70s but quality has slipped as the company entered the 21st century.


If you've got the engine apart, you're going to need a gasket kit. I like using Permatex 3 non-hardening aviation glue to seal every gasket I use, it makes pulling the thing apart again (if you ever have to) very easy, and you can re-use the gaskets if they're not too old. The only thing I don't use it on is the water pump, I use Hylomar there instead.

Aftermarket gasket kits work fine, the only genuine consumable (besides the oil filter) which I've found is worth buying is the distributor o-ring. When my distributor was leaking I tried several generic o-rings with no success. Luckily, this part only costs a few dollars from Toyota.

Timing Chain & Tensioner

You really should to put a new timing chain and tensioner on any K motor you're modifying. This keeps the valves opening where they should and ensures your motor is running to its full air potential. There is sometimes a single row chain on 4Ks, but you can steal the MUCH better double row timing chain gears off a 3K and they fit in place with no modification. As the chain wears, it usually won't break, it will just stretch and stretch until it rattles against, and eventually wears through, the timing cover.

Camshaft Gear

The cam is aligned on its' gear by a pin on the cam sticking through a hole in the gear. If you compare several cam gears from different engines, you may find they are out a little bit, as the factory machining wasn't too precise. Doug has found some that were out by as much as 7 degrees! If you want to change the gearing of the cam, you need to have some way to change the position of this hole, so the teeth of the cam gear are in a different spot in relation to the camshaft lobes when it's bolted back together. A good way to do this is to have new holes drilled at key points you might want to adjust (say +2, -2, +5 and -5 degrees), then you simply put the camshaft pin in the relevant hole. You can do this yourself if you're handy with a protractor and a drill press, or an engine builder should be able to modify one for you. When this idea came to me, I thought you might only be able to drill a few holes or you'd reduce the strength of the gear too much, however I've seen one on the internet with 7 additional holes. If you're making it yourself, it's probably worth knowing that one tooth on the camgear is 20 degrees.

Top End Mods

An engine is an air pump, the more air it moves, the more power you can make. The head is how the air gets inside the engine so there are alot of things which affect the power levels you can make. A full-house bottom end with a stock head is nowhere near its' full potential. A stock bottom end with a worked head will make more power any day (though may have reliability issues depending on what else you've done to the motor).


I wouldn't suggest running much over 10:1 in a K motor. The head design really isn't the best for promoting good combustion, and detonation sucks once it sets in. I run 10.2:1 in my motor, and it get piston shakes quite badly below 2500rpm, but I think that's more to do with my amateur head porting. One of the guys on toyotastarletracing had a 4K he didn't care about anymore, so for a laugh he shaved the head down to 11:1 and slapped it back on. He said it absolutely FLEW but didn't last very long. If you want to push the limits, work out what dynamic compression you're running and keep it below 8.4.

A good rule is to take the octane number of the fuel you'll be using, and divide it by 10. BP Ultimate has a RON of 98, so 9.8:1 is pretty ideal compression for an engine running on that. Standard unleaded is 91 RON, which is alot less compression. You can usually get away with deviating 0.2 to 0.3 compression points from this rule, if your head porting and fuel supply is up to scratch.

Head Porting

I'll write more on this at a later date, probably in its' own section. Basically, clean up the short radius, and keep it on the same radius point as the outer bowl wall. Taper the valve retainers, teardrop the bowl around them for good flow. Deshroud the valves and valve seats but retain as much quench area as possible. If you have stock-sized valves, leave the inlet port size stock, it's perfect for best airspeed. Anyone who tells you they've taken the inlet ports of their head out nice and big and has stock valves is an idiot. Oversizing the exhaust ports is probably a good idea, to create a kind of vacuum effect to suck more exhaust out of the cylinder, and more intake charge in its' place.

There is also usually a casting ridge just below the spark plugs, and it's a good idea to grind that ridge smooth to promote quicker oil flow back to the sump.

Valve Stems, Seals & Retainers

The valve seals and retainers on a 3K are a bit old and silly. They have an upside down cup on top of the valve, with an o-ring on top, and a quite tall valve stem, the combination of which is supposed to direct enough oil away from the stem to stop the inlet getting oily. The spring retainer sits in another groove below this cup. Unfortunately, if you have an engine with altered vacuum characteristics (ie: a camshaft) then you have the tendency to suck oil into the inlet via the valve stems. The higher stem also stops you from running as much lift as you can with newer stuff.

Later heads, 4K onwards, have a much better system. The stem is shorter and hence allows for more lift, and has a clip-on rubber seal which restricts the amount of oil that lubricates the valve quite well. The retainer is at the very top of the valve. To change over, you need to change the lot - stems, valves, retainers and seals - and re-cut the valve seats to suit the new valves.

Valve Seats

Valve seats made after February 1972 are made of hardened steel and are fine for use with unleaded petrol. Toyota Corporate themselves told me this.

