Aftermarket TBI vs Multiport

Hello all,

I've been having fun considering fueling options for my currently carburetted 302 powered car. For now I'm not ready to spend the cash, but its fun to think about. The current consideration:

Aftermarket Throttle Body Injection vs. Port Injection

I have always assumed multiport is the obvious choice (more modern, used by oems, used by most fancy builds). However while looking at Sniper/Fitech type systems, it occurred to me that the operating costs may actually favor the TBI kits. My thoughts are centered around the cost to buy all the components of a multiport setup (ecu, harness, manifold, injectors, iac, sensors, etc.)  I'll easily spend the same 1k-1.2k amassing the parts to use either system. After that, the simplicity of a kit made to work together with "self tuning" is attractive. 

So what am I missing? I have heard some of the horror stories of troubleshooting with the cheaper throttlebody systems, but the reviews for the high dollar systems seem stellar. Is there a case to be made for multiport injection on a car that is not emissions controlled, used daily, maximum power isn't the goal, and simplicity is a goal? 

TBI has most of the benefits of MPFI, but some of the drawbacks of a carb.

One of the thnigs that MPFI opens up is the fact that the intake manifold is only carrying air instead of air/fuel.  Air is a lot less dense, so the manifold design doesn't have to take that additional inertia into consideration.  It's splitting hairs, but it makes a small difference.

MPFI also tends to be more accurate.  With TBI you're always going to have some fuel come out of suspension and puddle/recondense as liquid.

For a fun street car where you want the benefits of EFI without the cost, TBI is fantastic.  Bolt on, tune, go.  For the home enthusiast, it's hard to beat the simplicity, and MPFI doesn't add enough benefit to outweigh the cost/effort IMO.

The other option would be to grab a complete MPFI/ECM setup off a later 5.0L.  That saves you the trouble of mix-n-match parts, buying expensive aftermarket stuff, or modifying things.

Having tuned some aftermarket tbi cars, hard pass. Only if there is a good reason to not use multiport would I even consider it. Generally atomization isnt as good as with a carb and cam/manifold choice and getting it to run well will be limited. 
The only time I have done one and felt good was a bbc twin turbo we used three 454 tbi units on top of. Low rpm "bog truck" and needed to be mostly free. 

TBI sucks, it has all of the fuel distribution and wall wetting problems of a carb, and it has the poor atomization of EFI.

It is cheaper to tune than a carb but it will never run as well as a carb.  I would go with actual port injection, or a decent carb from a good company like Quick Fuel if you must have a Holley pattern, or a Q-jet.

I have installed Sniper drop-on systems and did one MS3Pro install on a TBI engine.  (Long story)  The MS3Pro setup ran way, way better than the Sniper POS did, although it was kinda tricky.  Getting cold start and idle drivability was "fun" because TBI routes its idle air through an IAC valve, but the injectors spray at the closed throttle plates, so it was a balancing act to make it work well when cold and not idle at 1200rpm hot with the IAC fully closed.  But I got it running really well.

Every, and I mean EVERY, Quick Fuel carb I have installed was plug and play.  Bolt it on, set the idle speed, DONE.  No need to rejet anything and usually didn't need to even mess with the idle mixture screws.  Perfect off idle drivability with no flat spots that require drilling out the idle feed restrictors to correct (and if one did, those are JETS and not fixed holes)  Part of that is very good calibration from the factory, most of it is extremely good atomization.

I kind of imagine rolling your own MPFI setup using a late 5.0 Explorer donor for just about everything and control via Megasquirt would be cheaper and about as easy as anything else you might come up with. 

I think I'd go carb over TBI as others have already mentioned. 

Is there still an advantage if using something like an MS2 with batch fired injectors? I appreciate the discussion thus far, interested in any other experience the hive has.

So, i hate megasquirt with a passion. I fought it for YEARS on my duster. 

A 750 holley out of the box 4k mioes ago... .

