This article was first published in 2003.
There are plenty of magazine articles around on fitting intercoolers. In fact we've covered them ourselves: a workshop story where a shiny new intercooler is plonked into place. But this story is right at the other end of the spectrum: it's about installing a secondhand intercooler, doing the plumbing yourself at home, and then improving efficiency of the system as much as possible.
The car in question was my 1988 Maxima VG20ET turbo, but the points are applicable to pretty well any turbo car that was sold non-intercooled.
The first step is to decide on the location - and then pick an intercooler to suit. My first inclination in the Maxima was to place the 'cooler inside the front guard (fender). It's a common position on cars that run less than monster intercoolers from the factory. But why not in front of the radiator, where nearly all aftermarket intercoolers go? In the Maxima, as is the case with many cars, putting the intercooler in front of the radiator would have been a very major job. Not only would plenty of body cutting be needed to get the plumbing through, but the grille and bumper would also need to have been cut away.
Simply not worth it for a budget, Do It Yourself job.
So in the guard, then. But the space available in the Maxima proved to be very narrow and short: 'narrow' meant that the core couldn't be put in sideways, and 'short' meant that it would project out the bottom if it went in vertically. Well, that was my supposition with an intercooler of about 350 x 200 x 60mm dimensions, anyway. Getting the plumbing through into the guard would have also needed plenty of cutting, in addition to the moving of the battery and some relays. However, given that a battery move is straightforward and that it's easy enough to lob up at a crash repairer and ask them to cut out the marked hole with their plasma cutter (if access is easy it would probably cost you nearly nothing at all), the in-guard possibilities were at least feasible for a home-based DIY'er.
About that time I came across an ex-Skyline RB25DET intercooler. It was given to me from the bin of a workshop that fits lots of aftermarket intercoolers - that gives you an idea of where perhaps to go looking for a cheap or no cost core... Now that I had a real, live intercooler I again tried it in the guard, but this time - with the addition of the actual plumbing connections on my mythical box - it looked even more difficult.
I then decided that the Maxima intercooler was going to be located under the bonnet. Taking out the battery gave a spot pretty close to both the turbo and throttle butterfly (the V6 is transverse), which would result in fairly short intercooler plumbing. (The short plumbing will reduce lag and also lessen the chances of the charge-air being re-heated on its way back from the 'cooler.) However, while it could be fitted in the spot, the RB25DET intercooler was just a touch too big - but maybe by only 25mm in the long dimension.
Picking an Intercooler
It was time to visit a wrecker. I lobbed up at the local Japanese import wrecker with three things on my list - an intercooler a little smaller than the RB25DET one I already had, a selection of hoses and bends, and a bonnet scoop. But before I got stuck into that list I enquired as to the price of an RB25DET intercooler - if they had one. In addition to getting mine for nothing I knew that in Adelaide - the cheapest wrecker city in Australia but a long way from where I live - I could pick one up any old time for AUD$100, so I had a benchmark to price against. (If the intercoolers at this wrecker were going to be stupidly expensive, then I wanted to know first - rather than spending 20 minutes digging through a shelf of grubby intercoolers only to walk out empty-handed when I heard the dollars being asked.)
And when I heard the price on the Skyline RB25DET intercooler, it just about knocked me flat.
Two Hundred and Fifty Dollars!
After I had recovered, I suggested to the staff that I may as well leave without looking around any further. However I was encouraged to have a dig through the pile of intercoolers and was told that a price would be suggested depending on what I found.
A more realistic price, I hoped.
So what was I looking for? AutoSpeed has run some very good secondhand intercooler comparison tests (eg "Cooler Flows"), but in this case I knew they weren't going to be much good to me. Why? Well when I had access to only one local wrecker (the others are in the next major city, 100km away), and the space constraint was tight, it was almost certainly going to be a case of whatever was the right size, was available, and was reasonably priced was going to be what I got! However, I was looking at one other characteristic - plumbing size. If the plumbing is decent in diameter (eg 60mm) then it's also a reasonable bet that it is from an engine that flowed plenty of air - and so developed some good power. In turn this is likely to result in less pressure drop across the core, which means that more of the boost developed by the turbo gets to the engine.
