Part Two - Throttle bodies and pressure regulator:
Part one of this article described how to create a high pressure fuel supply into the engine bay, and a return to the swirl pot, and subsequently fuel tank. Let's now move our interest to the front of the car.
The first consideration is which throttle bodies to use. The budget constraints really leave no option other than motorbike throttle bodies. There's a whole variety that could be used and I guess that with a little improvisation any with an internal diameter in the 36-45mm range could probably be made to work, but there are definitely some that are recognised as being easier options than others. The favourites tend to be those from early (k1 and k2) Suzuki GSX-R 600 and 750 variants. Many bike TB's are cast in pairs and do not readily adapt to the port spacing of the Zetec (or any other larger bore engine than that for which they were designed). The early Suzuki TB's however are made individually and can be respaced to suit. The 600 has a 38mm body diameter and is generally reckoned to be perfect up to around 180-190 bhp when it starts to limit the airflow. The 750 has 44mm throats and should be ok up to around 230-240 bhp but will not give the low down flexibility of the smaller unit. For what it's worth, I used a low mileage set of k1 GSX-R 600 throttle bodies.
Before we move on to actually modifying and mounting the throttle bodies I want to discuss the objective of the project as this will determine some of the finer details. At one extreme there is the aim of just getting the TB's fitted as quickly as possible with the least amount of work and getting the car running without worrying too much about completely optimising the performance, or bothering too much about what it looks like. At the other extreme there's the aim of creating an installation that gets maximum available performance and incorporates best engineering practice, but may well encounter some genuine technical challenges along the way. And then there's the middle ground solution, which performs well, looks technically sound, but contains performance or engineering compromises. My aim was certainly towards the best engineering practice end of the spectrum, and one of the constraints (and hence this photo) was that the air cleaner after the TB's were fitted had to end up in exactly the same place as it was when the Weber carbs were fitted, as I had a hole in the bonnet to fill!
Before getting into further details of the installation on my car let's consider the available options should one be prepared to compromise on the optimal engineering solution (and there's nothing wrong with that, it all depends what you want from the package).
The first option is to leave the TB's together with the throttle linkage, throttle position sensor, etc as removed from the bike. To install on the engine will require a manifold, probably just the mating flange to the head with some stub tubes, onto which can be clamped some rubber tubes which can in turn be bent to connect up to the TB assembly. The downside is that the fuel won't be sprayed directly onto the back of the valves but into the air stream that will have to follow the route of the tubes, so there will be a performance compromise. There may be sealing issues and there will need to be some type of mechanical support as the rubber, or silicon, tubes won't in themselves be robust enough to take all the weight and vibration of the TB assembly. On the positive side the tricky issues of respacing the throttle linkages etc won't be necessary.
The middle ground, and probably the most common solution, is to respace the TB's to suit the porting but still use a stub manifold and flexible couplings to connect to the TB's. The advantage is that the air flow is much closer to optimum. The downside is that there's still a potential sealing and vibration problem, and some (including me) would consider four lengths of hose with eight jubilee clips instinctively can't represent the best engineered solution.
As well as having to position the air cleaner in the same position as it was when fitted to the carbs, I also wanted to leave the Weber manifold intact just in case I ever wanted to revert back to carbs. The solution I therefore chose was to design, and have made, some simple sandwich adaptor plates that bolted to the Weber manifold on the one side and would allow the TB's to be rigidly mounted to the other. This would ensure perfect spacing between the TB's and position the air cleaner in the right place.
The individual TB's were a good fit in the bores in the adaptor plates, and with a smear of high temperature sealant are completely airtight. Let me mention an occasionally overlooked aspect of the installation, and that is the consequences of thermal expansion. There's a lot of heat around an engine, and especially during warm-up there's quite a lot of expansion going on. If you are trying to seal a join with different materials the relative expansion between them can cause movement and leaks. In this installation the head, manifold, adaptor plates and throttle bodies are all aluminium, so minimising differential expansion between them and maintaining airtight sealing between the components.
I think in hindsight I could have made them a couple of millimetres thinner and profiled them a little tighter around the periphery, and indeed I may yet if I ever have to remove them, but 3000 trouble-free miles on from the initial installation I'm going to follow the old adage, if it ain't broke don't fix it.
You can see from this photo the extra spaces between the TB's when they're mated to the Zetec head, and to bridge this gap necessitated making some spacers to go between each TB, getting the threaded link rods extended and welding some additions to the intermediate throttle butterfly arms. As I mentioned earlier, to get the most from this installation does require some effort and attention to detail. There's another feature that can be seen from this photo and that is that compared to the installation on the bike there is a throttle butterfly missing from each body. On the bike there is a butterfly just inside the throttle body which has been removed and the bushes filled with a high quality plastic metal. I have heard these erroneously referred to as the choke butterflies, and indeed, they do look like the chokes on carburettors, however, of course, fuel injected systems don't need chokes, they compensate for cold conditions by increasing the amount of fuel delivered to each injector. These extra butterflies, which are electrically actuated by a motor, again removed in this photo, compensate for an airflow characteristic of the Suzuki TB's just off idle. We'll come back to this in greater detail in the section on engine tuning (accessed via the Westfield page, Megasquirt link).
