Replacing the Type 3 EFI
With the new engine for the Razoredge now procured, I've been looking towards what I need to do to it to get it up and running in the razor. The engine itself is a standard Type 3 fuel injected unit that is currently still in the '71 fastback in my garage. It leaks a little oil, but runs okay, and shows 80k on the odometer. this could possibly be 180k or maybe even higher, but this doesn't bother me as it runs fine, and I will eventually rebuild it anyway.
The choice of a fuel injected unit as opposed to a naturally aspirated unit is, to me, a no brainer. EFI is far superior in terms of low range power, fuel economy and drivability. Not to mention that they are more efficient than a carburetted engine - this is the reason that pretty much all modern cars are EFI'd - stricter emmissions regulations mean that higher efficiency is needed.
But anyway - I digress. I say that a fuel injected engine is better - but this is not regarded as the case for the Type 3. Over the years the L and K jetronic systems have come into a lot of bad press, especially amongst those who maintain thier own cars. The system is hard to diagnose issues on, as there is no diagnostics funtions available to assist in determining what's wrong. Fixing a faulty system is really a case of eliminating all possibilities, one by one, until the fault has been found. There is no laptop port to plug in your laptop, and get an instant readout of what is going on, and to many, this means that the system is considered complex and unreliable.
As I blogged about previously. the other main issue with the stock fuel injection systems is that they are not tunable. There is no way to modify the injector 'maps' to account for even simply modifications, such as fitting an extractor system. Change something on the system and it cannot compensate. Modern systems can overcome this. Either by closed loop feedback, from a lambda sensor, or by simply having the ability to be reprogrammed. This is the main reason that I want to utilise an aftermarket system, plus it is more suited for adding a turbo to later down the track.
I plan on using a distributorless system. this means adding a crank or distributor trigger, and utilising a coilpack instead of a distributor to take care of the ignition side of things. Distributorless systems have a couple of advantages over normal distributor controlled systems. Firstly, there is no electrical 'switch' in the HT (high tension) circuit, it has a direct lead from the coil to the spark plug. This means a stronger spark, and no degredation over time (wear). The other advantage is that the systems generally work in what is called a 'wasted spark' mode. This means that for a four cylinder engine, there are only two coils. This means that they fire twice each cycle, once on the ignition, and once on the exhaust stroke. The firing on the exhaust stroke brings a massive advantage, it helps scavenge the cylinder by ignighting any unburned gasses. All of this together generally means, more efficiency, and in turn, better performance.
So I set to thinking about what EFI I should use. I already have a DTA system. DTA are a UK company that produce EFI systems for aftermarket applications. They have a long competition history, and have even been used by some OEM race teams. So I could use this, but then I need to find another EFI system for the Type 4. Plus, I have already machined the trigger wheel for the Type 4 - and it's already fitted to the crank.
Originally when I was developing the Type 4, I wanted to use a Haltech system. At the time, these had a much better featureset than the DTA system, but were a lot more expensive. The only trouble is, that both the DTA, and the Haltech system cost more than the complete engine, which is more than I want o invest at this stage.
So I started to look back over the systems that I came across in my research for the Type 4. One system that came to mind is the megasquirt system http://www.msefi.com/. This is available as a self assembly DIY kit. You build it yourself thus saving lots of cash. Another system that I stumbled across is available form CB performance http://www.cbperformance.com/catalog.asp?ProductID=323. This looks to be based on an old GM ECU. From reccollection, there used to be a big following for this style of controller. It was easily available as it was commonly fitted to cars back in the 80's. The only issue with the system was that it was not real time programmable. Fuel and ignition changes needed to be blown to an eeprom, and then fitted into the ECU. This made tuning a long and labourious process. This was fine for one make ECU upgrades, where a standard map could be used, but a PITA for custom set-ups. I cannot find a lot of info on the CB system, but it is very cheap - even cheaper than the Megasquirt.
Originally I avoided these 'generic' systems, as I considered that they did not have the flexibility and resolution that I required, but for this new build, they will probably be adequate.
Two initial considerations are the type of triggering that I can use, and the method of load determination. On the Type 4 I have a crank mounted trigger wheel, but this is something that may will not be possible on the Type 3 as the cooling fan is mounted on the end of the crank. I have considered fitting a magnet to the stock cooling fan and utilising a mag trigger sensor. Also, there are already 'teeth' on the edge of the fanwheel, and it might be possible to modify these to work with a hall effect sensor. The third possibility is to find a distributor trigger - this is basically an optical wheel to replace the distributor. This is by far the easiest, but I cannot locate one as an aftermarket accessory. My guess is that they are NLA. Probably originally fitted to cars where they were upgrading old carburetted distributor based engines to EFI. This kind of setup is now integrated into the engine design, and so no longer needed. Ideally I need to pull the FI system in the fast back to bits to determine how it works, and what type of components I can re-utilise.
The type of load determination is very important. Generally there are three methods used. Throttle position, Air flow and manifold pressure. The original Type 3 system is an air flow based system, these generally use a mechanical method by which to measure how much air is being consumed by the system, such as a flap. This is unfortunately prone to wear.
The manifold pressure system, detects the pressure or vacuum within the system to determine the air consumption. The advantage of the manifold pressure system, is that it is more suitable to forced induction applications such as turbocharging.
The throttle position sensor based system, may be used in addition to all of these systems, or it may be used alone to determine how much air is being used. When used in conjunction with the other system, its function is generally to provide fuel enrichment under acceleration conditions.
With these considerations in mind, I need to check out the specs on the systems and determine which will be better suited. At the moment, I am leaning towards the el-cheapo CB system. The Aussie dollar is pretty strong against the US dollar at the moment, plus, it's cheap enough that even if it is not 100% suitible for a turbo, I could theretically replace it down the track. The $$$$'s I save could then go towards something else, such as a nice header system.
So this is where my headspace is at the moment, planning the build out, and working out which way i need to go. I've even noticed a cheap Haltech E6H system on ebay at the moment. So maybe my decision is already made?
More next time...
Interesting post on EFI trigger wheel for the type 3 - http://www.thesamba.com/vw/forum/viewtopic.php?p=5508153#5508153