SpeedTECH The definitve guide to EFI transplants

NOTE: This article was originally written in early 1994 and published in ‘Sports Car Talk’, the magazine of the Sports Car Club of NZ. I have given it a quick once over to correct a couple of minor errors. Prices are in $NZ. If you want to know more, please do not hesitate to contact me. 

Phil Bradshaw

WHEN 16 VALVES JUST ISN'T ENOUGH! (Or 'Easy is only Skin Deep')

This article is a follow up to one I wrote in early 1993 ago concerning the installation of a 16 valve Toyota 4AGE into my car. I apologise to the casual reader who may find the content excessive, but hope that it is of use to those contemplating buying one of these marvelous little engines, or indeed contemplating any engine installation. As before, I have endeavoured to make the article as accurate as possible, and can be contacted at the above address if you would like to know more. And now on with the story...

Recite after me: 'In 1992 I built up a Leitch Supersprint Lotus 7 Replica (similar to a Fraser 7)...' You know the story! Well, what you may not know was that at the Practice Day at Taupo back in May 1994 I put in one 'race' and retired due to a nasty little rattle emanating from my motor...

Expert trackside opinion was consulted, and the general consensus was that my hitherto faithful 1983 Blue and Black top 4AGE had killed a conrod bearing. It still ran fine, oil pressure was OK and although the rattle was definitely there, it was not too bad (well, as they say - 'Is it broken bad?' - Have you ever seen anything broken good?)...Anyway, I digress.

So, I spent the rest of the day marshaling, which was a bit of an irk, as I felt I was doing really well until the engine died, and was looking forward to more laps. I figured it was better to retire and let the engine live to be able to drive home, rather than destroy the motor completely. End of the day we drove back to Auckland without a hitch, and I drove around, albeit gently as it is still my only car, until Queen's Birthday weekend when I pulled the car off the road. I was fortunate that a good friend was overseas with work for four months and had loaned me his car.

The Leitch had been on road for 18 months at that stage, and through one thing or another, such as me heading overseas with work, had been driven for about 16 of those months, and clocked up 23,000 miles, on top of the 84,000 km on the engine initially. I guess having the motor die at about 115,000 km, given the life its had, is pretty good really.

My car was in need of some major TLC, which I must confess was not part of my plans when I built it initially, but circumstance went against me - I had a rear mudguard crack and mostly tear off when I flicked up a muffler lying on the motorway, ripping the nutserts out of the aluminium skin. I was rear ended by a Gentleman who not only didn't notice the red traffic lights but really made my day by being a sickness beneficiary with no insurance. Legal proceedings are being pursued, damage was minor but necessitated reskinning the back of the car - $1450 worth of work, but a very good job by Terry's team at Terry's Panelworks in East Tamaki. (Note - some 2 years later I am still only 50% successful in getting money out of the other driver...)

I had a minor wiring loom fire, the cause of which is still a bit of a mystery, and I can only guess that some stray condensation shorted across terminals under the scuttle. Why it took 18 months to occur I have no idea. Then there was the broken panhard rod mount - mine is the only Leitch to ever have this happen, which was small consolation when it happened at 5 pm 30 km north of Taumaranui on Easter Thursday, when I was heading towards Wellington! Fortunately a local welder took pity on our sob story and did a patch up job.

A few other minor niggles had got to me, so a major rebuild commenced. How major? Well, it took me 3 months to build the car initially, but I managed to stretch the rebuild out to almost six. Admittedly I had more time available, like a solid month off work, when I first built it, but as you can appreciate, I have virtually completely rebuilt the whole car, and don't intend doing so again. Mind you, that is what I said when I built it originally...

On to more interesting stuff, and the main thrust of this article: what do I do about the engine? I talked to a number of people, rang around and scratched my head a lot. As best I could tell, the options were:

rebuild the 4AGE, 
replace the 4AGE, or 
repower the car with something else. 
I'll say now that I had no money, but was very conscious of the old adage that the cheapest option is not always the best, hence I did what I consider to be all the homework first.

Rebuilding the 4AGE was going to be a major, as the engine always smoked a little on deceleration, had a dead bearing or worse, had a worn clutch, delivered 118 hp/88 kW when new (and I was hungering for more) (sound familiar?), and had oil problems resulting in it turning into a two stroke on occasions. Now, all these maladies were curable (except for the power output - remember I used the injection, and was basically restricted from putting in hotter cams as the computer can't handle it), but the question was at what cost? I reckoned on at least $1000 to do the job properly, even if I did all the spanner work myself.

Next option was replacing the engine. I rang around, and found I could get a similar vintage Blue and Black top for $550 cash from Engine Imports. There was a choice of two, which were the last of a large shipment. Both were incomplete, and with no computer or wiring loom, however I had all the missing bits available from my old engine. Thing is, such an engine is probably not much healthier than my old one, and to me three month's warranty is not sufficient consolation on an 11 year old motor that is going to be given a hard time, and must be reliable as it is normally my only transport.