It is beneficial to airflow to have multiple angles on the valve seats. The stock seats are just cut at one angle, so the air deflects off to the sides of the chamber, the sudden change of direction also causes it to lose velocity, which is something you don't want. Getting a three-angle valve job lets the air gradually curve around into the combustion chamber, retaining velocity for better air movement, and putting the fuel around the spark plug where it should be.

You can get as many angles cut into the seats as you have the money for, or time if you're doing it yourself. I've heard of some perfectionists who cut 7 angle valve seats, almost like the seats are round. I think it's alot of work for a diminishing return, 3 angles is fine.

Valve Springs

The most lift you can run with stock valve springs is about 0.400", after that you need heavy duty springs. Whilst stock springs will definitely hit 8000rpm without a problem, they'll de-tension fairly quickly (say within a couple of years?) and you'll need to replace them more often than a set of heavy springs. My springs were $110, compared to $80 for stock springs, I think it's worth the upgrade if you have larger-than-stock lift and revs planned.

What height aftermarket springs bind at depends on the thickness of the coil and how many of them there are. If in doubt, assemble and measure. Rob's head uses heavy Holden 202 red motor springs, but you need to bore the spring seat down more into the head so these don't bind.

The stock spring seat pressure is 70lb (77lb for bigport Sprinter heads). From memory, Stewart's running some springs which fit and are 120lb. Anything between those two numbers should be fine.

Bigger Valves

3K and 4K engines will benefit from larger valves (and the accompanying larger ports that accompany these), and they're definitely THE big restriction point on 5K motors. Realistically, any valve can be used as long as it doesn't shroud too much against the edges of the combustion chamber (hence, running oversize pistons enables you to run bigger valves more usefully than stock bore size does). This is likely to be a very expensive head job ($600+) and is beyond the scope of this article. David Vizard has some good rules for calculating ideal port size based on valve size, buy his books.

Rockers and Posts

Understand that the rocker pivots on an axis, and the valve moves vertically, hence the rocker must "wipe" across the face of the valve stem at some point whilst it moves the valve down. It's better to have this lateral movement at lower lift, as it gives less wear to the valve stem and puts less strain on the whole lot. You accomplish this by lowering the axis upon which the rockers move, which is supported by the rocker posts.

Toyota had it pretty well setup from factory, so a good rule is to have the rocker posts accurately machined down by the amount of valvelift you have added to the engine. Stock lift is about 0.340", so say you're running 0.400" lift, you need to have the rocker posts machined down about 0.060".

There are two types of rocker post, early alloy and later cast iron. The cast ones are better, as the suffer less from thermal expansion (it gets pretty hot in your engine you know) and when the posts expand, they raise the axis upon which the rockers pivot, hence open up the valve clearance more, this is bad.

Changing from bolts to studs and nuts for the rocker posts is also a good idea, make sure you use high tension stuff.

Valve Clearance

Valve clearance is there to allow the valves to cool in between combustion events (spark), and to make sure the valve closes properly and contains the combustion explosion. You need to cool the valves a little bit, the exhaust valve moreso than the inlet, as it's what sees the most heat as spent burnt gases go past it into the exhaust port. Run too little clearance and you'll cause the valve to "weld" a little bit to its' seat every time it closes, and then take seat metal off again when it opens. Do this enough times and you burn out the valve seat and need the head reconditioned. However, by running less clearance, the valve stays open for that little but longer, and you get more air in and out of the motor, and make more power. Run too much clearance and your motor just sounds like a diesel and loses power and tappet rattle annoys you while you're driving, so it's better to err on the side of caution.

The stock valve clearance (0.008" inlet / 0.012" exhaust, warm) is pretty good for any of the cams I've listed above, they're well-sized street cams, not really wild top heavy race cams that require super high compression to work, so don't be afraid to close the valves up a bit from the suggestions of the cam grinder, which will usually say between 0.010" and 0.014". There are a few ways to set your valve clearances:

Yellow Book Method

This is from the factory 3K/4K/5K engine manual, aka The Yellow Books. It's the quick and lazy method.

Set engine to #1 TDC. Crank pulley timing mark should be on 0, distributor rotor should be pointing towards spark plug 1 or 2. Set tappets (counting from the front) 1, 2, 3 and 5. Rotate engine 360 degrees, crank pulley to 0 again, distributor rotor turned 180 degrees. Set tappets 4, 6, 7 and 8.

Nine Method

I've never really gotten into doing this, takes too much time and maths for my liking :P Rotate the engine through its' range of movement, as one valve is "on the rock" of its' max lift, subtract the valve number from 9 and that's the tappet you should be setting. For example, if #8 valve is open, set clearance on valve #1. If valve #7 is open, set clearance on #2. And so on.

Stewart uses this method, and rotates the camshaft through its entire base circle and sets the clearance at the highest point. Which is a pretty good idea, as it compensates for the possibility that the base circle has been ground unevenly, and allows you to run the proper valve clearance that you want to run, instead of something that may be a few thou larger in reality.

