A12 - time to get busy

That looks like a damn nice resto on that car, I don't see much in the engine compartment i don't like, and from what I can see of the car it looks like a quality build!

From watching the vid and reading the article I don't think Holley offers a complete Six pack set-up....just the EFI system and the individual throttle bodies, which by the looks of it that was a custom build by the builder is my guess. That plate the throttle bodies mount to is a single piece and there looks like additional provisions for brackets and such as well.
 
I haven't watched the video

I found this but it doesn't mention self-learning so I'm assuming it's not. It's self tuning too.
FiTech Fuel Injection 39610 FiTech Go EFI 3x2 Tri Power 600 HP Fuel Injection Systems | Summit Racing

I looked at Holley's site and it didn't say a word about 2-barrel throttle bodies. I do see them now at Mancini but they seem intended for use in a single throttle body application.

I'd hate to think you have to sink $2200 into throttle bodies and brackets and then another $2k in the ECU. But I wouldn't be surprised.

Since I'm starting from scratch on this car, it kind of makes sense. I've got to buy fuel tank, lines, and a pump anyway. It still look stock, but there's no fight with the tuning and you can't get to the idle adjustment screws.

I've already got non-original carbs but there are holes drilled in the butterflies and the end bowls are from a 4-barrel with the accelerator pump location filled with epoxy. Obviously I don't have much $$ in those so starting over won't hurt as much.
 
Last edited:
b-body-bob said:
This article installs the FiTech system
Kick the Carbs! Installing FiTech’s Go EFI Tri-Power 600HP system on a '71 Challenger
Click to expand...

That looks like a really nice set-up...and fairly reasonable too!

What is the cost on restoring OEM Six pack carbs? Quick internet search...it looks like at least $1K for a good rebuildable set of carbs, with yours having mods it would probably be a bunch more money. And at the end of the day you still have to tune and adjust the carbs frequently.
 
so...since its been mentioned before, i have to ask the stupid...i was under the impression that NONE of hese systems were "install and learn as you go" but required a "baseline" dyno tune$$$$$ to get you into the correct range at which point the kits would then "learn and tweak" minor things based on what it can read

i mean if shits come round to the point where you dont have to actualy get on a dyno ..i will have to re-think the brains in the lloyd
 
That FiTech system is fully self learning and tuning from what it looked like in the vid. The Holley kits from what I saw and read look like they are might be the same...self learning and tuning as well?
 
69.5CUDA said:
so...since its been mentioned before, i have to ask the stupid...i was under the impression that NONE of hese systems were "install and learn as you go" but required a "baseline" dyno tune$$$$$ to get you into the correct range at which point the kits would then "learn and tweak" minor things based on what it can read

i mean if shits come round to the point where you dont have to actualy get on a dyno ..i will have to re-think the brains in the lloyd
Click to expand...
From what I read they come with a baseline tune to start with where you set a few parameters like cam profile from 1 to 4, and it takes off from there.
 
thrashingcows said:
That looks like a really nice set-up...and fairly reasonable too!

What is the cost on restoring OEM Six pack carbs? Quick internet search...it looks like at least $1K for a good rebuildable set of carbs, with yours having mods it would probably be a bunch more money. And at the end of the day you still have to tune and adjust the carbs frequently.
Click to expand...
Well, first, I wouldn't restore them anyway since they're not original to start with. But yes it could get expensive even so. If these 6-pack EFI systems had been available I would've never bought carbs.

3 new carbs will cost about the same as the whole FI tech kit. (990+660*2 at Mancini) The special tank, fuel pump, and lines will probably run the FI cost up some.

I think the only difference in tuning vs a 4 barrel is you've got two secondary circuits and all of the carbs have idle adjustment screws and 2/3 of them are hard to get two.
 
I'll once again remind that throttle-body EFI is just an electronic carburetor and lacks many of the benefits of multi-port EFI--it's a huge waste of money. You have the same distribution/puddling issues as a carburetor. Guys that talk about how much better their TBI injection drives simply never spent quality time improving the drivability of their old carbs. And most of them are starting with way-fat, poorly-configured aftermarket carbs (usually Holleys of which they have no grasp of tuning) or beat-to-death originals that haven't been rebuilt and tuned via the FSM in decades. Or, y'know, they work for a magazine whose existence depends on selling their advertisers' parts.