Furthermore it would still require modifying to prevent it suffering from the oil problem, and it would be prudent to replace cam belt, etc. At the end of the day, I would be conceivably driving on an engine that was no better than the one I broke, and more than likely to be on its last legs, which might last 1-2 more years if I was lucky.

So, that left option three. I talked to more people, and paid a visit to Neil Fraser, who listened to what I was saying and showed me a collection of engines he head lying around. We talked about cost and output and installations etc, and I came down to three options:

Red top 16V 4AGE 
20 Valve 4AGE 
Purchase of a red top 4AGE would cost around $1500, and would deliver about 135 hp/100 kW. The engine would be up to about 5 years old, I think. That was the good news. Bad news was it was an East/West front wheel drive motor, so the inlet manifold faced backwards, and the computer is different, necessitating a full rewire on the engine side.

A 20 Valve 4AGE would deliver 158 hp/117 kW and be up to 3 years old, however have the same problems as above, plus an air flow meter to squeeze under the bonnet somewhere (as opposed to the map sensor fitted to the 16 Valve), and cost $2500.00.

The 3SGE is the latest 2 litre motor fitted to the MR2, and is looking like being the formula motor for the Clubmans' Series. It has an inlet manifold for the injection that wraps over the head, which makes it too high to fit in a Seven - the solution being to put twin side draughts on it. This is fine if you already have a 4AGE or similar running on twin side draughts, however if you are like me, adoption of the engine means that you need to buy:
engine $2500
manifold $ 220
carbs $1100
jets/throats $ 400
linkages $ 200
air filters $ 250
bell housing $ 400
black box to make distributor go $ 400
Total: $5470

Add to this new exhaust system, low pressure fuel pump, new engine mounts, shortened sump, plus the other complications involved in turning an East/West motor North/South mentioned above, and it all looked too expensive for me. Which is a damn shame because the motor puts out 180 hp/135 kW in stock form, even more on carbs, and a fair whack of torque.

If I ever got rich (or someone paid me a lot of money for my car and I built a new one), I'd like to attempt building up a 3SGE on injection with a custom built inlet manifold to fit under the bonnet, and either the stock computer or an aftermarket one. I reckon it ought to be possible, and believe it is an option being looked at by one or two people.

I believe that the rationale for adopting this engine as the Clubmans' motor, is that people were spending $15,000 on a 4AGE, getting 230 hp and eating those of us with fewer resources. The 3SGE offers a hell of a lot of power, out of a stock engine, that once set up should last for quite a while.

Granted that speed costs, we all know that, and realistically spending 5 or 6 grand to get 180+ hp out of possibly a two year old stock motor is bloody good value for money. Unfortunately it was also beyond my means. Additionally, perhaps mistakenly, I was concerned about my standard T 50 alloy case Toyota AE 86 Levin gearbox and Escort diff handling the power and torque. I know they are both reasonably beefy, and a seven doesn't weigh much, but as I keep saying, my car must be reliable, and I'm just not as rich as I'd like to be.

As I said, if I was building from scratch I might do things different, and would probably go injected 3SGE, 5 speed and fully independent based on a LSD RX7 rear end. Dreams are free, thank God!

So, after much deliberation, I decided to go for a 20 Valve, for the following reasons:
I was concerned that a replacement blue and black top would see me back in the present situation in 12 months time (if I was lucky), and would not cure my wish for more power or alleviate the ...
... oil surge smoke (Which I could fix by using a surge/collection tank, external breather etc, however I was concerned that when cornering at high RPM oil was pumping into the head quicker than it drained), 
rebuilding my motor would cost at least $1000, and again not cure the desire for more grunt or the oil surge smoke, and for approximately $500 more ...
... purchasing a red top would solve the desire for increased power and oil surge smoke due to redesigned oil returns, however for the effort involved in fitting the engine to the car, and spending a further $1000 ...
... the 20 valve offered a state of the art motor with a very high output, that should last a long time, should bolt in relatively easily, and cost an 'affordable' (or is it justifiable?) amount of money. 

Wait a minute, I thought I had no money. All I can say is God Bless my Bank Manager! Poor man was subjected to my analysis of what I wanted to do, and refinanced a loan that I got initially to build the car with, and had almost paid off. I think he would have given me the money just to get me out of his office!

At the end of the day, getting my car back on the road has not been cheap, however as I have already explained, I have undertaken a complete rebuild, with the re-engining of the car being just one aspect of it.

So, what is a 20 Valve? I must confess to not knowing a lot about them, but what I do know is that it is a heavy duty 1587 cc 4AGE bottom end with a 20 valve head (how surprising). It has three inlet and two exhaust valves per cylinder, and variable inlet valve timing. My understanding is that at high RPM a solenoid valve directs oil through an additional camshaft drilling that fills up a fluid coupling within the inlet cam drive pulley, rotating the camshaft 15 degrees ahead of the periphery of the pulley, thereby advancing the camshaft (and the inlet valve timing) by 15 degrees.