But why was I happy with a small core? Didn't I need a monster intercooler? Two points: firstly, having any intercooler (one that doesn't cause much of a pressure drop, anyway) is better than having no intercooler at all. I'd measured the Maxima's standard intake air temp at no less than 85 degrees C on a 20-degree day - with just 0.45 Bar boost! That's a scarily high intake air temp. Secondly, even a small core can have its efficiency improved, as you'll later see.
Anyway, there were about 20 intercoolers for me to look through, and after narrowing the selection down to four I picked this one. Including end-tanks it is 280 x 270 x 80mm with a good sized plumbing diameter. It also wasn't an intercooler that I recognized, so I thought that its obscurity was likely to make it lower in cost. Orphans - especially small orphans - are always going to be cheap... Hmm, maybe that needs to be rephrased!
This time the quoted cost was AUD$88 - much more reasonable.
I then asked if they had any plumbing - hoses, bends, clamps, etc. Luckily they had two 44-gallon drums full of bits and pieces, and I spent a good 30 minutes looking through them. What I was after was factory post-turbo (rather than pre-turbo) plumbing. (The hoses and fittings designed for connecting the airflow meter to the turbo, or the airflow meter to the filter box, aren't designed to handle pressure.) Obviously the plumbing also needed to suit the diameter of the intercooler pipes and the throttle body pipe. Since on the Maxima - as in most cars - the throttle body is much larger in diameter than either the turbo exit plumbing or the (new) intercooler plumbing, I also looked for adaptors that changed in diameter. (But why go to this trouble? Why not just make all the new pipework? I'll cover that point a bit later.)
After a while I had a good selection of plumbing bits and pieces (not all are shown here, just the ones I ended up using), and a price - including the intercooler - of AUD$120. That's more like it! However, the wrecker couldn't help me in the way of a bonnet scoop. No matter - that could come later.
The new intercooler fitted into its chosen underbonnet space far better than the previously-tried RB25DET unit - so that was one hurdle overcome. The next step was to organise all of the hoses, bends, adaptors and clamps (oh yes, I forgot to say - pick up lots of clamps as well from the wrecker; you'll never get them cheaper!) into something that might channel the air from the turbo to the intercooler, and then from the intercooler to the throttle body.
There are a few tips when assembling the pipework. Firstly, if a rubber tube is a little tight to get onto a fitting, use a heat gun to soften the rubber. It can then be pushed over the fitting without splitting the hose or drawing too many beads of sweat. Secondly, use a spray lubricant like WD40 and the assembly process becomes even easier.
After trying all the permutations and combinations I could think of, I ended up with a satisfactory system - the aim was to have pipework that had as few joins and bends as possible. Once in place, the intercooler had almost nowhere it could move, so just one bracket was used. This went from the bonnet locking platform to the forward edge of the intercooler, and was made from 4mm thick aluminium strip. In this pic the intercooler and its plumbing are just temporarily in place - explaining the lack of hose clamps and paint.
Making the intercooler plumbing from a selection of straight hose and rubber bends, from factory adaptors and metal pipes, means that you'll end up with plumbing with lots of clamps and lots of separate sections. A custom bent-and-welded stainless steel series of tubes looks much nicer and probably also flows better (depending on how many sections it's made from and how much the welds penetrate the walls of the pipe).
But taking the bits-and-pieces approach is just so much easier to do at home. The rubber bits have a degree of flexibility, so if things don't line up absolutely perfectly there's no drama in bending them a bit. Bends can also be rotated, allowing the other end of the pipe to move through an arc. It's easy to adapt different pipe diameters using rubber fittings that change in diameter, and most factory turbo plumbing has at least a few very tight bends, making it a lot easier to fit things in. Importantly, those bends are usually well made with a good radius on the inner part of the bend (rather than a 90-degree square edge that creates turbulence and so drops flow).