Ive installed fitech, msd atomic, fast ezefi, edelbrock pro flo, and holley stealth. 

Id buy the stealth today. It was the best out of the box tbi setup ive ever run. It worked as it was supposed to the first time, and every time since on a vortech supercharger equipped 360 dodge with way too much cam, too little converter and 4.56 gears. Its almost daily driven by the owner now.....

If i could do megasquirt/carb berkeleyery over, knowing what i know now, id buy the stealth. 

Batch fire only works if you have extremely long runners, so reversion pulses can't starve some cylinders and overfuel others.

Note that when Ford and Chevy batch fired V8s, they had those manifolds with 22" long runners.  Buick turbo V6s had very short runners and they also had sequential fuel injection, in the 80s.  GM went to sequential for the LT1 which had - surprise - very short runners.

I tried running a BGN six batch fire.  It ran on four cylinders.  Set it up as sequential like I should have and it ran smoothly.

In reply to Dusterbd13-michael :

Why do you hate megasquirt so much? Coming from 2 cars and probably 125k miles between them running MS2, one a small block chevy with multiport fuel injection and the other a BMW M42 with a turbo. Once I got the tuning stuff sorted out, I never really had issues with them, so I'm genuinely curious.

I have seen a slight improvement in idle smoothness on seq but all the batch installs I have done don't have the issues that Pete mentions. I think you would need really short runners, a lot of cam, and low rpm to get into that zone. 
 

I have 1000+ customers on various ms setups and nearly all issues were self inflicted by the customer. I could post a list but it's long. 

The runners in a BGN manifold are maybe three inches from the injector site upstream to the plenum, so yes very short runners. 

Also analogous to sticking injectors in a manifold designed for a carburetor, so I feel that it is a valid concern.

The V6 engine in batch fire would inject once or twice every six intake strokes.  Having smaller injectors would allow more injection events without running into pulsewidth issues, but that limits you for power.  Can't run 60lb injectors in a 245ci engine with three injections per revolution 

Think off all the man hours and expertize it took to get your DD's EFI to run as well as it does. Now, think of you having to figure it out and  get the same result on your own. Aftermarket EFI merchants make much ado about "self tuning" but in reality it is a meaningless term.

What I found out about aftermarket EFI is that you have to accept it as a hobby of its own. Also, the actual fuel squirting gizmo be it TBI or port flavored, is only one piece of the system.  Tuning the thing is an endless game; be prepared to spend many hours in front of your computer looking at data files and making tuning decisions that you may or may not be qualified to make. Enjoy your new forum friends until you learn to hate all of them!

In the end, after finally getting TBI to work reasonably well, I decided that a carb would be a better choice for a car that gets driven. Aftermarket EFI has a lot ways to fail and leave you on the side of the road whereas most carb ailments can be fixed with a hammer and screwdriver.

So I now have a Holley 670 cfm carb that is  reasonably tuned and out of sight and out of mind so that I can finally move on to the other endless list of stuff that you need to do on a project car.

In reply to NOHOME :

That is actually one of the problems with Megasquirt: the community support assumes that tuning and modifying EFI is the goal, and not simply a means to an end.

They go ballistic if you don't want to run the latest firmware or play around with EAE enrichment or whatever.  A E36 M3storm unfolded when it was discovered that I adjust my idle speed with a screwdriver instead of berking around with tuning.  Hey, man, I just want the engine to run well enough to focus on driving.  My focus is not on engineering engine management, my focus is on roosting on dirt.