Otherwise the motor is pretty normal with a knock sensor, air flow meter and not one, but four throttle bodies - the inlet system looks a lot like a collection of stubby sidedraught carbies sucking from an air box.

The 20 Valve is designated 4AGE-10, whatever that means, and has 'TWIN CAM 20' on the cam cover (which makes everyone think it is a two litre), and '4 Throttle' on the inlet air box. The cam cover is silver and the plug leads are hidden beneath a central cover. The distributor drives directly off the back of the exhaust cam, and the variable cam timing solenoid sits in about the same position as the normal distributor mount.

Ordinarily the motor lives in a Toyota AE 101 Apex or Levin, both of which are front drive and apparently Japanese domestic models only. I have seen a couple in New Zealand, one of which was a 16 valve, and there was a 20 valve advertised for $30k, but I didn't get a chance to have a look at it. They come automatic and manual, and run 11.25:1 compression, as far as I know, which is why they aren't available outside Japan, due to the requirement for high octane unleaded.

I bought my engine from Neil Fraser, and received motor (fanbelt to clutch, air flow meter to extractors, alternator and starter motor), uncut loom, computer, coil, igniter etc, and a wiring diagram. Neil also gave me a warranty that everything was in going order, as unless the motor rattled or smoked from start up, it was unlikely to die in the foreseeable future.

This cost me $2500 cash, which I thought was reasonable, bearing in mind other prices I had been quoted for a 20 Valve. Furthermore the engine arrived in NZ only 2 days before I bought it, and I had the choice of two. Additionally it came from the same crowd in Japan that Neil uses to source his FC7 engines, so I figured it would be a safe bet. I also felt happier dealing with Neil as he knows what he is on about, and if he sells you a motor with the statement that 'all you need is here', then it is. I was concerned about having an engine from an importer that I would not be able to get to go without paying someone lots of dollars for a few seemingly insignificant parts.

Neil is not really in the engine supply business, and as I hadn't actually bought a Fraser 7, he explained that I was pretty much on my own (ie they would not hold my hand during the installation), but he did give me the number of a guy who did work on the 20 valve they raced in the Targa Tasmania. I thought that was fair enough, and at the end of the day, I think I managed quite well on my own.

Incidentally, Strong Brothers at Penrose was recently selling a 1600 Honda VTEC motor, gearbox, computer, loom etc for $2000 inclusive. Be warned the damn things rotate backwards though, which is fine if you are building a mid engined car, otherwise I guess you have to turn your diff upside down...

I talked to Peter Sullivan in Southland, who is building a Leitch up slow-time with a 20 valve, but he has yet to get it running. He was planning to use an aftermarket computer, as is Darryl Jeffares. Neil Fraser's TT one was extensively modified, and other than one rumoured to be running in a Chevron, I have not heard of any others actually on the road, or that are running in a stock configuration.

First hurdle was deciphering the wiring diagram, which was all in Japanese. I quickly discovered that Japanese is a very complex written language, so much so that it can apparently take until late teens before people can really understand a newspaper. One of the guys at work lived in Japan until he was 12, and he managed to decipher about a third of it. A friend of a friend, majoring in Asian languages, got stumped due to the technical content, and I was on the verge of placing an add when my flatmate, who is a sales rep, came home from a day trip up North with a Japanese exchange student that his Boss was hosting. She was being picked up in half an hour, so I collared her and she translated the entire lot!

Whilst chipping away at the wiring diagram I was also installing the engine in the car as a trial fit. As it is a front wheel drive motor, there are a few differences in component configuration. Firstly the starter motor is on the inlet side of the engine, which would not clear my foot well and also not mate with the gearbox bellhousing. The solution was to use the old motor's flex plate, starter motor and bellhousing, etc. This all bolted straight up, however if you buy a 20 Valve and a T 50 gearbox as separate units, you will not have the flex plate. All is not lost, as the 20 Valve has the blank where the rear wheel drive starter motor goes, so with a bit of care, accurate cutting can be done, followed by a simple amputation of the now redundant other position.

As I said, the gearbox mounts straight up, and I re-used the two block to bellhousing supports, however one of the holes didn't line up, so I am one bolt short, but I appear to have got away with it. Darryl's motor has a big cast aluminium bracket that does this job, but his engine was out of an automatic, and the casting does not line up with the bell housing. He is contemplating his options at the moment.

The Leitch uses the central engine mounts on the block, and apart from drilling a hole to clear a dowel pin, my old mounts mated exactly. I believe the FC7 uses a different location on one side, and the 20 valve block is different at that point, necessitating a new mount, which is not really a biggie.

So, next step was installation. The motor and box went straight in (well, as straight in as any attempt at squeezing a motor into a 7 goes!), except for having to stop half way to lop off a chunk of redundant bracketry that was interfering with my pedal box. I installed the motor complete, except for the air box and loom.