"It cold-starts so much better." Of course it does; they either started with an untuned choke or no choke whatsoever. Kev fixed his "cold-start" issues with a cheap inline electric fuel pump and verifying his OE AVS's choke was exactly where the FSM said it should be. One pump and it lights off instantly.

"Self-learning" is a tremendous joke that leaves plenty of power and fuel economy on the table. People believe that BS because new cars have adaptive elements built into their PCMs, so how hard can it be for the aftermarket to copy?

Here's the problem: Targets. Every factory PCM has literally thousands of hours of dyno time spent figuring out exactly everything that's perfect for that particular engine (series--they don't do this on a per-engine basis) to make maximum power and fuel economy within the given restraints of emissions. Since every actual installation is going to have production differences, the adaptives are necessary. They also adapt to a particular driving style using information gleaned from the engine sensors. The long and short version: The factory guys have already learned where that particular engine makes peak power, torque, and economy... then they have to make it comply with emissions and adapt to either Hot Rod Johnson, Grandma Grunt, or somewhere in between. The answer? Adaptives. Yet, as EFI Ed so profoundly proved, even a factory adaptive EFI PCM based purely on airlfow (he went to the trouble of building a MAF-based system--no mean feat) is completely lost when installed on a significantly-different engine. Even though it was reading the actual amount of air entering the engine and had sufficient built-in fuel/spark targets, the car ran poorly. After custom modifications to the PCM and a couple of dozen hours of dyno time, it finally ran great. Still, he seriously regrets the OE PCM MAF decision.

So factory PCMs are self-adapting, but not self-learning. Aftermarket PCMs function exactly the same way, but they have extremely-general target values based on the scant information you provide during setup.

Let's discuss your particular engine. Where is your peak volumetric efficiency? At what air-fuel ratio does it make the best power or torque? What about the best economy? Which AFR provides the best idle characteristics? I can tell you this: The answer is never 14.68:1. How much fuel enrichment does your exact engine want to see during full-throttle acceleration? Or part-throttle, at a given engine load, RPM, and throttle opening? What's the ideal cold-start enrichment AFR? How long should it last in engine cycles for your engine? At what rate should the cold-start enrichment cycle decay to the fully-warm program? It's not a straight line based on temperature. What's your optimum injector size based on VE, AFR, duty cycle and displacement? Do you really believe one size fits all?

You have no idea what the answer is to any of these questions. Nor do I. Yet they must be answered for a "self-learning" PCM to work correctly. If you don't know the target, for what do you aim? Rumor? Generally-accepted averages? "This worked for me" posted on some forum by a guy with a kinda-sorta similar engine? You can get "this worked for me" just as well, if not better, from a factory carburetor setup--especially if you're considering an electronic carburetor (throttle-body injection). The factory guys had to meet drivability, power, and emissions goals too.

Here's the best part: Most of the benefits guys seek with aftermarket EFI are much easier to get with aftermarket programmable ignition. Power and economy are more easily gained through an absolutely-correct spark map. Even with dyno time, which is essentially required (although much like EFI, guys have had decent success with a ton of seat time), it's often still less expensive than EFI--with greater benefits. The truth of the matter is that a tenth of a point or two of AFR won't make much difference in performance or economy but a degree or two of spark advance or retard can have a dramatic effect.

You don't have to believe me; read the works of Greg Banish (a former dyno-cell engineer for both Ford and GM), Jerry Hoffman and Matt Cramer, or Mark Warner's excellent book on turbocharging (all the same rules apply below boost). I have read their stuff, at length and repeatedly. After doing so, spending thousands on "good enough" seems like wasting money on something I could've done using carburetion and a distributor. All agree that "self learning" does not and cannot exist. Every engine is different. Banish's constant harping about injector sizing alone will make you say "OK, OK, I get it!" aloud halfway through the book... but it continues throughout.

All of the above authors tell you the same thing in their books after the EFI bits are done: It's all in the spark. What do they know? They only do this stuff for a living, and not for magazines.

Based on what I've learned from the above-mentioned writings and further research online, my stillborn turbo project is (was?) MPFI with Ford EDIS-8 ignition built into the PCM. My Challenger will have factory Six Pack induction, but I have a standalone programmable ignition controller for it. I want the Six Pack look, but I don't believe the benefits of an MPFI-based Six Pack setup warrant the enormous cost over what I ended up investing. I ruled out the FITech nonsense the moment it appeared. That decision has been confirmed by a long-suffering customer's experience with it; it cost him a shortblock. He now apparently knows more about tuning a FITech than anyone at Holley (they just read excerpts from the manual on the phone) and is completely on his own. After a couple of years of trying to get it sorted, he'll gladly tell you 1) the system is a joke 2) self-learning is an even-bigger joke, and 3) he should've gone with carburetors.