This ties in with my overarching rant over people being offended that your priorities are different than theirs.  I run a distributor, fuel-only.  "But crank trigger COP perfection blargh blippity bloo"

With a distributor I can bump start the engine with the clutch if it stalls in traffic without the engine having to rotate a full revolution to find itself before it send spark.  "u need to fix ure tune OMG y duz it stall"

Idle speed needs to be low enough that it stalls sometimes.  If it doesn't stall sometimes, the idle speed can be lower.  The goal is to idle as absolutely low as possible, slightly lower, stalling every now and then is an acceptable possibility, and it is nice to be able to bump start the engine before the car stops rolling, so I use a distributor.  "OMGZ stop liking what I don't like"

OEMs spend thousands of hours tuning because you have to be confident in 500,000 powertrains in every conceivable condition every customer might see. If you are in the desert, you don't need to spend the hours doing one -20F start every 8 hours. Just like if you never go to the mountains you don't spend the time doing altitude work. 
 

But getting into a rhythm of taking and looking at data is important. Find the error, figure out what caused it, try the change, and see if it worked. 
 

Like Pete said KISS. There is no need for the more complicated until the simple does not meet your problem. And know that "problem" is relative. An A/F excursion was a problem for me when working on SULEV30, but if it's not causing a hesitation- it's probably not an issue for the nominal tuner. 
 

It is interesting to see tuners get caught up in the "newest" thing. Especially when they are putting in an OEM like code to "fix" things that don't need fixed to that degree. 

In reply to NOHOME :

I can only hope you are wrong.   To use Turbo chargers EFI is required.  Carbs just don't work well. 
     The stock Jaguar EFI  no ability to tune so I need to go aftermarket.   I've watch countless  people be successful with  Megasquirt , Maxx, and others.  
     The one difference the successful ones have is they seem  to understand lean from rich and have a process to systematically achieve their goals.  
 In short they understand cars.  

In reply to alfadriver :

Well put.    When I've gone to the Megasquirt site regarding the  Jaguar v12 it's nearly impossible to get basic answers but fueling at 12,000+ RPM  is deeply discussed and debated. 
      Thanks to people on this site I've persisted until I feel confident it's not magic.  Just simple rules to be followed to achieve goals. 

 In Short, thank you for your help and guidance. 

Pete. (l33t FS) said:

This ties in with my overarching rant over people being offended that your priorities are different than theirs.  I run a distributor, fuel-only.  "But crank trigger COP perfection blargh blippity bloo"

So your megasquirt doesn't have control over timing advance?  That may be one reason it stalls -- IME timing-based idle control is much more stable than trying to do it all with the idle air control valve. :)

Pete. (l33t FS) said:

In reply to NOHOME :

That is actually one of the problems with Megasquirt: the community support assumes that tuning and modifying EFI is the goal, and not simply a means to an end.

They go ballistic if you don't want to run the latest firmware or play around with EAE enrichment or whatever.  A E36 M3storm unfolded when it was discovered that I adjust my idle speed with a screwdriver instead of berking around with tuning.  Hey, man, I just want the engine to run well enough to focus on driving.  My focus is not on engineering engine management, my focus is on roosting on dirt.

 

This ties in with my overarching rant over people being offended that your priorities are different than theirs.  I run a distributor, fuel-only.  "But crank trigger COP perfection blargh blippity bloo"

With a distributor I can bump start the engine with the clutch if it stalls in traffic without the engine having to rotate a full revolution to find itself before it send spark.  "u need to fix ure tune OMG y duz it stall"

Idle speed needs to be low enough that it stalls sometimes.  If it doesn't stall sometimes, the idle speed can be lower.  The goal is to idle as absolutely low as possible, slightly lower, stalling every now and then is an acceptable possibility, and it is nice to be able to bump start the engine before the car stops rolling, so I use a distributor.  "OMGZ stop liking what I don't like"

That really make good sense Pete.  I loved the simplicity of a distributor.  ( and I've got a distributor machine so I can test  them,   But
    I'm ditching my beloved distributor  because the parts for that distributor are so brutally expensive.  Just the rotor is $249.  Then the  sensors and cap etc.?   I'm afraid to total it all up.  Plus,  I'm constantly hearing the aftermarket parts have a short life.    My original thoughts when my goal was the $2000 challenge was to sort through all my spares  and toss the bad, keep the good.  
     However when my goal changed to vintage racing. That didn't seem viable.  Entry fee,  fuel for the race car, fuel to get to the track, and home again, food, and a place to stay that's a $15-2000 commitment. 
      Ignition problems ruins that weekend.  So coil on plug with a few spare coils is a lot less risky.  

