Gonna show my ignorance here, hope yall appreciate it. I have a 914 2 liter 4 cylinder, and like everybodys, it does have some injection deficiencies, why can't i go down to the local junkyard (or dealer when i win the lotto) and find an injection system off a ford, chevy, mazda, vw, fiat or similiar car that has 4 cylinders and is a 2 liter and hook it up to the 914, granted the first system i find may not be the best, but if i look enough (or just plain get lucky) and find a 4 cylinder 2 liter that had a manifold pressure system won't it work? and really couldn't even a non MPS work if i got all he pieces? these cars are old enough that couldn't you actually look at the engines systems as separate entities? just my thoughts on how to get an ill (but running) car back on the road simply and cheaply! O.K. BASH AWAY, i'm holdin on.

the problem is that you would have to find a donor motor that has almost the same exact VE* tables thoughout the rpm and manifold pressure ranges...if you found one that is close, you might be able to fake it or tweak it

the chances of it running perfect is pretty slim, more than likely it'll run too rich or worse, too lean

the max VE of the type IV is approx. 70%...basic term would be that for each time the motor intakes an air charge, the cylinders are only being filled up to 70% of their capacity

the problem that can happen is not being able to fully modify or tweak the new FI..that is what is so nice about aftermarket FI such as Haltech or even the megasquirt, everything is user tweakable...once set, it should be set-and-forget

*Volumetric efficiency

I have talked to one owner who is using MegaSquirt and is happy with it. Some "do it yourself" work, but you don't need to break the bank either.

Why not just fix the djet?

What issues are you having with the Djet. An Ohm meter can go a long way with a Djet. Loose ground at block by intake, trigger points, TPS and MPS and HPS all can be ohmed out with meter. Bring it to Sarasota or me there.

Some type of closed loop injection might work, wouldn't it? If you install O2 sensors, then the injection system should be able to adapt. Would VE still affect this? I don't know a whole lot about FI, so I'm mostly talking out of my ass.

A CIS system would likely be the easiest to adapt. Mega Squirt is pretty easy to do as it has been done and that means resources.

Nope. Any of the OEM EFI systems will rely on a narrowband O2 sensor, so the loop can only be closed at a mixture of 14.5-14.7:1, which is fine for idling and low loads on a water-cooled engine, but is completely inadequate for good power or an air-cooled engine. Nearly all EFI systems go "open loop" past 1/2 to 3/4 throttle, as well, since you have to run richer than 14.7:1 with nearly any engine to get good power.

All of the OEM EFI systems I know of that use a wideband O2 sensor use the sensor to run substantially leaner than 14.7:1 during cruise, and the engine is adapted to run this lean safely. Such lean-burn engines are one way a lot of automakers have acheived the low emissions levels seen in some current cars. Running a Type IV this way would kill it in short order.

If you're REALLY into DIY and really cheap, there are a number of people who fooled around with reprogramming a common GM ECU to remap it and tune for other engines. Indeed, Bruce Bowling and Al Grippo, who designed Megasquirt, were active members of this group. It's a hard slog, and MS is actually substantially easier to deal with, and really isn't all that much more expensive (considering the tools you need to reprogram a used ECU).

I'll generally agree with the position that you should try to just fix the D-Jet. Stock and properly functioning, it works well, and has no serious "deficiences". If you desire to do things to the engine that makes life difficult for D-Jet (like a wild cam or a big engine), then MS is a route that would be much easier than trying to adapt a junkyard system to the Type IV.

CIS is a nice option, and you can pick those systems up in a junkyard in operating condition (look in any 80s VW). Unlike EFI (CIS is mechanical), CIS really can adapt itself to another engine with the same number of cylinders, since it measures the airflow into the engine at all times. The curve of airflow to mixture is somewhat tunable (at least the "height" of the curve). The downside of CIS is that you can't run a cam any wilder than what D-Jet will accept, so you're really just swapping one relatively limited system for another.

Megasqurt has a new code that will run closed loop with a wideband O2 meter. It has an 8X8 table of target air fuel ratios. Manifold air pressure and RPM are the axis. I'm gonna try this but, to be honest I don't expect the O2 sensor to last too long. The previous ones I have had died a ghastly death because our cars run soo rich.

To be honest, I really think our cars will run just as well open loop. The problem is that means the fueling map must be tuned very well as there are no corrections on the fly to compensate for a bad area.

My new system will include rev limiter (fuel cut), over run fuel cut, and wasted spark ignition with multispark at idle. There are lots of other options that can be included but, I'm not going to include them

Thanks for the replies guys. Both of them make sense. I never really thought about how an air cooled engine liking to run so rich would affect a closed loop system.

Another option is the Kit Carson FI system, which sounds like it should be out pretty soon, and probably will involve the least amount of work for installation and setup.

If anybody wants to try running a CIS system, I think I have about 3 or 4 complete system for 1.7 and 1.8 VW Rabbits kicking around.

Keep thinking Ford. I have some very cool information I will share soon regarding Ford EFI systems.

Lets say this now: They have a higher resolution than a Motec...