Degreeing In a Cam

This gives the most accurate method of setting the cam timing, regardless of what clearance you're running. If doing this, it might pay to check the clearance at the base circle instead it winds up being too big/small for your liking. To degree in a cam, you'll need the specs of your cam, and to print off a degree wheel about the same diameter as the K motor crank pulley, and stick the degrees onto the pulley. Alternatively, you can buy a degree wheel at a speed shop.

A 25-65 cam opens the inlet 25 degrees before top dead center (BTDC), closes the inlet valve 65 degrees after bottom dead center (ABDC), opens the exhaust valve 65 degrees before bottom dead center (BBDC) and closes the exhaust valve 25 degrees after top dead center (ATDC). Start at 0 degrees TDC and rotate the engine through its' range of movement. When you get to the timing when a valve should open or close, based on the position of the piston in relation to TDC or BDC, set the clearance on that valve to 0.001".

Spark Plugs

Stock spark plugs are NGK BP5ES or Denso W16EP. These are what you should use in a stock engine.

I find with my higher compression and what not, my engine gets some weird thermal thing going with stock heat range plugs, and my exhaust tappets would unwind and over double in clearance, in the short driving distance of about 400km. So I've changed to one step colder NGK BP6EY (aka Denso W20EP) and it fixed up.

As an engine gets really old and starts blowing a bit of smoke, you might even find a hotter plug (NGK BP4ES) gives it a bit of oomf back, however do realise that the bottom end is approaching its' wear limit and it's time to start thinking about a refresh or new motor.

I don't really think fitting a $100 set of "platinum" spark plugs or similar is of any benefit on a K motor. Maybe in a very high compression EFI application where air-fuel ratios are fine tuned (such as a 10.3:1 100kw 4AGE where they come stock) but not in a crude little carburetted thing like a 4K. If you try some and they make a huge difference, let me know.

Head Gasket

A standard type head gasket will work fine. If you have oversized pistons, then you need an oversized head gasket to suit. I'm using a graphite head gasket, which doesn't use sealant of any kind. Graphite acts as kind of a lubrication between the two different metals on either side of it (alloy and cast iron) as they expand at different rates because of engine heat. A graphite gasket is a little thicker than a normal head gasket. Don't make your gasket too thick or you'll destroy valuable quench area.

Things like running a huge copper head gasket or doubling up head gaskets to lower compression are not good ideas, there are different and better ways to lower compression, such as porting the combustion chamber, modifying the pistons, or using a different head or pistons altogether, if possible.



There are a few types of distributors for these motors, which are covered elsewhere in the Wiki. I don't really have a problem with points, but some people do, so change to an electronic distributor if you want to.

What you really should do however, is get the distributor rebushed and recurved. Rebushing means to recondition the distributor so it has no play in the shaft, and gives reliably timed sparks. Recurving refers to modifying the advance weights and springs and sliders inside the distributor to change the way the distributor advances the spark as engine rpm increases.

Having the spark at an ideal point is important. If you ignite too early, you're exploding fuel which is working against the piston still rising to compress air, and wasting power. If you ignite too late, you're exploding fuel which is "chasing" a piston down the bore and not providing as much energy as it could to move the motor, and wasting power. It's said that a modified motor generally needs less overall advance than stock, but a quicker advance curve.

I sent my Bosch-type distributor down to Performance Ignition Services in Victoria who charged me $160 to reco the dizzy to new, and curve it to suit my cam timing, compression ratio and fuel supply.

Base ignition timing is 8 degrees BTDC. I got my distributor back with a sticker "set to 10' BTDC" on it, so I do that. I've found any more than 12 to be useless. It makes the motor quicker down low, but restricts the power the car makes at highway speed. When I set my stock 3K to 20 degrees, it launched like crazy around town, but couldn't get over 105km/h, as the spark was too advanced.

I think a good rule is to set the timing to suit the octane of the fuel. Stock is 8 degrees on 96 RON leaded petrol, so 10 degrees on 98 RON BP Ultimate works pretty well. In the same fashion, you should reduce your timing if you're running 92 RON regular unleaded, maybe to 4 degrees, maybe 6, try it and see.


Having good spark is important, as it helps to make a better burn in the combustion chamber, so you make more power for the amount of air-fuel you have in the cylinder.

Fitting a decent ignition coil is a good way to get better sparks. An efi coil would probably be a good idea, or a performance oil-based (round cylinder) coil. I don't really like Bosch parts so I won't recommend a GT40, but I'm running an Echlin GX80 performance coil and it's noticeably better than stock. It cost about $80.

Upgrading the ignition leads is a good idea, especially if your leads are the old stock ones. Over time, leads increase in resistance, and this reduces the power of your spark. I've had old stock leads that won't rev over 3000rpm, they just break down inside. The lowest-resistance pre-made leads I could find were, surprisingly, Repco 8mm leads. You'd probably get a set for $40? I'm running custom 10mm Topgun leads, I got them through a shop in town and Topgun made them up for me for $80.