FITech and its ilk--"self-learning" TBI setups--are a half-assed solution to a questionable problem. Banish and Hoffman/Cramer should be required reading if you're considering one.
 
that there is what i figured the truth was.....as its impossible to "on size fits all" a set of cloths let alone engine management

this factual bit right here is why i keep going back to..ya know..carbs on the alfa would be a whole lot easier and cheeper.....even if it was a full set of mikunis off a street bike

i WISH there was a magical way to avoid my dyno time...

some day there may be enough computational power to actually have self learning..or..atleast in theory..but the computational power and amount of sensors to doit?!?!?!??...im certain youd need port by port readings of both intake and exhaust at the very least for fuel flow rate, afr, maf, o2, temp in/out

theres just toooooooo many variables

i have 3 standalones for alfas, and can tell you this much..someone else with the EXACT same setup with a TON of dyno time...will plug n play into my engine but even it is leaps and bounds off ..it runs but its fat, the spark timings off ....ive spent a ton of seat time butt dynoing and tuning on the fly and HAVE improved it..yet..its still wa way way off...how much 8mpg off on the fuel consumption side i can tell you that much
 
69.5CUDA said:
i have 3 standalones for alfas, and can tell you this much..someone else with the EXACT same setup with a TON of dyno time...will plug n play into my engine but even it is leaps and bounds off ..it runs but its fat, the spark timings off ....ive spent a ton of seat time butt dynoing and tuning on the fly and HAVE improved it..yet..its still wa way way off...how much 8mpg off on the fuel consumption side i can tell you that much
Click to expand...
No two engines are the same, and on a MAP-based system--which virtually every standalone is--every change is significant. Cam, intake, even a different muffler will throw it off. Adaptives can't make up for a different engine. That's why EFI Ed went mass airflow: "If I want to change heads or cam later, it'll be able to account for it." It wasn't. He gets pretty impressive economy for a near-500-inch engine, and his RWHP is definitely adequate, but that's after a half-dozen dyno days. It was quite a learning experience, and he'll go MAP if there's a next time.
69.5CUDA said:
some day there may be enough computational power to actually have self learning..or..atleast in theory..but the computational power and amount of sensors to doit?!?!?!??...im certain youd need port by port readings of both intake and exhaust at the very least for fuel flow rate, afr, maf, o2, temp in/out
Click to expand...
The average standalone has more computational power than the anything in the space program in 1969. Processing is not the problem. Simply put, without direct readings of peak torque and horsepower whilst making various changes in timing, AFR, or any other variable, no computer known to man can determine what the optimum settings are for any particular engine.

Every standalone comes with a "setup" program into which you enter your displacement, etc. but the resulting initial "tune" is merely a best guess. Consider this: a 383 Mopar and a Chevy 383 stroker have the roughly same cubes, but the Mopar has better-flowing heads while the Chevy has nearly a half-inch more stroke. All the PCM knows is that it's dealing with a 383. Even if you enter bore and stroke during setup, the Mopar has, at a minimum--and not likely one--2.08"/1.60" valves, which won't even fit in a SBC. Head flow has a huge effect on load v. RPM. In this case, one size doesn't even fit one size.

69.5CUDA said:
i WISH there was a magical way to avoid my dyno time...
Click to expand...
There isn't. You might get "I can live with that" but if you want the maximum benefit, you have to do the homework (dyno time).

Once again, I'll return to the subject of targets. One engine might make peak HP at 12.5:1 air/fuel, whereas another does it at 13.0:1. The first engine might make best economy at 16.2:1 air/fuel, and the second at 15.8:1. So, which target do you enter into your FITech for your engine? Do you enter a guess, and when you're not happy after it's "self-learned" to that AFR you then try another guess? The only thing a "self-learning" PCM can do is adapt itself to hit the target AFR that you entered. And without a dyno, you have no idea what that AFR should be at any given engine load/RPM. Engine load v. RPM is almost never a constant. Think about it: You're driving on a level interstate at 80MPH with the cruise. You encounter a sudden gust of headwind. Your engine load just changed in relation to your RPM. The cruise feeds it more throttle; now you have a light-throttle transient. How much additional fuel is required for that momentary transient? There's another transient when the gust subsides. Transients are where aftermarket EFI is toughest to tune... and where so-called "self-learning" PCMs simply overfuel to cover a potential stumble. Transients happen a lot. Fuel's $4/gallon.