Unfortunately getting the motor in place was the easy bit. The challenge was getting the EFI system to fit! There were two main problems: the air box was too big to fit, and the air flow meter had to go somewhere. Additionally in the standard East/West set up, the engine sits with the exhaust manifold forward, which makes it sensible to locate the radiator hose connections on that side, at the flywheel end of the head due to all the cam drives and variable timing bumpf at the front. Unfortunately, when you rotate the motor North/South, it puts the inlet at the back of the block, and you've got an exhaust manifold to negotiate around as well! The thing to remember is that this engine is going into that shrine to minimalism known as a Seven, which was initially designed to take a 1000cc pushrod motor, not a twin cam fuel injected beastie.

Anyway, I digress, yet again. The decision on the Air Flow Meter had to be made first, as it is attached to the Air Box via a bit of flexible tube (which I didn't have, and probably wouldn't have been much good anyway), which mates into the flywheel end of the airbox. Dilemma: do I mount the air flow meter behind the engine, above the foot wells, where it gets all that nice, warm, low density air, or do I perform some surgery on the airbox and have the air flow meter attached at the front of the engine? Trouble then is that it would be sucking from directly behind the radiator...

I thought about these things, and figured that the old motor basically sucked from behind the radiator, and still went ok, and I had seen a picture of a Fraser (I think) with an air flow meter, sitting on the footwells. I've gone with this option, as there is a little more room on the footwells. Additionally I had to cut the airbox to fit below the bonnet, and after surgery the front of the box wasn't big enough to fit the inlet tube onto anyway!

The Airbox was a bigger worry, literally - it sat too high (above the bonnet line), too low (its base wanted to sit where a spaceframe tube lived), and too long (it interfered with my pedal box cover). As you may recall, a Fraser with a 4AGE on injection has a bonnet bulge to clear the manifold, whereas the Leitch does not, due to chopping the inlet manifold by 30 mm. I could have made up a bulge for the bonnet, but it looked too hard, and I figured it would be a big, ugly bulge (I personally think the wee bulge on the Fraser is quite elegant and subtle, but this one would be in a different league!), and it still wouldn't cure the other areas of interference.

As a sideline, if you have the luxury of space, particularly height, as you may do in a Lynx or a 'normal' car, you can quite literally bolt the air box on upside down, which will then put the inlet from the Air Flow Meter towards the front as well, or If you don't mind a bonnet bulge, this may be a workable solution.

Of course the other option is to build your own air box, which is what I considered initially. After all, people have been doing it with carburettors for years. I went and had a talk to the friendly chaps at Weber Specs, as I figured they'd have the good oil on such things. My first plan was to construct a plastic air box out of suitably sized PVC conduit, drainpipe and glue, with a good coat of paint on completion to cover my sins. This appeared to have merit, until the Weber guys started talking about sufficient volume for intake depression when the throttle is floored until the air flow increases and other scary things. Like everything I guess, it all appears simple from a cursory glance, but gets more complex as you look into it.

Additionally the Toyota has intake trumpets within the air box, which are curved like a cow horn and about 150 mm long, as the box is deep but not wide. I thought more about it and decided that in order to get a decent volume within the box I would have to use a large diameter tube that again would not fit under the bonnet, and probably look rude. Time for plan two. I rationalised that Escort Rally Cars survived for years on stubby trumpets (or none at all), due to the proximity of the carbs to the strut tower, and they still produced heaps of grunt in race tune, so I figured I could chop down the steel air box and put some stubby (ie inch long) trumpets in later. (The throttle body inlet is 47mm wide and vaguely similar to a Dellorto bolt pattern).

With a plan of action, it was time to work out where to chop the box. I figured that the actual air box volume wasn't that great, as the trumpets took up most of it, so I reckoned that reducing the air box and turfing the trumpets (with only short replacements) would work. The Weber Guys agreed in principle, so away I went. I then had to suss out where to chop the box. Of course it isn't parallel, but with a bit of cunning, I sliced 30 mm out of the bottom half and with a small amount of tin bashing got all the seams to line up close enough for welding.

I can't weld to save myself, but I talked nicely to one of the guys at work who is a very good welder, and he did it for me. He had a bit of a mission as the airbox halves are basically laminated from thin sheets of steel (they are quite deep pressings), and proved very difficult to weld. Ultimately he used stainless welding rod to get the job done. Next step was to take 15 mm out of the rear wall to clear the pedalbox. What was really scary was the amount of time it took me to section the box, make up the new inserts, have it welded and grind the welds smooth on completion - the Guy who supervises the Welding shop reckoned that as a commercial job it would probably be worth about $700 to have it all done...