Fitting a spark box (high energy ignition or capacitor discharge ignition) is the best way to get big sparks, though it needs the above two items to work to its' full potential. These electronic boxes usually use the points as just a trigger, instead of passing the spark through them like the stock system, so your points don't suffer from bounce as easily, and don't wear anywhere NEAR as badly. Jaycar make a HEI kit which you solder together yourself, it costs $50. Doug and Stewart have these and they go quite well. They don't work with GT40 coils either, another reason to stay away from them. A capacitor discharge ignition kit makes huge sparks, and usually multiple sparks. Dick Smith used to make a CDI kit similar to Jaycar's HEI (they're all from Silicon Chip magazine) but nobody makes it anymore, you can still buy the PCB from RCS Radio and get the article from Silicon Chip and buy the component parts yourself. And MSD is probably the most well-known CDI system, and Victorian Performance Warehouse do an MSD 6A and Blaster II coil for $440. I want to get one of these one day.

When you get either of these kits, you could probably gap your plugs out from 0.8mm to 1.0mm and get even bigger sparks without any adverse side effects. You could probably go even further to 1.2mm with an MSD.

Water Pump

If you have a water pump that's working and doesn't look too rusty inside, keep using it. If you have one that's been sitting for a while, probably best to ditch it - the bearings don't like sitting for a long time and the seal must harden or something - they usually fail pretty quickly when put back on an engine. I bought a new water pump for my engine, it was between $90 and $140, I forget. There are several different types for K motors, the only difference being the fittings and sizes. Try get one that matches your current pump, or the hoses in your car if the current pump could be improved on. I haven't found any problems revving water pumps hard.

Other Cooling

The stock radiator will cool a modified engine if it's working well. However these cars are getting pretty old and it's not uncommon for the radiator to be in pretty shoddy condition, my KE35 one had one whole core of fins and 1/3rd of the other core missing. I got it re-cored with a 3-row center and it works great, only ever goes above half when I idle for a while on a hot day. It cost $230.

Make sure you have a thermostat installed or you'll cavitate the top housing. I don't really think removing the thermostat and reinstalling the empty shell is a good way to solve cooling problems either. Drilling the thermostat is also of dubious value.

Early Corollas had metal cooling fans, later ones had plastic fans, and the last KE70s had clutch fans.

The metal fans are dangerous and you shouldn't use them, they're all getting really old and under stress, and like to shatter at high rpm. Alex related to me a story of his old KE10 which he was in the engine bay of, revving the throttle wheel, when something went BANG next to him. The fan had shattered, and a blade had whooshed past his head and gone straight through the wall of the garage and continued on outside. If it happened while he was driving, it would have munted the bonnet pretty bad, and possibly harmed him or a pedestrian. Don't use them.

I like the engine driven plastic fan, it moves alot of air even without a shroud. You should fit a shroud, and the engine tray, to help create good vacuum to draw air through the radiator and through to the back of the engine bay faster. The shroud is positioned best when the blades of the fan are half in and half out of it, looking at it from the top. I wouldn't use a clutch fan, just something else to break, but if you have one and it works and you can't be bothered getting a new fan then just leave it.

People fit thermofans sometimes, which is a good idea if done right. Less reciprocating mass the engine has to spin, which is the fan AND the resistance of the air it's trying to move, gives an effect similar to lightening the flywheel. A thermofan needs to be mounted on brackets which reach out to the side of the radiator, NOT bolts through the fins of the radiator with rubber washers. It also needs to have a shroud made up, which covers the whole face of the radiator and just has an opening where the circular frame of the fan blades is, or the fan basically just sucks around the circle and you don't use half of your radiator area.


There are two types of stock alternator. The Denso one is of smaller amperage, but seems to handle big revs alot better. The Bosch one fitted to Australian KE55-onwards Corollas makes 45A but old stock ones don't seem to be as hardy at 8000rpm. You can underdrive the alternator with a larger Sigma pulley, or a reconditioned alternator usually works fine, maybe the bearings wear out quickly?

The good thing about Bosch alternators is that you can upgrade them very cheaply and easily, by rebuilding parts from a higher amp Bosch inside the Corolla housing, you'll just need longer stator bolts. That's too hard for me and requires good soldering skills, so an auto elec should be able to build you a 60A alternator for about $90. This is what I have now.

The stock alternator will work fine with the stock car, but if you're adding headlights and stereo and big ignition you probably need more amps, especially in a KE10-KE30 with a tired little 35A Denso. There are upgrades available to take a Bosch alternator as high as 160A, this is what alot of large stereo cars have.

Engine Mounts

Because you can make these engines pretty torquey, they'll bring out any weaknesses in engine mounts pretty quick. Usually the manifold side goes, as it's exposed to the heat of the exhaust all the time, and it gets stretched under engine rotation, not compressed like the distributor side mount does. I'm onto my 3rd one in 7000 miles, but I'm using almost 40 yearold KE10 mounts so it's to be expected.

