D

Has anybody been following the thread on Pelican regarding CIS conversion to EFI using a "Megasquirt" computer

The most practical setup seems to be:
Keep the existing intake. Get rid of the CIS head and fuel delivery.
Replace the Airbox with a simple cone Filter directly to the turbo intake.
Modify the injector blocks to hold new EFI injectors.
Build a fuel rail.
Replace the WUR with a generic fuel pressure regulator.
6 cylinder EFI control computer from Megasquirt
Misc wire and connectors.

The whole setup would cost about $1200.

Geoffrey

FWIW, You get what you pay for...

What about ignition?
What about tuning?
What about boost control?
What about idle control?

Steve Weiner-Rennsport Systems

I must admit I question that as well, based on my experiences with various EFI systems over the past 10 years.

Even though there are wider choices of EFI systems out there, there are no shortcuts,.......

The consequences of an improperly equipped and configured EFI system on a Turbo would be VERY expensive and there are no cost savings in that.

D

For the sake of discussion!... I agree the consequences of a an improperly equipped system would be expensive. The system mentioned is tunable (hooks to a laptop) over the entire RPM range, it stores an 8 by 8 fuel map and according to the specs the processor is fast enough to handle 16,000 RPM. The savings is mostly in the "do it yourself from off the shelf parts", even the computer needs to be assembled!.
As for boost control... is that not a seperate issue and not a necessity (if your not trying to run excessive boost) as long as the system can provide enough fuel for the amount of air.
Also does a system have to "control" ignition or can it as this system does just monitor it. Not to over simplify, but if the system knows when the spark occured and how often (RPM/6), along with injector capacity and a/f ratio , it can predict when the injector needs to start and stop providing fuel.

Geoffrey

In my experience, an 8x8 table does not provide enough resolution, especially in a turbo car. Consider you will need to provide load sites from 0kPa absolute to 300kPa absolute (2bar), 300/8 = 37.5kPa increments of control, that is every .375 bar! RPM sites will be every 1000rpm. The very early Bosch EFI in a 3.2 Carrera used a 12x12 table, and even that is small. It is difficult to acheive a good quality idle with a proper A/F ratio with such large increments. Even the early Bosch EFI had a separate table with fine resolution for idle as well as an idle air control valve.

Boost control can be a separate control, however, ECU overboost protection is great.

EFI and engine management systems are different. EFI controls fuel delivery only. Engine management systems control both fuel and ignition systems. Typically, a fuel only engine management system does not provide a feedback mechanism for timing. Even in an engine management system, the ECU does not "monitor" ignition events. Ignition events are triggered by the ECU based on crank position sensor and the ignition timing values in the the table. Ignition is an output, not an input. On a turbo car, the ability to control timing is perhaps more of a benefit than the ability to control fuel.

Most inexpensive systems only provide narrow band O2 sensors and therefore, do not provide accurate AFRs for closed loop operation. In fact, the ECUs use the main tables combined with the compensation tables to provide for proper fuel and timing more than O2 sensor, even when wide band is available.

D

Great analysis Geoffrey. 2bar! yes that would require better map resolution! and a wide band O2 sensor would have to be part of the equation.

Geoffrey

A wide band O2 snesor is not a requirement, but if your EFI system logs AFR, then it is an invaluable sensor. I can't think of a car where I've used an O2 sensor for closed loop and isn't required for running.

D

So which systems are "acceptable" for a turbo Porsche.

anh911

I agree with Steve and Geoffrey, the hardware is the easy part. Getting the system to work well is much much more difficult. I love the Megasquirt in principle - the linux of FI - but in practise it wasn't designed for what you want to do with it.

Geoffrey

...Being a MoTeC dealer I would have to say MoTeC...

I feel the most important part of the EFI conversion is to find a tuner with a good reputation regardless of the system. I've seen so many "bad" tuning jobs using all kinds of units, epsecially the tuner's who claim to have a plug and play program for your car.

In general, you want to look at systems that have:
- full sequential ability
- good table resolution
- Wide band lambda support
- proper compensation tables
- data logging
- inputs and outputs sufficient for your application
- flexible trigger sensor options
- sufficient ignition outputs for your ignition system

Most importantly, you get what you pay for in both the tuner and the hardware. You will want someone who suggest and builds into your quotation 1 day of dyno time to tune your car.

D

Ideally I would want a system which works with my existing ignition system (MSD). Makes use of any existing sensors, except O2 would need upgrading to wide band. Allows use of the existing intake manifold. Is easily tuned through a laptop and stores more than one map,(power and economy) and can easily switch between the two. And also allows good idle control. I can build my own rails, and wire it up if the instructions are reasonable. One of the systems I have found nearly meets these demands EMS-Q4, but the EMS company in Australia never returns emails. The other option is the TEC-3 unit but it is more costly because of the distributerless ignition which I think is required for this system. Are there others out there which are proven... how does the Mo-tec compare to these.

Geoffrey

the 87 Turbo will have no sensors that you can reuse in an EFI conversion. If you want to keep the restrictive 930 manifold you can purchase injector blocks and fuel rails made for EFI injectors, and I'm assuming that you'd want to retain the stock single plug ignition powered by your MSD CDI ignition. If you were looking for big power 500+hp, I'd go with the 3.2 Carrera intake system with proper port modifications and twin plug ignition. You then need to figure how to trigger the engine management computer and build a harness which can be frustrating especially if you don't have the proper crimping tools and connectors.

MoTeC is recognized as a leader in engine management and produces race quality components and wire harnesses with all the cost associated at that level. The systems are very flexible, reliable, and simply work very well.

richard glickel

Hi Geoffrey,

Happy New Year! Would you recommend MoTeC for my 3.6? What would we need to add to my set-up to make it work?

See you soon.

Richard

P.S. Will you be attending Zone 1 Tech Tactics next month?

Geoffrey

Richard,

Happy New Year... Unfortunately most likely I won't be at tech tactics with the Sebring race, and some other projects I have going on as well as my own car. I would recommend that you just learn to drive and enjoy your beautiful car. It is largely stock and without internal modifications, or move to a turbo engine, you'd have a small return going to a MoTeC setup.

P.S. I'll be down there today working.

NineMeister

As a Motec agent / Porsche specialist in the UK I agree with most of what Geoffrey has said. We fit Motec regularly to the 964 3.6, you can see the results on the following link:

9m Racing 964 Motec conversion

We are also just about to fit a similar system to a 965 3.3using Carrera inlet manifolds, if I get time I will post the results when we have them.