The final touch to the TB fitment was to design and get made some round adaptors that mated the outboard end of the TB's to the original Weber trumpets and the ITG air cleaner backplate.
The next consideration is getting fuel at the right pressure to the injectors. If you recall we have fuel at high, but unregulated pressure, coming from the back end of the car through an 8mm copper pipe. This has to be connected into a fuel rail that fits onto the injectors and has a connection back into the return pipe. This return pipe then has to go into a pressure regulator which maintains the pressure in the fuel rail at whatever we choose to set it at. Firstly let's consider the fuel rail, and this is probably the trickiest engineering challenge to overcome (with the GSX-R 600 TB's especially). I mention the 600 in particular as the bike installation uses a one piece cast aluminium fuel rail that doesn't lend itself readily to respacing. The 750 uses a fuel rail made from multiple parts and lends itself more readily to respacing. Getting back to the 600 fuel rail, you could split it into sections and thread some extra hollow spacers into it to lengthen it. You would also need to add a return end fitting as the bike installation doesn't actually use a return, which isn't appropriate in a setup like this. Personally, having some access to fabrication facilities (and an ace fabricator...) I designed my own and had it made. There's no doubt the best solution is the one with the least joints to leak. I can't imagine it would be that expensive to get made even by a commercial fabricator, it's not that complex when you get down to it. It's also possible to buy billet fuel rail sections and these require machining rather than fabricating to fit, so I guess the choice may come down to what skills you and your friends have, and what access to manufacturing equipment.
The mounting brackets were cut off the original fuel rail and welded onto the fabricated one, other than that it was made from scratch.
You can also see in this photo the manifold pressure sensor tubes under the throttle bodies. There are four, one from each TB, which are linked together and go off to the manifold pressure sensor in the ECU. The third throttle body from the front has a second manifold pressure tube which, if used, connects to the fuel pressure regulator, which brings us in a rather contrived fashion to the prv (pressure regulator valve).
As mentioned previously this fits in the return line from the fuel rail and maintains pressure in the fuel rail and therefore available to the injectors. I used a Power Boost valve from ebay (new and at about one-third the price of one from a retail store). You can connect these with the pressure sensor tube in the top cover vented to atmosphere and they will work like a normal prv, attempting to maintain a constant pressure in the fuel rail (if you do this the additional manifold sensor tube on TB number 3 will have to be sealed). However, the full functionality is achieved by connecting the manifold pressure tube to the cap of the Power Boost valve.
The way this is works is to adjust the maintained pressure such that the difference between the fuel pressure and the manifold pressure is constant. What this means in practice is that when the throttle is opened from an idle or low throttle condition the absolute fuel pressure increases which in turn gives an increased burst of fuel from the injector which manifests itself as "crisper" acceleration. The same effect can be achieved electronically via some ECU's, and whether the full features of the Power Boost valve are worth the effort or not polarises opinion. I think it may well depend on the quality of the manifold pressure signal, mine is particularly smooth, so the prv seems to work well. I know of similar installations where the transients between each cylinder firing completely obliterate the quality of the manifold signal, so there's no logical reason under such circumstances why it should be of any benefit.
Anyway, while we're on about manifold pressure signal, it might be worth, if you intend to use it as the load indicator from which to determine the fuelling, putting a small orifice, such as a MIG welding nozzle, in the sensor hose to just dampen the fluctuations a little bit. I have heard of people putting welding nozzles, plus fuel filters, plus accumulators etc in the sensor hose to dampen the signal, but there has to come a point where the signal that is seen by the sensor just bears no resemblance or relevance to what is actually happening in the engine so the fuelling value becomes a bit of a lottery.
So, here's the whole lot fitted together. I had to make some small mods to the throttle cable bracket to clear the alternator. There's an air temperature sensor fitted into the air cleaner backplate. I've used the throttle position sensor exactly as fitted to the original bike. There's also a couple of extra braces added between the adaptor plates and the air cleaner backplate.
A brief word about testing the system. A leak in a high pressure fuel fitting can be potentially quite dangerous, so I would suggest firstly have two fire extinguishers to hand plus someone to help. If you have retained the low pressure pump, disable the high pressure pump and just pressurise the system with the LP. Check all the connections for any signs of leakage and fix them before attempting any high pressure testing. When there are no leaks at low pressure, enable the HP pump and just run it for a couple of seconds. Look around for any signs of leaks, also note any smell of petrol as a pin hole type high pressure leak can emerge as a spray and this is potentially the most hazardous leak. Before starting the car attach a gauge into the return line (the Power Boost valve has a port available) and with the HP pump running but with no ignition, set the system pressure - usually around 3 bar. Next, and still with a spotter and a couple of fire extinguishers, go for it, and good luck.
It might be useful to have a look through the article on the Megasquirt even if you're not using a Megasquirt ECU itself, as there are some additional tips on tuning the engine, especially with the Suzuki bike TB's.
If anybody has some notes that could be added to this article, or some photos of different approaches I'll certainly consider adding them if it would help clarify the process.
Part 1: Fuel tank, swirl pot and pump mods.