That done, I could finally bolt the bottom half of the air box to the throttle bodies in the car. Naturally it was still too high with the top half of the box in place, so it was time to put the nose cone and scuttle on and start stretching tape between the two, to get the bonnet profile against the air box. After much head scratching a decision was made, the top of the box was lopped off, and a new one welded on. The net result is an air box that clears bonnet, spaceframe and pedal box by 10-15 mm, which is about the same clearances as the original 16 valve 4 AGE installation. Not forgetting that I have very rigid engine mounts, so the motor doesn't move about much (except when the sump hits the road).

The Air Flow Meter was a tight squeeze, particularly with trying to site the ignition coil/igniter pack, battery and windscreen washer tank in the same area, not forgetting pedal box and master cylinders. I eventually overcame this, but not before spending a considerable amount of time and effort with cardboard and then aluminium mounting brackets. I swear I made 5 of each before getting it right. I used a modified sidedraught K&N airfilter on the inlet to the air flow meter. Next step was connecting the air flow meter to the air box. Trouble was, it needed 70 mm tubing. After much anguish, I used a section of Escort heater concertina tube, that normally connects the heater to the eyeball vent in the dashboard. I was intending to use it only as a temporary measure, but it seems to work fine so far, and looks OK as well. Time will tell, I guess.

I reused the standard 16 valve exhaust manifold as I could not afford to make a new exhaust. The cast manifold required slight modification to the outermost mounts but otherwise matched right up. I intend to use the factory extractors as part of a 4 into 2 into 1 system in the near future.

There was the usual messing around with fuel lines etc to get the hoses to mate up with the right bits, as again the inlet and outlet from the fuel rail were at opposite ends to the 16 valve. I used the new alternator with the old mount and belt type, as it was more compact than the 20 Valve arrangement, which uses an idler pulley, I guess in an effort to clear some obstacle that exists in the Apex/Levin. In so doing, I had to modify the dipstick tube mount, as it connects to the other alternator bracket. I just copied the 16 valve installation, whereas Darryl has just cut the rest of the 20 Valve alternator mount bracket off beyond the dipstick mount. You cannot swap the dipstick with the 16 valve as the levels in the sump are different.

The throttle linkage was another source of fun and games - the 20 Valve has a cunning mechanism utilising levers and pivots that sits above the throttle bodies, with the cable inlet from the front of the engine. Two problems here - the linkages sit above the bonnet line, and the cable comes in from the wrong end. I figured I'd just rotate the whole assembly 180 degrees and mount it underneath the throttle bodies, but after much messing to get angles right discovered the spaceframe was in the way anyhow! I ended up using a bottom entry direct wire set up similar to a twin sidedraught installation, which works really well, but needs a little modification to some angles to prevent the wire abrading the side of the adjusting barrel. I used bicycle brake cable components for the set up.

The oil filler cap has a small spine on it for you to grab and quarter turn to release it, but of course it hit the bonnet also. The solution was to chop the spine off, but the wily Japanese made it hollow, so I machined the top flat and superglued a flat piece of 'Engine Oil' engraved plastic to the top. Looks like a bought one and still easy to undo. The injector wires are run in a small square section conduit the sits above the injector heads, screwed to the throttle body castings, which also hit the bonnet (surprise, surprise). Initially I was going to turf it and just use concertina cable coverings, but ended up mounting it below the injector heads, where it clears the bonnet just fine, mounting on the redundant throttle linkage points.

I threw away the vacuum lines that went from the engine to bits of the Apex/Levin that I didn't have, and also flicked the little throttle return dashpot. A couple of bits looked far too integrated with lengths of wire going to the computer, so I left those, but as there is only two of them, and you can't see them from above, why worry?

Then it was time to yank the motor and gearbox out again, clean and paint them before re-installing them for good (hopefully). I used Epiglass PA-10 as primer and blue Hammerite for the block/gearbox with black sump and detailing, which looks really good, but isn't all that visible with the injection in place! As an aside, I have found that a relatively light coloured engine makes niggly minor leaks easy to find.

The next step was the wiring. I guess I like making things hard for myself, but I don't like the idea of running an EFI engine without having a warning/diagnostic light. What this means is that you must run it with all the sensors, particularly the speed sensor, or else you get a warning light. The speed sensor does not affect the way the engine goes, but I don't want an intermittent warning light from it, otherwise I'll never know when I have a real problem! Fortunately I had the inductive pickup from the old engine, so I just wired it in.

My main concern was the exhaust/Lambda/oxygen sensor. I had run one in the old engine until it broke off against the body when the engine shunted back (after the sump hit a big lump of tar seal at speed). The engine ran fine without it so I never replaced it. I always ran the old engine on unleaded, but was concerned that the new one had to be on super 'cos of the compression ratio (lead poisons the sensor), and had a four wire sensor vice the single wire of the old motor (the four wires are for sensor, earth reference, and a heater circuit to get the unit responding quicker). I was aware that some of the newer engines will give an error if the exhaust sensor isn't in the loop, as a means of ensuring emissions compliance, and I was concerned that Toyota had got a bit more worked up about such things since 1983.