B

whatcha got

Be nice if you could chip some of that stuff. Maybe sequential injection and coil on plug ignition.

The distributorless ignition with my MS is all Ford stuff. The EDIS module controls dwell and has the coil drivers. Later versions of the Ford EFI have the ignition module in the EFI brain so you would have to reprogram the whole thing instead of supplying a single command for advance.

Dave

Howdy, the reason i posted the original Q, was it seems everytime i get the car running (dare i say) good, another "system parameter" changes (or just quits) resulting in an embarassing push to the edge of the road or lots of middle fingers directed at me but pointing to the heavens. I really do enjoy driving my 914 when it is running well. Just thought that a fresher complete system may stay more stable longer. thanx for the thread.

just a thought:


With the LM-1, on a bone stock, good running 914 with D-Jet or L-Jet (or carbs even), the LM-1 can do a datalogging on a running engine...this would give you a VE table which should be darn near perfect to get an aftermarket FI up and running from the start (such as Megasquirt) without having to do any "real" tuning. It should be close enough for a total plug-n-play solution.

A newer TPS would have to be fitted to the throttle body (you still need the factory TPS for the D-Jet/L-Jet to function) and an external MAP sensor would have to be plumbed into the intake plenum housing. The LM-1 can datalog the factory intake air temp and rpm, engine head temp plus a few other inputs.

Basicly it would be like doing a dyno run to get all the information of what the engine/ECU is doing and just copy it to the new FI system.

If you did hook up an LM-1 on a car that runs well to log data, and then got all of the VE tables, temps and other variables right for another FI system, is it possible to go to the junkyard, pull off a 4cylinder Ford EFI and have the brain reprogramed with type4 parameters.

If this is possible, (special stuff is obviously needed to reprogram) how much do you guys think it would cost just to have a new program loaded. I don't care about the actual program writing because I've programed tons of stuff from lego's to actual robots (in labview) so I think I could write the code. I only need a brain to monitor the systems and make quick injection changes.

Sorry for rambling; I gues my main question is can the brains be reprogramed and adapted to our engines.

Jim.

yes, if you knew how to modify the Ford ECU....more stuff is out there for burning new tables into the GM and Motronic brains...but with all the aftermarket chips available for Mustangs and such, someone has to know how to do it...I've just never researched it much...there was someone that I "thought" was going to use the FI from a Saab and throw on a turbo for a type IV, I'm guessing the Saab ECU was "tweakable"

I remember someone talking about a saab FI. Ford just inerests me because I have more access to it, butknow I think I'll look into the saab stuff

how about using a suby fi system ( obd not obd11 ), the ve should be close as the engine and intake system are similar. the suby system is programmable, as well as i think if you dont hook up a speed sesor they run closed loop. the suby system should be just about bolt on as far as the tps and airflow meter. the hardest part might be the crank and cam position sensors, which could be tied in to the dizzy, and the suby does spark as well. so the dizzy is avalible to hook up the cam sensor buy using a magnet from a petronox and a stock suby cam sensor mounted into the stock dizzy body.

also with the o2 sensors i wonder if they could be tricked in to running richer with an air bleed right in front of the sensor and they might even last longer.

I have looked at using the Honda PGMFI system. It is OBD-O so no crap to deal with. I have a chipped ECU already and can adjust for bigger injectors, 3 bar MAP sensor, hot and cold rev limit, clutch cut, traction control, data logging, etc.

It is already a four cylinder, it is a speed density system and there are plenty of freeware programs to hack it. The only thing that had me stumped was the ignition drivers and signal output.

You can get them for free and parts are plentiful, so is knowledge.

subi is watercooled.....it can run a much leaner air/fuel mixture......if it does not know to increase the fuel, kiss the motor goodbye...........

Brett, can you point us to a site for hacking the honda stuff?? not that I know what I'm doing or looking at, but it would be interesting to read it

yes i know suby is watercooled i have one hanging in my 194 by rope

but its programmable so it could be made to run richer, like i said i even think it runs closed loop till 20 mph or something like that, so if you dont hook up a speed sensor it stays closed loop. so bigger injectors or higher fuel pressure could get it to run richer, or reprogram it richer

what is the cost to reprogram them? if it's in the thousand dollar range for the system then it's too expensive for this particular application (FI on the cheap, or reasonable )

not really sure i think some aussy's on the subaru vanagon yahoo chat have it, i will look there and email it to ya

Dave,

Do you have details posted on your EDIS implimentation? Like what car you got your trigger wheel from and mounting location.

-'Chung

Chung

It's on the bench right now. With any luck it'll go in the car over the holiday. The trigger wheel and all the EDIS parts are off of a 91 Escort.

Dave

http://www.pgmfi.org/twiki/bin/view/Library/WebHome
The basic low down.


http://www.ecucontrol.com/
For Data loging and other stuff

http://forum.pgmfi.org/index.php
Forum for knowledge

http://www.turboedit.com
One of the better freeware programs.

Fix the OEM Djet how bad could it be?

thanks Brett.....pretty cool, this is pretty much how Megasuirt started, but it was with GM ECU's instead....