But you can of course buy a selection of mandrel bends in stainless steel and use a hacksaw and a file to fit them together into a plumbing system that works. But with only hand tools and basic power tools it's a nightmare job - and then you'll need to take the assembly somewhere to get it welded.
If you're on an extreme budget and you don't have too many tools, much money or patience, build the plumbing out of bits and pieces.
The first step in improving the efficiency of the small core was to fit it with a fan. The fan, which is triggered when the vehicle speed is low, keeps air flowing through the core when otherwise it would have stooped. (Be careful that under-bonnet intercoolers don't become pre-heaters when the car is stationary or moving only slowly!)
A 12V motorcycle radiator fan is a good choice - small, weatherproof and powerful. However, on this budget job I had to make do with what I had - and that was an ex-Daihatsu radiator fan. Trouble was, the fan was much too big in diameter. Mounted underneath the intercooler (ie sucking air through it) the plastic blades would have hit the bottom-exit plumbing. Even when removed from its frame, the fan blades were just too big - the easy answer to the problem was to shorten them. This reduces airflow, but there's still plenty of air being pushed along.
(Computer fans are often looked at for this application, however there are real question marks over their durability at higher underbonnet temps and how weatherproof they are. Not to mention that most flow little air and do so at a low static pressure - so their flow just about stops when you put any obstruction in the way.)
I removed the blades from the motor (just a simple nut to undo) and then used a hacksaw to take the same amount off each blade. I then smoothed the ends with a sanding belt. (A file could also have been used.) To get the balance back as close as possible to standard I inserted a Philips head screwdriver through the hole in the centre of the fan and then balanced the shaft of the screwdriver on a horizontal edge, so that it could roll along. The blade that consistently ended up pointing downwards I lightened a little by removing some more material from the end of the blade. It only took a few minutes to get the balance reasonably good.
The next step was the most involved - making a shroud. It's no good having a bare fan (especially one with shortened blades!) just sitting under the intercooler core - it will tend to suck air from wherever the supply is easiest... and that's almost certainly not through the core. Instead you'll need to make a shroud, just like the one employed on radiators. A shroud forces nearly all of the air that is being moved by the fan to come through the core.
Using hand tools (hacksaw, file, bench vice) and a few power tools (electric jigsaw, electric drill) I made a shroud from aluminium sheet. The materials didn't cost me anything (or anyway, a dollar at the most) because I used sheet alloy taken from my scrap box. The scrap box is replenished by visiting non-ferrous scrap metal dealers - there you can buy aluminium sheet offcuts by the kilogram very cheaply indeed.
The fan needed to be mounted slightly towards one end of the intercooler - but that doesn't matter because the shroud will ensure that air is drawn through the core across its full area. Once I had finished making the shroud, I connected the fan to a battery and then checked the flow through the core. The most dramatic way of showing the flow is to lay a sheet of paper across the top of the intercooler core - in this case the sheet was immediately sucked flat.
The Bonnet Scoop
Having a supply of cool air to the intercooler is vital. And just as important is having a low pressure exit. The exit was provided by the undercar flow of air passing beneath the engine bay, but what about the feed? A bonnet scoop is the obvious answer and that's what I went for. However, quickly and easily obtaining a simple 250mm square (by say 40mm height) scoop proved difficult. One company said that they'd be happy to make me a fibreglass one for a reasonable AUD$60, and to come down to their factory to look at pics of various designs. When I got down there, they didn't appear to have any scoops of the desired dimensions - but would one about twice the size still be OK? Well, no.