A good way to fix this problem if you have it, is to make up a mount from the stock piece of metal which bolts to the block, and a Ford Escort or Cortina rubber, which is basically a thick rubber disc with two metal bolt threads in it. You may have to weld a piece of metal plate to the stock mount to space it out enough, (weld it to the face that sees the rubber, not the block) but it works and works well. Ford mounts aren't dear from Repco, and usually wreckers have a big container full of the things you can fish thru and get for a few dollars.

You can still get new stock-style mounts for Corollas, but they're like $130 a piece. And that's aftermarket, Toyota probably don't even make them anymore.


Before I owned a Corolla, I'd never even changed a spark plug. I don't really have the money (or the inclination to spend the money) to pay someone else to build my engine for me, and I like to try different things for myself and have fun learning about the internal combustion engine and how it responds to changes in running conditions. So I built the motor myself, with no prior knowledge other than reading books and talking to people and pulling dead old free motors apart and trying to put them back together again. K motors aren't hard to work on, they're actually very simple compared to other cars I've had the pleasure (or pain) of working on.

I also take pride in the fact that I can look at my car and think of every bit that I've had down to nut and bolt components, and think "Yep, got that sussed if anything goes wrong". And it makes the repair bills SO much cheaper when something does break. And these cars are getting 20-30 years old or more, so things break and wear out all the time.

However, I understand not everybody's that mechanically inclined, so if that doesn't inspire you to have a crack and slap the old girl together yourself, then you motor will be getting built by someone else. Hopefully it's a mate and it will only cost you a carton. Generally, over half of the cost of a professional engine build is in labour, so think of what you're paying for parts and double it.

The only things I don't do myself are jobs that require a workshop full of expensive machining equipment I don't own, or stuff which I could do, but by the time I buy the tools it's cheaper just to get it done anyway. Putting a head together is a good example, considering how expensive a valve spring compressor is, and how much of a pain in the ass valve seat lapping tools are, I really don't mind paying someone the money to use their professional and precise machine shop to do this for me.


If you've never built an engine before, you're going to need some tools - torque wrench, feeler gauges, measuring calipers, metric socket and spanner set to name a few - and a nice clean work area. You'll need an engine stand. Lubricate any moving metal-on-metal part such as bearing surfaces, lifters, etc in graphite grease (aka: engine assembly grease) before putting it together. The only exception to this is the piston rings, just wipe the bore down with a light coat of oil before you put the pistons in. Buy a Gregories or Haynes manual as well, so you know what goes where, in what order, and what torque setting to use. I personally find the Toyota factory service manual (The Yellow Book) to be most useful, but they're sometimes hard to find. First time you start the motor, wind it over without spark until the oil pressure light goes off. You'll probably flatten at least one battery doing this.

Run In

There are alot of different ways to run in a motor, and it seems to be a bit of a "black art". The basic things to remember are that a cold engine doesn't wear much, and a hot engine doesn't wear much, but an engine between 20c and 80c does about 90% of its entire wear between these two temperatures as it's warming up. At the start, you want the engine to wear a bit, so the rings bed in and seal properly, and everything frees up and becomes happy with its new surroundings. Some things take a long time to bed in, such as the rings, some things don't take as long. For example, the big end bearings and crank bearings will be worn in the first time you start the motor. Here's how I did mine.

The first time I started the engine, I let it idle to operating temperature, giving a few revs a couple of times a minute to keep oil pressure up, then turned it off and let it go dead cold. The next time, I drove the car for about 5 minutes, turned it off and let it go dead cold. The time after that was 10 minutes, then 20 minutes, then half an hour, then an hour, then I didn't worry about temperature cycling it so much.

I ran running in oil for the first couple of hundred km or so. This is a fairly thin oil with lots of graphite particles in it. The graphite lubricates the tight metal parts as they try to rub against each other while the clearances are still really small. The oil is thin, so you shouldn't rev too hard with it in, as you can squish thin oil out of the bearing gaps, and the oil can wipe away at high component speed, leaving a fairly bare surface with chunks of graphite and metal on metal which will score surfaces which should be lubricated. After that, I changed the filter and swapped to a straight 30W oil, and kept this for a couple of thousand miles, then renewed the oil and filter with the same, and drove it for another couple thousand miles. I then changed the filter and started off with the 5w synthetic oil, but later changed to the 15w semi-synth as described in the oil section.

I tried to keep a sensible but gradually increasing rev ceiling. I took it to 6000rpm on the first day I had it, then 7000rpm on the next day, then I couldn't resist an 8000rpm test run the next day. I didn't flog the arse out of it from then on, but I kept it below 7000rpm until I got rid of the running in oil. Once I'd changed to the synth oil, I was pretty confident the motor was going to hold together, so 8000rpm became more of a regular occurrence. I didn't hold it flat and reach the 8500rpm limit of the stock carb until it was run in fully.

