Some reasons for EFI conversions tend to work out better than others.

[MadScientistMat]

We have threads pop up from time to time where somebody is wondering if they should switch from carbs to EFI. I worked at DIYAutoTune in technical support for 14 years, and some EFI conversions have turned out much better than others. I'd like to share some tips to make sure that you can make an informed decision. I'll start with a couple of big truisms of EFI conversion.

1. There are a lot of "self-tuning" EFI systems on the market, but nobody's got a self-troubleshooting system. When something goes wrong, you will need to make sure you have the right resources to solve the problem - your own skills, a shop or buddy that knows EFI, the seller's support team, etc.

2. Self-tuning systems generally only tune the main fuel table and maybe a few other fuel tables. There's often still a bit of manual tuning. And nobody has a workable way to tune the spark table. A good spark table is both the key to good power and good fuel mileage. To get the most out of EFI, you will want to dyno tune the spark table.

3. Put some thought into the fuel delivery, but don't overthink it. Just adding a high pressure pump, regulator, and return will do OK on a street car except you may have problems with it not picking up fuel below 1/4 of a tank when cornering. There are a number of ways to avoid fuel slosh issues, some of which are pretty cheap if you have the right equipment, such as welding a sump to the tank. You can spend hundreds on a custom baffled tank or a complicated surge tank system with two pumps, but you don't have to.

4. A throttle body injection system is about halfway between a carburetor with an OK-but-basic tune and multiport EFI in terms of what you will see in terms of drivability. Sometimes that may be what you're OK with; other times you will want to go big or go home. Or rather, go multiport or stay carbed.

5. "Self tuning" generally works best on mild engines and needs a fair amount of distance between the O2 sensor and the exhaust outlet. If you have an aggressive cam or some sort of really short exhaust that has a foot between the O2 sensor and outlet, expect to tune the idle and low RPM by hand.

 

[Lefty71]

Id like to think that your post is step one towards an important sticky thread, but I personally dont have enough experience with conversions to say......hopefully im not the only one. Good info tho.

 

[matthon]

Imo, none are 'plug & play.'
One has to rethink and revamp the systems to support it, electrical, ignition, charging, fuel, cooling, etc.

 

[TT5.9mag]

#6 most problems with efi installations are user induced. i.e., not following wiring instructions, lack of good grounding, using wire taps, using (god forbid) wire nuts, or having a good alternator or clean battery voltage.

Hopefully you agree with that @MadScientistMat i don’t mean to step on your toes here.

 

[MadScientistMat]

#6 most problems with efi installations are user induced. i.e., not following wiring instructions, lack of good grounding, using wire taps, using (god forbid) wire nuts, or having a good alternator or clean battery voltage.

Hopefully you agree with that @MadScientistMat i don’t mean to step on your toes here.

Those are certainly pretty common newbie mistakes. Old car wiring wasn't designed to support anything more sophisticated than a transistor radio. And don't forget non-resistor spark plug wires.

And the next item I wanted to go over: If you're considering an EFI conversion, consider your goal. EFI is a means to an end... usually. Some goals tend to come out better than others, and it's possible you may have a couple different goals at once.

• "I want to build a drag car with a lot of features." EFI can do very well here - even an entry level system like a Microsquirt can get you a two step and fan control in addition to your fuel and timing, and high end EFI systems will have sophisticated boost control, trans brake delay, progressive nitrous, and more. And you typically expect a drag car to be a major project. EFI is a great choice here. Just make sure the system you pick has all the options you might need and implements them well. (Same goes for Bonneville and land speed racing.)
• "I'm swapping in a later factory EFI engine." If the engine is an updated version of a carbed motor, like the 5.9 Magnum, you might want to weigh EFI against a carburetor. If it's something that was designed for EFI and distributorless ignition from the get-go, like the Gen III Hemi, I would only consider carbs under very specific circumstances, like class racing that allows newer engines but doesn't let you bring their EFI. You're looking at a big wiring, tuning, and fabrication project any way you slice it, so you might as well get all the benefits. The only question is whether you want to use a factory ECU or aftermarket, and that comes down to how modified the engine is and how much work it takes to get a factory ECU to be happy with being transplanted into an earlier chassis.
• "I want to put a turbo on my motor." EFI also does quite well with turbos, and these builds also typically bring the right expectation that you're in for a big project. There are some people who can get a carbed turbo working, usually on an all out race car, but there's a lot more people who can tune EFI for a street turbo application than who can pull that off with a carb.
• "I want to get more mileage and drivability." This is one where you really need to weigh your objectives. You're going to put a lot of money into installing and tuning the system. If you want to maximize mileage, don't skip the dyno tuning - remember the note about spark tables being the real key to mileage. Even the $1000 price tag, before tuning, on a cheap TBI type system could buy a lot of gas. And if this isn't a particularly wild engine but still has drivability problems, you may just need to fix a problem with your carb. Or a mechanical problem. If you want to take this step, keep in mind you're setting yourself up for a pretty big project and the fuel savings aren't going to pay for the cost of the project.
• "It's a road race car - even a few tiny improvements in throttle response and fuel consumption could mean the difference between first and second." OK, the mileage and drivability is a stronger argument here, and you are probably going into this with the right expectations about what it will take to get your tune to that level. Plus, onboard data logging is cool.
• "I want the wow factor" or "I'm building a show car". Only you can decide what you want to spend on this. EFI does have wow factor.
• "I bought these plans for a hydrogen orgone generator, and it's doing absolutely nothing, but I'm convinced if I could just do this and this I'd be getting double my mpg!" If a picture is worth a thousand words,
  

 

[MadScientistMat]

Next one - how do TBI systems fit into this? When it comes to drivability, you end up about halfway between a carb and a multiport EFI setup. If you're after drivability, maybe half the gain for half the work of an MPFI system appeals to you, although you may find it's closer to 3/4 of the work than half. Sometimes they fit into the show car / wow factor category as well, particularly if you're doing something with a rare and unmodified intake, or are using multiple TBIs. There are some setups out there to work like a Six Pack. But this design has inheirent fuel distribution problems, and very few of these systems give you the tools to adjust the injectors independently. Even if they do, each injector adjustment is going to affect several cylinders.