I figured it was time to get some advice, and made the mistake of ringing around. I got replies as follows:
You don't need one, 
You do need one, 
We can sell you a black box that fools the engine, 
so and so's black box doesn't work, we spend a lot of time pulling them out of cars that people put them in, 
we have one that is lead tolerant for $300, 
uses an EA falcon - they're lead tolerant, 
no such thing as a lead tolerant sensor, 
use an aftermarket computer and you won't have to worry, and the common reply 
don't know mate, try (one of the above people)! 

It was time again for plan B - get the sodding thing going and stuff a meter up the exhaust. I had a contact that was willing to let me have a play with some sensors that came from full race Cosworth Escorts in the UK if I needed them. I found it unsettling that I couldn't get a consensus of opinion. Yet again, a word to the wise - very few Fuel Injection/Electronics experts know much about setting up the more obscure projects. I guess the answer is to use an aftermarket computer, but I'm not that rich and decided that to go that way only as a last resort.

The rest of the wiring was pretty straight forward, however my freshly translated diagram had only the engine electronics on it, so if a plug had 10 wires and only 2 went to the engine or computer, it didn't tell you what the rest did! I already knew that I couldn't buy or source a manual in English for the engine, but found that the Toyota EFI system is a derivative of the Bosch system, and the 20 valve set up is much the same as the system laid out in the 1987 4AGE and 1983 AE86 Levin manuals I do have. So, with a bit more head scratching (no wonder my hair is falling out) I managed to sort it all out, which I thought was pretty good going, even if I do say so myself.

I had managed to sell my old engine, loom, computer, spare sump, photocopied manual, original Toyota subframe and mounts to some guys for $750, which I thought was excellent as it was broken (I still do not know what was wrong with it - either bearings or maybe a piston - but it still went well) and I figured that fixing it would not increase the amount I could sell it for any more than the cost of the parts, to say nothing of my time.

Unfortunately, in selling the old motor I also sold the circuit open relay. So what you say? Well, this relay shuts down the fuel pump if you have no air flow into the engine, a situation that occurs when you crash and stall the engine. This prevents that fractured fuel line becoming an 80 PSI flame thrower. Knowing how I drive, my luck, and the inputs to the computer from the relay, I figured it might be prudent to have one.

The local Toyota wrecker sold me one for $25, which I thought was pretty good as they cost about $60 new. Trouble was (of course) that on closer inspection the relay was arranged differently internally - the relay has two operating coils, and a resistor and capacitor were in parallel on the opposite coil for the new engine EFI system. And (naturally) it wasn't just a case of swapping the wires. I went back to the wrecker, who didn't have any others in their drawer... But they did have an AE 101 front cut! (less engine). 20 minutes of digging around later (the wiring diagrams are schematic only and don't tell you the physical location of the bits) I found the relay and the plug, and swapped it, which was a bonus as this relay was out of a 1994 car and costs $100 from Toyota (and you can't buy the plug). (Note: You don’t need this relay if you wire in the oil pressure cut out - you can slave the fuel pump off it.)

The next hurdle was the cooling system. Remember me saying that the inlet/outlet were at the back of the head? Well, wasn't this fun? The shortest route to the radiator was over the top of the exhaust manifold, which I initially rejected for two reasons - the hoses would look ugly and be damned difficult to mount clear of the exhaust, and would run higher than the waterjacket and radiator, which seemed an awfully good recipe for air locks and baked engines.

Which meant I had to run lines down and along the spaceframe then back up to the radiator. So what's the problem? Simple. Each run is 41/2 feet long. Measure it, it is deceptive. And I don't have the space to weld in fixed tubes. If I was building from scratch maybe, but not now. OK, I'll do it in flexible hose. Wrong! You cannot buy radiator hose in more that one metre lengths. Which means you have to use braided line - at about $80/metre. For 3 metres. Plus Fittings (four in number, at about $25 each). Ouch! I thought about steam hose, which is rated at 180 degrees Celsius and 150 PSI, but it looks ugly and is about $50/metre anyway. Non toxic PVC is rated to 250 PSI, but only 60 degrees Celsius.

All this looked pretty nasty until I spied a roll of used PVC tube in the come in handy pile at work. This had potential. I liberated a chunk, figuring that I could dummy up the installation and that way not buy a millimetre more braided line than I had to. Additionally it gave me the opportunity to fire the sucker up before I bought the proper stuff. (Note: Graeme Bluett in Hamilton has put a 20 valve into his Fraser and made rigid aluminium pipe that are connected to the engine to reduce the legnth of flexible hose required.)

So, after much messing about, it was all wired up and ready to go. I thought it would be wise to crank the motor without the plugs in to get some oil pressure up, and I was concerned that I may have made a mistake in wiring the computer, so I checked all the wiring with the diagram, and checked it again. I was concerned that the plug for the car side of the computer that Neil gave me bore no relation to the diagram colour coding, but one of his employees advised me to take the lid off the computer and check out what was written on the circuit board adjacent the plug pin. This confirmed that the drawing was correct and calmed my nerves considerably.