Paul_VR6 (Forum Supporter) said:

I have seen a slight improvement in idle smoothness on seq but all the batch installs I have done don't have the issues that Pete mentions. I think you would need really short runners, a lot of cam, and low rpm to get into that zone. 
 

I have 1000+ customers on various ms setups and nearly all issues were self inflicted by the customer. I could post a list but it's long. 

That's interesting. Jaguar's long runners I assumed was for low speed torque. 
  I know a set of  IDF Webers on short manifolds really are torque less  and don't really make that much power over  the EFI  no matter how much time you spend on the dyno burning through Jets and air correction.   Etc  
      Group 44 was even able to get 450 horsepower from the factory carb set up  that has 4 Strombergs set down below the intake on the heads.   ( Rumor has it the Huffaker gained 30 hp going to SU's.)

 I digress.    My real question is, runner length isn't critical to power?   A big  Helmholtz plenum with short runners  would be as good?   

I discarded a sequential set up for a batch fired one on s V6 engined MG and never saw any negative effects (I did it because I wanted to be able to tune by sending chips back and forth rather than the entire ECM, given that I am in Canada and most tuners are in the US).  No negatives that I could see, although I wasn't in  position to compare fuel economy.

codrus (Forum Supporter) said:

Pete. (l33t FS) said:

This ties in with my overarching rant over people being offended that your priorities are different than theirs.  I run a distributor, fuel-only.  "But crank trigger COP perfection blargh blippity bloo"

So your megasquirt doesn't have control over timing advance?  That may be one reason it stalls -- IME timing-based idle control is much more stable than trying to do it all with the idle air control valve. :)

If it doesn't stall occasionally, the idle speed is too high

An automated idle control algorithm would require taking into account things like battery voltage, coolant temperature, time of day, local housing density, and the distance to the nearest local authorities.  And I do not think even at the best of times that the setup can handle an idle speed shaped like a saw blade.

Y'know what works?  A screwdriver, and a little right foot assist when cold

In reply to Pete. (l33t FS) :

I'm sorry Pete.  
  Too slow a idle speed may mean insufficient oiling or heat extraction  due to too slow coolant flow.  It may also fail to charge the battery. 
          

In reply to Shavarsh :

Imho, a huge question is "what is the goal?"  
Im not a carb person, but get the impression that there are some really good ones out there. 
But if you want some basic digital tuning, a TBI system could be fun. 

If making it run really well, I too would suggest getting a whole intake system for port injection. And then look into the boards for tuning files from an engine that is really close to yours. Those exist and shorten the time to get it going a massive amount. 

I ran Midlana on an AEM Infinity ECU. Took it to a pro for the "flat out" portion of the tuning, so I wouldn't blow it up, and I handled the rest of it. The "rest of it" meant cold start, hot start, steady state cruise, idle, slow acceleration, and deceleration... you know, what's used 99% of the time. Getting  those to "OEM" level of drivability took me hundreds of hours, but was successful. Point being, it takes a LONG time to get it really right, or you pays your monies and have someone else do it. Doing anything less will be death by 1000 cuts in the form of stalling, bucking, surging, loading up the plugs, etc, etc.

kb58 said:

Doing anything less will be death by 1000 cuts in the form of stalling, bucking, surging, loading up the plugs, etc, etc.

It really depends on what quality of base map you can get.  If you're putting it in something where that ECU is widely used (like, say, a Miata), the base map will be 95% of the way there on the "rest of it" stuff already.  Building a complete map from scratch is definitely a lot more intensive.