TBIs on independent runner manifolds (e.g. replacing three DCOEs on an inline six with three Borla DCOE-pattern throttle bodies) are going to behave more like multiport injection, as you have a dedicated injector for each cylinder.

While you can run a TBI with a turbo, it's not my first choice (although as I noted, the number of people who can tune this is probably a bit larger than those who can correctly set up a blow through carburetor for good street manners). You're likely to have to run it rich to make sure you don't have one hole lean out under boost. It leaves a bit on the table compared to multiport. There's only been two factory TBI / turbo cars, ever. One was the Mitsubishi Starion, and the other was an earlier Mitsubishi. And I had a lot of requests from people with those cars looking to ditch that TBI.

 

[matthon]

EFI is also fantastic on a tunnel ram.

I went from 390s, cool, very little hesitation, 600s, more better, slim hesitation, efi, different animal entirely.

 

[MadScientistMat]

Yes, it also works great in cases where vacuum signal is what's making it tough to get a carb in tune.

And next I would like to elaborate on one thing TT5.9 mentioned, grounding. Thanks for bringing that up.

It's easy to think ground is ground. But not when you're running 10 or 20 amps through it. That's enough to make little faults that don't show up on a multimeter really show up.

So I recommend running a dedicated ground wire, sized bug enough, straight from the ECU to the battery negative terminal. While some other locations might work with some testing, grounding straight to the battery always means a good ground. Nobody ever had grounding issues with a clean battery terminal.

Why not the block? It has that big ground strap to the battery in most cases. The trouble is that strap has to carry a hundred or so amps from the starter motor. And sometimes you run into issues with things like paint or cases where something that looks like a block ground is really just run through a rusty bolt. So it's better to ground the EFI straight to the battery. (Bonus: Please don't use the starter terminal as a 12v power source for your EFI wiring either. I've seen this mistake in commercial engine swap wiring harnesses. The ECU sometimes registered only 6 volts while cranking....)

Sheet metal grounds are typically the biggest problem - you have the current run through spot welds, rust, etc, and these can give a lot more resistance than you would expect.

Then, the sensors need to ground to the ECU sensor ground wire if at all possible (Sometimes you may want to wire something like a fuel gauge to the ECU and don't have this option, but any mission critical sensor should have a separate ground wire). If you ground the sensors elsewhere, like to the block, your sensor may be grounded at a different voltage and give bad data.

Sure, you'll see OEMs breaking these rules with their wiring, but they usually put a bunch of testing into it. These rules, though, will let you get great grounding without needing a bunch of testing and calculations.

 

[toolmanmike]

You talk about TBI, I had an old Holley analog system on a 454 Chevy powered tool truck. (Manual control box with dials) The system was old when I bought the truck and after 6 months I started having issues, mostly component failures. I needed to be daily dependable so I converted back to a Quadrajet. Other than setting the choke on the carb, I didn't see any difference in operation, mileage, or drivability.

 

[matthon]

I mounted a ground hub next to the battery, large terminal to battery, the rest to ground anything else, alternator, headlights, whatever.

Keeps a gazillion wires off the battery and everything clean and organized.

Put a ground hub and a power hub under the dash.

Even put a small ground block in the trunk, grounded tank, sending unit, pump, then ran a 10 gauge up to the hub by the battery.

Swapped out the glass fuse box for a Painless 7 circuit box, integrated it into the harness, Denso alternator, etc.

Also shielded the coil and distributor power wires.

 

[MadScientistMat]

You talk about TBI, I had an old Holley analog system on a 454 Chevy powered tool truck. (Manual control box with dials) The system was old when I bought the truck and after 6 months I started having issues, mostly component failures. I needed to be daily dependable so I converted back to a Quadrajet. Other than setting the choke on the carb, I didn't see any difference in operation, mileage, or drivability.

I've had a look inside one of those control units - there's no microprocessor, the only load input is a TPS, and there's a total of five or six adjustments you can tune. I'm kind of impressed it was able to perform anywhere near as well as a good Quadrajet.

Fun fact: These Holley Pro-Jection TBIs were based on the hardware used on the late '80s TBI 318, but with a control unit farmed out to MSD.

 

[Tsully]

Fitech TBI is based off of a 1994 Chevy truck TBI… any replacement or upgrade parts can be ordered by looking up that year truck and TBI and this is from their own mouth from one of their troubleshooting videos and Tech forum.

Also these systems, as many have stated, are normally user install error and they do require serious grounds as any Interference will wreak havoc. I used Faraday Tape on the wires and covered with normal electrical to hide the hideous silver fabric.

Also on these systems and I don’t know it is like this…but all of the TBI’s ones that I have looked and installed…only have one O2 sensor!!! Really!!! They know most people are running dual exhausts and the O2 is only reading one bank!

I looked into the dual hookup’s for O2’s and that’s an additional $4-500 bucks for a good system… not some eBay special!

They are not plug and play as they would like for you to believe…if your throwing it on a stock motor… then yeah… you’ve got a good shot at getting it right out of the box!

Any EFI system you buy… make sure you can tune it with software and not just the handheld that they lead you to believe!