I also had the coil unhooked so that there weren't any stray volts running around. Time for the big moment. I hooked up the battery, and was rewarded with a pleasing absence of sparks, smoke or untoward aromas. Then it was ignition time. Again, everything seemed OK, warning lights came up. Start. And nothing! At all! The lights stayed the same, a relay went click, and that was it. Hmmmm. Maybe it does need the computer after all (there are some 'start' inputs into the computer). Tried again. Still nothing. Hmmmm. Maybe it needs the coil as well. Nope, no change. Oh well, at least the computer hasn't gone fizz...

Time to check the starter motor, which I had rebuilt, and previously tested. The terminals were shiny. Out with the multi meter. No volts at the solenoid. I traced the wire back to the start EFI solenoid and found a wire stub I hadn't hooked up, and had overlooked because it was tangled up amongst a bunch of wires, and wasn't shown on the wiring diagram. I hooked it up, and away it went. Once oil pressure was established, I put the plugs in and started it. It ran first time, happily at idle, as smooth as you could wish for, with all warning lights out.

Then it was time to top up the radiator fully. Fortunately I had been warned by Fraser's that they were bloody difficult to get the air locks out of the head, and after an hour of messing around and almost cooking it a couple of times, got it sussed. There were two problems - I had the hoses hooked up to the wrong end of the radiator, and the thermostat was stuffed. So, I sorted that out, and managed to burp the air out. As it turned out, having the clear hoses in enabled the engine to be checked clear of air and for good flow.

Next step was test drive, as I wanted to be sure it all ran OK before I formalised the wiring loom. And I needed a car fix, it had been too long. So, with a chase car in case of glitches, we headed off, with crash helmets on (to alleviate the hazards of bug strike), as the car had no nosecone, scuttle, bonnet, windscreen, lights, mirrors, mudguards, etc, etc. The only taillight, other than numberplate, was the high stop light taped to the roll bar. Being law abiding citizens we were driving on 'private roads'. Yes. Well, what can I say? We put about 100 miles on it (on PVC radiator hoses) at speeds up to 100 mph (I had some shoddy tyres on that weren't really safe to go faster on), and cruising at slow speeds, and it went great. No warning lights, grey exhaust and oh what a beautiful engine.

It doesn't feel that much more powerful than the old motor because it is so smooth, but given the amount of oversteer and wheel spinning I'm getting, it certainly appears to be delivering the goods!

I removed the exhaust sensor wires, but kept them incase things change in the future and we get high octane unleaded, although it certainly runs OK without it. I've yet to get the exhaust analysed, but it does appear to be fine.

The cars first outing in anger was at the end of year motokhana at Botha Road in Penrose. The car was mostly complete, except for proper radiator hoses and a few minor things. The day went well until a relay played up and the radiator fan didn't cut in. The temperature gauge was steadily climbing but wasn't too bad so we thought we'd do one more run and then sort it out. We did the run and stopped the car with the gauge on three quarters, which is where it runs if you give it death on a hot day, and as the ignition was turned off the PVC hose burst with an attention grabbing bang!

I was then subjected to a dozen knowing looks of 'should have known better', which I guess I deserved, however the truth of the matter is that the car ran a total of 1200 miles on the PVC and they only blew once (which is enough) due to the fan not working. Whilst I certainly would not advocate the use of PVC as radiator hose material, it does provide a cheap means of dummying up an installation. I found it invaluable, as it enabled me to end up with the inlet to the engine being 1.5 metres of braided line running along the bottom of the spaceframe and up to the inlet, with the return being via a length of stainless exhaust tube above the exhaust manifold with braided line to connect it to the radiator and engine. All up I needed six feet of braided line, and six clamps.

The next step will be plumbing in the oil cooler I bought, lack of space and money delayed me initially, and following a minor collision when I was backed into, I will install it when I repair the damage once the insurance claim is settled, but that is another story.

My engine lives on a strict diet of Mobil 1, and has a thermostatic electric fan to keep it cool. The fan no longer whines like a jet engine because the collision broke it, and the interim replacement is considerably quieter. Additionally I have a 7 PSI oil warning light that is hooked into the ignition circuit, so that the motor will not fire until the oil pressure is up and the warning light goes off. This also works in reverse so that if the engine loses oil pressure, it shuts down at 7 PSI. Hopefully with the above measures I have eliminated start up wear, and fully protected the engine. Time will tell.

Since getting the car running in late November 1994, I took it down the South Island over Christmas, and have now clocked up 70,000 km on the 20 Valve. I had a wee fire just prior to heading down South. My pressure gauge line was PVC, same as the old engine had, rated to 600 PSI. It ran about three inches away from the exhaust manifold, but could not get any closer, which I thought would be fine, but was (yet again) wrong! One day I got a whiff of smoke out of the bonnet, noticed the oil pressure gauge was on zero, and so killed the ignition before popping the bonnet, when I found that oil had sprayed onto the exhaust manifold and created flames about a foot high.