A K motor is an old design, and uses fairly high tension rings. At the start of the engine's life, you want to put the engine under large vacuum conditions, so the rings "suck out" against the bore walls and wear themselves into the cylinder hone, so they provide a better seal. The way to do this is to get a long section of road with no stops, and drive along at normal town speed, shift into 4th or 5th and put your foot down, letting the revs slowly increase as the engine sucks air in. Hills are really good for this, as you don't end up going too fast and speeding. Do this sort of thing as much as you can during the first 500 miles, as this is when the most ring wear occurs. Don't unnecessarily labour the engine, but don't just cruise with it either. The point is to be constantly changing the vacuum conditions of the engine. Don't take it on a long steady drive, what you want to do is constantly vary your speed and gear, so you have to put your foot down, and that makes the rings suck out and wear in better.

After about 1000 miles, I just drove the thing normally, there's not much you can do but wait, and you'll know when the engine is fully run in, it will get alot more powerful for no reason. Mine took 4 or 5000 miles. Run in period over.

There's also no point trying to tune a new fuel supply on a fresh engine, its' needs for air and fuel and vacuum are constantly changing, so you'll find yourself changing the tune every week or month, and if you get it to lean out and detonate, or you can't get it started, or you put too much fuel down the sides of the bores, the engine won't seal up as well and you'll have lost power from the word go. I'd say leave the fuel supply stock until it's run in, and you know a tune is going to last longer than the next 1000 miles.

Air and fuel

Stock Manifolds

There are three types of stock manifold. The early KE10 manifold has straight runners and smaller inlets in the plenum area below the carburettor, Stewart used one of these manifolds when he did up his 4K, and continued to do so until he switched to EFI. I think one of the later ones might have gone better. There are two late type of manifolds, they both have curved runners which I think might be better for airflow, as well as larger holes in the plenum for the runners, as the runners are more spaced out than the straight manifold. The difference is in the port sizes, sometimes you can find one which has larger ports approximately 27mm round, compared to the normal type which are about 23-24mm round.

Removing the manifold casting between the throats (known as "hogging" the manifold) is an old performance trick from the old V8 days, and it is believed that the increase in plenum area (the bit between the throttle butterflies and the valves) helps give the engine more response if it needs it, as it has extra air to draw from.

Downdraft Carbs

A downdraft carb is easy because they're plentiful and cheap, and you don't have to modify much to fit them on. Realistically, any downdraft is going to be a compromise on your power, because the air still has to turn 90 degrees inside the inlet manifold, from down to sideways.

I could write volumes on how to select a well-sized downdraft for your motor. You can't over air an engine, but you can slow down that air too much so that the engine isn't responsive and combustion suffers greatly. Generally, anything that goes well on a stock 1600cc engine will go well on a worked 1300. You've got a bit of leeway with a 5K, as it draws more air. A 32/36 DGV from a 2L Escort is too big, unless you can get one of the rare ones with small primary throats. The stock carb is 27/27 (or 27/32 on 4K onwards), so anything from there to 32mm, with primary chokes between 20mm and 25mm will probably go fairly well.

I used a modified stock carburettor for ages when I did my motor up. They only flow about 160cfm, but the small throats makes them responsive and nice to drive. They start to gasp for air about 5000rpm, the slow down alot at 7000rpm, and you can coax them to 8500rpm if you try hard. I lathed my primary throat out 1mm and it improved the response of the engine out of sight, the thing was more snappy than some EFI cars I have driven, no kidding! If you continue to use one of these Aisan carburettors, rest assured that it is your restriction point. I don't think Sprinter twincarbs would go too bad on a worked K engine, though you can get better fuel efficiency, and just as much power, out of a properly sized single, or sidedrafts.

Sidedraft Carbs

A sidedraft is ideal for airflow. The ports are on the side of the head, the air should come in the side too. A set of 40mm Weber DCOEs or Dellorto DHLAs is probably the ultimate carburetion you can put on any motor, but they're expensive. Buying new, don't expect much change from 2 grand. Buying second hand is likely to cost you over $500 to just get a pair carbs in good working order. Twin 40s with 28mm chokes would be ideal, I think the 32mm chokes (next size up) would be a little too big but you could possibly get away with it.

I personally believe that SU carbs are fantastic. They're a throttle-type sidedraft with a vertically sliding piston which raises and lowers to both give a smaller throat, which improves airspeed, and to raise and lower a tapered needle, which seats into a fuel jet, to meter a different amount of fuel based on the airflow of the engine. SUs can also be quite expensive, but you can get them cheap if you look hard enough. Twin HS2 (1.25") SUs would be good for a 3K or warm 4K. HS4 (1.5") SUs would be great for a fairly worked 4K or a 5K.

Any sidedraft is going to need some sort of custom linkage design. Genuine Weber linkages are ghastly expensive, but they work really well. You can make up something out of redline linkage arms and balljoints and a stock throttle wheel if you're clever. Or you can pay someone to make them for you, same goes for tuning. Most carby places work for between $50 and $100 an hour.