Instead I changed tack (pun, that) and went to a marine supplies store. There I bought a chrome-plated 'Transom Ventilator' for AUD$30. I admit that it's not to everyone's taste but it suits the Eighties chrome highlights on the Maxima as well as easily bolting to the flat bonnet panel. But what about its size - isn't it way too small? In fact its opening has an area of about 55 square centimetres - a bit larger than the cross-sectional area of a 3-inch duct. A little like radiators and modern cars' grille openings, the intercooler can be much larger than the area of the scoop opening without heat exchange suffering much. Also, the fact that the scoop stands 55mm high means that the flow into it could be expected to be greater than a scoop that stands - say - only 20mm high and is proportionally wider.
Anyway, after buying the scoop the first step was to cut out the plastic base-plate to increase the air throughput - it is designed to connect to a 3-inch hose and has a plastic flange to suit. This opening enlarged, the bonnet was then marked and a jigsaw used to cut out the metalwork. A piece of chipboard was placed over the intercooler to protect it and the bonnet was propped open with a block of wood. The paintwork was protected with wide masking tape and the motion of the saw was eased with some penetrating lubricant. One bonnet support rib needed to be cut through, but the fine-tooth metal-cutting blade did this without a worry.
I initially left the bonnet sound insulation in place but when I saw the tight clearance to the intercooler that existed with the bonnet shut (amazing what you learn when you can look through a hole!) I removed the insulator. It might be better to remove it first anyway, cutting it separately. The raw metal edges of the bonnet were then painted over and the holes for the scoop retaining screws drilled. The scoop was then bolted into place.
In the same way as the fan shroud makes sure that all the air being moved by the fan flows through the intercooler, the pressured air being provided by the scoop must also all flow through the core... not around it. To make sure that this occurred, I used high density scrap foam offcuts bought from a recycling shop, cutting them into strips with an electric carving knife. Little crescent-shaped cut-outs were needed to clear the bonnet ribs and the water spray nozzle.
A spray can was used to paint the foam black and then using contact adhesive, the strips of foam were glued to each other and then to the inner of the bonnet.
Testing showed that the intake temps in normal use have fallen nearly 40 degrees C! Instead of the intake air temp being 75 - 80 degrees after about 15 seconds of boost they are now 35- 40 (on the same 20-degree C day). The temperature with a burst of boost rises by only 3-4 degrees, rather than 20 or 30 degrees. Additionally, the measured peak boost is exactly the same - the intercooler is posing no measurable restriction at all.
However, stay into it up a long hill and after about 20 seconds of boost the intake air temp with start to rise. At the time of writing the highest I have seen is 50 degrees - still vastly better than before. But that temp rise after a longer period of boost is just the sort of thing that can be addressed with an intercooler water spray, which we'll be fitting next week.
(At this stage the fan has not been used at all: its inclusion was primarily to allow the water spray to work better at low road speeds. I live at the top of a 20km country road climb and the thought of being stuck behind a slow truck all the way up the hill meant that the spray-and-fan combo was picked.)
In addition to being cheap and doable, in this application the 'cooler is performing very well.
A valid criticism that can be made of this story is this: why not simply put a full-size aftermarket core in front of the radiator? Forget the bonnet scoop, forget the mix-and-match plumbing, forget a water spray. Just do a proper job in the first place!
It's a valid point but it forgets two important things. This intercooler cost very little - in fact, it owes me $159 complete. That price includes the intercooler core, the plumbing, three extra hose clamps and the scoop. It doesn't include the bits I already had - the fan, the aluminium sheet, the spray paint, the nuts and bolts. But even including those bits as well, the total would still come in under $200.
These prices are chicken-feed compared with doing it any other way (hmm, maybe a DIY water injection system excluded!). Anything commercially available and fitted is likely to cost you a minimum of double the dollars, if not three or even four times the amount.
The other point the criticism forgets is that all this can be done at home - in your driveway or in your shed or in your carport. No welding is needed, no plasma cutter, no pipe bender.
And in the real world that's a helluva advantage.