I had visions of my car burning to the ground four days before driving down South, and was about to beat them out with my T-shirt when they went out of their own accord. Apart from soot on the bonnet, there was no damage, however I received a fire extinguisher for Christmas, which was one of those things I'd never quite got around to buying...

I bought a new T-piece that leads the line away from the manifold, and the first foot of the line is now copper, as an added safety feature. I had resisted copper at first, as I was concerned about it work hardening and fracturing, but figure PVC bursting is more of a worry. Incidentally the compression olives for copper and PVC tube are different - the copper ones are barrel shaped whilst the PVC ones have a collar to prevent them being compressed too much and cutting the tube.

Installing the thermostat was a cause for concern until I cleared the forest for the trees. To cut a longer story short, I installed the thermostat, which resides horizontally in the back of the head, whilst I had the garden hose in the radiator, ensuring a constant flow of water out of both sides of the thermostat until it nipped up. In this manner, all the air was expelled from the engine. Additionally I have a gradual slope from the engine outlet to the radiator, which relieves all air from the engine side of the thermostat. (The conceivable problem is air locking the inlet side of the thermostat.)

The temperature gauge sender, again a Triumph 2500 PI one, to match my gauge, screwed straight in to the Toyota mount. The problem is that the sender is on the radiator side of the thermostat, so it measures the cooled return water from the radiator rather than the engine water jacket temperature. The hose between the radiator and the inlet to the thermostat alone holds over a litre of water, not counting a heavy duty modified Cortina radiator, the return hose and overflow tank, which means that there is a hell of a lot of water to warm up before the gauge reads anything. Typically this takes over 10 minutes of driving to achieve. I thought of mounting the sender on the engine side of the thermostat like Escort motors are, but I am going to run an oil temperature gauge on the inlet side of the oil cooler, which will give me a quicker indication of engine temperature.

In the final analysis I'm very happy with the engine so far, although bolting the engine into the car was the easy part! I consider that I had effectively no teething problems and have been most impressed with the engine. The most surprising factor is that it is considerably more economical than the old engine - I drove from Tawa in Wellington to Pokeno in the Bombay Hills on $45 of petrol. This was on a trip that took 8 hours from Tawa to Glenfield, with a few brief stops along the way, and no holds barred when overtaking, on a 4.44 diff ratio. I reckon that's pretty reasonable.

Sure, it would be nice to have a 3SGE with even more horsepower and torque, but where do you draw the line? As I said in my original article about installing the 4AGE - my car is not a race car, is my only transport and is used every day. It was a bloody good road car with 118 horsepower, and with 165 it is realistically bordering on having too much power now, but it would appear to be the ideal solution for me, and I guess that is the important part - it doesn't matter what engine or drive train you choose to put in your car - if it works well and does what you want it to, then it is the right combination.

Once again my thanks to all the people mentioned in this article, and the many others not, for their time and advice.
Sump Shortening and Bash Plate
$ 168.75
Oil Cooler/Adapter Plate
$ 281.25
Fan Belt
$ 18.00
$ 90.55
Coil Lead Cap
$ 3.45
$ 28.70
Temperature Gauge Sender (Triumph 2500 PI)
$ 14.00
7 PSI Oil Pressure Warning Light Sender
$ 20.00
Accelerator Cable Bits
$ 19.00
Air Filter
$ 122.63
Crankshaft Spigot Bearing
$ 6.05
Braided Fuel Lines/Vacuum line
$ 226.62
Electrical Fittings, Wire, Conduit
$ 65.55
Oil Filter
$ 15.87
Sump Sealant
$ 7.06
Exhaust Manifold Sealant
$ 3.83
Exhaust Gaskets
$ 20.72
Radiator Flush
$ 8.79
Stainless Steel Nuts and Bolts
$ 25.00
Circuit Open relay
$ 28.12
Braided Radiator Hose and Clamps
$ 300.00
Stainless Radiator Tube
$ 35.00
Oil Cooler Hose and Fittings
$ 227.00
Oil Gauge T-piece and line
$ 30.00

Bits I already Had:
Engine Mounts
High Pressure Fuel Pump
Alternator Bracket
Clutch Flex Plate
Fuel Filter
4AGE Engine Manual
AE 86 Corolla Body Manual
Drive Shaft
Starter Motor
Compatible Gauges and Senders

Note: The 20 valve now has some 42,000 miles on it in my car, and despite running a 4.44 diff ratio returns a genuine 34.5 mpg regardless of around town or on a trip. The car will do 115 mph on the flat before the speed sensor speed limits it, although it is still accelerating. It will accelerate to 110 mph up long hills... I believe it will do a genuine 120 on the flat in full road trim, clamshell, windscreen etc, but it is largely academic as I want to keep my licence. Not bad for a car with a cd of 0.74...


Copyright © 2000 SpeedTECH Last modified: January 23, 2000