The early cars had a very basic emissions system, all the features of which are a good idea to keep if you can. Ported vacuum advance is good as it allows the distributor to change the timing based on the conditions your right foot is placing the engines under, and how the engine is responding to the airflow at different RPM. The fast idle valve is helpful in summer, when your car will probably idle a bit higher because the air is less dense.

The PCV (positive crankcase ventilation) system is so important it gets its' own paragraph. K motors have alot of crankcase vapours, pistons just generally have alot of blowby and the rings start to float above the factory redline, so you need some sort of way to have these vapours sucked out the engine, or your oil gets ruined within 500km or less.

The stock PCV valve and carby plate is fine if you're running the stock carb or a downdraft of some kind, just modify the plate to suit, or drill and tap another fitting into the manifold. A sidedraft or SU may or may not have PCV functionality, in which case adding something to the manifold is the only way to go. Because of the way the PCV valve functions, nothing gets sucked out of the motor at full throttle. If you're doing alot of full throttle work, I'd suggest ditching the valve altogether and fitting an oil/air separator (aka: catch can) with alot of pot scrubbers in it to separate the crankcase vapours from the air, and still use the inlet manifold as a source of vacuum to suck the bad stuff out the engine.

What you definitely don't want to do is not have a PCV system, you'll just pollute your oil quickly and make a big mess on the tappet and bonnet cover if you leave the pipes open to atmosphere. I've used filters (a chrome redline filter, and a little uniflow pod) on the pipes before, and they still just make an oily mess. Use a plug or cap if you're sealing one off, a 10mm emissions cap fits great over the middle rocker cover pipe. (note: KE10 rocker covers don't have this pipe, only the rear PCV valve pipe)

Later 4K-C engines had this god awful mass of hoses and valves and the exhaust going back into the inlet manifold and vacuum secondary throat on the carb and it was just awful. I don't know if you can rip all this stuff off and have it work again, but I'd be looking at doing something about it if you're keen on making power. Likely the large overlap and varying vacuum conditions of a decent cam would confuse all the little devices anyway.

Fuel Efficiency

To make power, you need to use more air, and that air needs to be mixed in a good ratio with fuel. More power means worse fuel consumption, deal with it. The good news is that these are only small engines, so they don't use that much fuel anyway. My stock 3K in is best state of tune gave about 34MPG, or 8.4L/100km. My modified 4K consistently gives about 28MPG, or 10L/100km with a mixture of town and open road driving. When I race, obviously more fuel is used, but that's what racing is.

Exhaust Manifold

The stock exhaust manifold is alright if you're just playing around trying to get your car running nicely. There are two types of stock manifold, one has just one outlet pipe, the other has two pipes and is similar to tri-y (4-2-1) extractors. Generally, as you add a larger cam, you need to have tuned length extractors and a larger exhaust in place, or all the extra air your engine is now moving is too much for the stock manifold to hold, and the exhaust gases pulse back up against the inlet charge and fuel comes back in a fine mist out of the carburettor.

Any extractors will do, new ones are about $280. You can pick second hand ones up for $50-$100, or sometimes even free if you're lucky like me. Second hand extractors tend to be a pest, in that either the middle pipe has bent out and doesn't like to seal, or the end pipe (usually the rear) is bent from a poorly hung exhaust pulling on it for so many years, and is hard to line up with the head.

I've always used second hand extractors which I got for free, and by far the best thing you can do to make the manifold easy to locate and seal is use studs and nuts instead of bolts. This also puts a more even force over the threads of the head, so you're less likely to strip a thread. The normal two-piece manifold gasket can also cause sealing problems if the middle pipe is bent, but an exhaust shop should be able to sell you a one-piece extractor gasket which is much better. You can get metal ones which are alright, but I like the floppy asbestos type gaskets which are difficult to get these days.

If your manifold gasket just won't seal, there's a chance the flanges of the inlet manifold and exhaust manifold (or extractors if you're using them) are different thicknesses. Any engine builder or machine shop should be able to grind the faces of the manifolds together for you, to ensure they're flat and sealing properly. This should cost you between $40 and $80.

Exhaust Pipe and Mufflers

These aren't huge motors, the absolute biggest exhaust you want on any K motor is a 2" outer diameter pipe. Even that's a little big for most applications, and a 1.75" exhaust would do fine. I'd say the smallest you'd go with performance in mind would be 1.5". I don't really think getting mandrel bends is worth the expense, you're not flowing a huge amount of gas.

If you want to reduce heat levels in the engine bay a bit, make a heat shield from metal that sits between the extractors and the inlet manifold. You'll notice the stock carb has an extension to the PCV plate that protects the float bowl from the heat of the exhaust. My heat shield is just two pieces of thin plate with a sheet of gasket material between them, bolted together and cut out in a rectangle with some indents for the manifold pipes. It's fastened to the extractors with hose clamps and holes drilled in the shield.