There is no such thing as self-learning timing, either. Contrary to popular belief, your best timing is rarely right at the edge of detonation. A knock sensor will retard the timing out of detonation, but there's absolutely no way a PCM can learn a damned thing about timing.

User-configurable computer-controlled ignition absolutely has to be tuned on a dyno. Installing EFI on an engine with a distributor is like installing air conditioning on a Model T roadster... most of the potential improvement goes right out the window.
 
Last edited:
belive me..i agree with you 120%, i spent enough time mokeying about with multiple maps from 3 different yet identical engines, and even tested it against 2 different yet identical engines i own....not one of them was "close enough"...and were talking about 3 guys who clamed to have the same cams in other wise stock engines(S-spec cams are a popular and BIG gain swap), and not one of them was close enough, in fact only one of them ran well enough to actualy monkey with...and runs well enough to drive and seat dyno..but the reality of the variable is and epic undertaking and promptpy made me understand the how and why a dyno is REQUIRED

its everything from altitude to pipe size/length, tire size, drive train drag, parasitic losses even the amount of milage on the engine and how far "out of spec" your injectors are, all of it has to be "tuned out" on a dyno....even driving and tuning is neerly impossible as your not getting any "feedback" from the road under you

thats why i figured something had to be "off" on the whole self learning....i mean in theory one day we might have something but like i said the sheer amount of sensors required would be overwhelming
 
If you read Greg Banish's book, and I highly recommend you do, when you reach the tuning section you'll think "Hell's bells, I'm going to be at the dyno forever." That's not necessarily the case. He does give some helpful ideas on how to narrow things down quickly, but what's most evident there's no real way around it. But if you read and comprehend it, you'll literally roll your eyes at the mention of "self-learning" EFI. I've actually read it multiple times, to both make sure I fully grasp everything and make it second nature when I do eventually get to a dyno. I'll know why the operator is doing what he is, and if not I'll have solid reason to question him.

A very-quick rundown:

An EFI table is a 3D map made up of cells formed by a scale of load--either manifold absolute pressure (MAP) or airflow, depending on PCM--in one direction, RPM in another, and throttle position in the third (you can't really do anything about throttle position beyond defining its operating voltages, but the PCM takes it into account for tricky stuff like transients). Depending on the PCM, you'll either enter a fuel amount or the target AFR in each cell. Your target is maximum brake torque (MBT) at any and every given cell outside of the idle area. If you enter a target AFR for a cell, it's the one you've already found to make MBT. With maximum brake torque, you get maximum everything else, including fuel economy. How? Very simple: If you can cruise at a lower throttle opening, you're doing it with less airflow. That means less fuel is required. Done correctly, a "maximum power" map will also provide the maximum fuel economy you can hope to get.

There is no such thing as a torque sensor for a car, nor can there be (if you know how a dyno works, you understand why). As such, "self learning" EFI is impossible. It can't measure torque, it can only adapt to whatever target AFR you enter for any given situation (cruise, full throttle, idle, etc.). That target will always be a guess. You'll never get better than "It's better than the carb I never seriously tried to tune" without tuning on a load-bearing dyno. The bullshit accelerate-a-drum county-fair dyno can't do it (and are notoriously inaccurate).

If the FITech and other TBI-style EFI systems can be map-tuned, you can certainly optimize them with dyno time. I'm not familiar with their software, but I'd hope they give you the option of optimizing their system. Edelbrock's later stuff could be map-tuned but I don't believe the old RetroTek and its variants could. Others, like Autotrend, make the TBIs but outsource the PCM.

The fastest path to maximum torque? Programmable spark timing. Period.
 
playing with those "cells" is so overwhelming its mind boggling

hear me out for a second doc, the way the AWD system works in some soobs in some semblance of theory be used to work like a dyno....im not saying its perfect or even a replacement for a dyno..."yet" or that it ever will be, but...atleast in my eyes its a step into the FUTURE of the possability of a MBT based sensor on a car
 
I'm getting cranky already - it seems like the shop is working on cars that weren't there when I visited back in October, and that makes me wonder if I'm being skipped over to work on other things - for example there's photos of a Mustang getting hacked up for a Coyote transplant, and they're picking the body off a 55.
 
You must log in or register to reply here.

SiteLock

SiteLock
Back
Top