What mufflers you end up getting will greatly affect the noise levels the engine makes. A K motor can actually end up pretty noisy if you let it, mine's over 115dB! I'd suggest the longest resonator you can fit under the car, maybe even two if your exhaust guy can find space (and there is space there), this will muffle the noise and give it a tough deep note. A shorter resonator will make it sound good but maintain the buzzy kind of noise (which plagues all small capacity engines) up high. I'm using an offset flow straight thru muffler on the rear, and it's pretty noisy. A reverse flow muffler would probably be alot quieter, but give you less flow which means less potential for performance. I wouldn't fit a reverse flow muffler, but if your car is too noisy and you get defected (as I expect to one day) then it may be necessary.

Everyone knows that black 4" droopy tips are the best way to end any exhaust :)



The purpose of a clutch is to transmit full power from the flywheel to the gearbox without slipping, to disengage during gearchanges to allow the gearbox parts to spin at different speed than the engine rpm, and to provide a smooth transition from one gear to another whilst driving on the street.

It is for this last reason that I don't really think a button clutch is really a good idea, they absolutely suck to drive with because a clutch, by design, is made to be slipped during normal driving. And because button clutches are designed to grab, when you do slip them, you wear the friction material and the flywheel down really quickly. If you have a racecar, then by all means they're a top idea, but not in a street car.

Really, even a new stock clutch is tough enough to stand up behind a worked 4K, Doug had the same standard clutch behind his for years, and it was making probably over 120hp. Perhaps the extra torque of a well-worked 5K would require more. A new stock clutch kit is usually about $120.

I have a Daikin Exedy Extreme Heavy Duty clutch kit in my car, which comes with a sprung organic driven plate (old school design one with bits of string in it) and a heavier pressure plate. It cost $240.


Depending on how you treat it, the stock gearbox will either be of no concern to you, or a constant source of problems. Lots of hard launches, snapchanges, burnouts and 8000rpm grindy shifts will destroy the gearbox pretty quickly. Usually they grind the teeth off 2nd gear, but I've heard of others going, or the layshaft gear strips all its' teeth. Generally, the K40 4-speed is tougher than the K50 5-speed, as they have less shaft length to twist up and break under stress.

There do exist bellhousings to mate a Celica T40 or T50 to a K motor, and they go for about $200. Then you need to buy the gearbox, which bolts into a KE20-55 no problems. I believe a TE72 gearbox bolts into a KE70 as well. These are strong gearboxes and should stand up to 200hp.

Beyond that, there do also exist bellhousings to mate a K motor to the alloy W58 supra gearboxes. What sort of K motor needs this I don't know, but you can buy a whole spare car for less than the bellhousing costs, never mind the actual gearbox.


The stock tailshaft is fine in performance applications, though you might want to change the universal joints just to make sure it's up to the extra stress. A uni kit should cost about $20, and you need two of them, or 3 if you have certain KE70s. Jap tailshafts (KE10-30) and Borg Warner tailshafts (KE55-70) have different types of unis. You might want to get your tailshaft balanced if you're going to be doing alot of high speed stuff, and it will probably make the load on the gearbox and diff easier too.

Out of interest, a custom carbon fibre tailshaft is about $1200.


Much like the gearbox, the diff may be fine if you treat it nicely, or you can break them often. Angela in SA is making about 215hp out of a turbo efi 5K and she has the same stock 5.7" diff as the car came out with. Then again, I've had mates who can manage to blow up a stock diff with a stock 3K. The things that kill diffs are sudden shock loading, such as dropping the clutch when you're standing still, and doing single wheelers which puts the spider gears are under too much stress. You generally don't break things when you're moving or when you don't have traction (as the shock goes to the tyres and causes them to exceed their grip limit), which is why a stock diff is fine in a dirt rally car.

I don't believe of either the Jap diff or the Aussie-only Borg Warner diff, that one is any stronger than the other. The Jap diff has a better center but thinner axles, BW centers are known to be crap, but the axles are stronger. Really, if you're driving hard enough, they're both bound to break soon. Nikki had both diffs in her 4AGE-powered KE30 and they both stood up to the punishment she gave them.

I have 4.3 gears in my diff. I think 4.1s make the car a bit sluggish. 4.5 or 4.7 would be nice, but it's already quite annoying to drive on the freeway with, though that's probably more to do with my exhaust. I'd like SLIGHTLY tighter gears, some 4.4s out of a KE26 wagon would be ideal I think. Really, it depends on what you're doing with the car. I hillclimb mine, so I'm one of the few cars that gets into 3rd gear, I want quickness. Longer gears gives you more top speed, as long as the engine is powerful enough to overcome the wind resistance.

Diff upgrades are covered on the main page. A drive in drive out shortened diff conversion is likely to cost you $1500 or more. Pick something with LSD.

The End

That's it. Build away. Hope you learned something!

Back to K Series

Back to Main Page