A good friend of mine has been building a '77 crewcab for maybe eight years now. He's as slow as I am when it comes to completing projects. Anyway, 2007 was "engine year", and based on a ton of aggressive ebaying, trading, begging, gifts and so on, he ended up with this combination:

I'm merely quoting what he gave me originally for EFI tuning:

Bottom:
- Flatlander 385 block taken to 0.080 over stock (4.440" bore)
- 4340 Forged H-beam Rods (Eagles I think)
- Hypertueric Dished Pistons
- Scat 4340 4.5" stroke crankshaft
Top:
- Blue Thunder SCJ heads, ported, machiend to 90cc chambers
- Valves: 2.250" intake, 1.880 exhaust
- Comp Cam 275DEH (275/285 + 110 deg) 2-6k RPM

Arp everywhere, felpro gaskets, clevite bearings, yada yada.

Induction:
- Blue Thunder CJ dual plane intake, port-matched to above, converted to EFI
- Large carb I didn't recognize (BG?), under-jetted, fed alcohol/water via seperate tank/electric pump.
- (8) Bosch injectors 160lb/hr
- Pair of Turbonetics T76s in unknown trim (big!)

Exhaust:
- fabricated stainless turbo headers
- 4" stainless dual exhaust from turbos to tail, cats and race mufflers.

EFI:
- GM 1227749 Speed Density EFI (3bar)

The end result is a little over 550cid with the cam being most happy in the 2000-6000rpm range, a lumpy idle, and a hair over 7:1 static c/r.

The reason why he chose GM EFI and the 1227749 specifically is because for my twin-turbo crewcab project I was originally going to use these parts, so since I changed my mind a year or two ago he got himself the ECM, a homemade wiring harness that's long enough to wrap nicely around a BBF, and a box of GM sensors including the 3-bar MAP which obviously is key.

Over the past few months we've been going back and forth via email sending BIN after BIN trying to get it started, then "driveable enough" where he could head down to me for a day of "serious tuning" to see what we could do with it, EFI wise. I had to promise to drive him back if it puked engine parts

I recorded the last run of the night before he headed home with my mp3 player/recorder while strapped in the passenger seat, laptop on my lap.

http://frederic.woodbridgedata.com/audio/tt558.mp3

We got onto the garden state parkway (a long, wide, flat major highway that goes the length of NJ), pulled over, shut things off, and I dumped in what we hoped would be the final bin image. I pushed "record" just before he started the truck. He shifted into drive and pulled into the highway after some unstable, lumpy idling.

Idle is still unstable unfortunately and we'll have to play with that later on.

He gave it a lot of throttle to start off and the truck bolted incredibly hard and the boost climbed steadily. The rear was fully locked and the tires chirped a bit as traction came and went pulling onto the Parkway but the engine was so loud the recorder didn't pick up wheel spin at all.

It shifted from 1st to 2nd (which you can hear in the recording) then as the rpms got close to the 2nd --> 3rd shift point he mashed the throttle to the floor and held it, keeping it in 2nd allowing the boost to go "wild".

That's about when the turbos became significantly louder than the engine

Unfortunately, these test runs kept knocking my laptop's cable off the ECM with every "romp" (its very fiddly) even with piles of electrical tape so I didn't get to record the actual boost level he hit nor did I capture the top RPM. He wasn't watching the tach either as he was paying close attention to the road in front of him as we were passing anything and everything and one bad move - you know how that could have ended up.

But the boost gauge he installed goes to 30psi and it was pegged long before the turbo screeching reached it's peak, but based on the pitch of the whine and some simple math I roughly estimate he hit about 40-42psi or so.

He backed off fairly quickly because we didn't have any idea of what the boost was and he wanted to drive home without breaking parts - he lives about 3 hrs away so I can't say I blame him.

The funny thing is my wide-band O2 meter indicated a "barely rich" condition while the turbo whine drowned out the engine roar, so there's some additional power to be had if the engine parts can endure it - to be played with another day.

So yes, it's possible to use a lumpy cam, big turbos, wild crazy portwork, 92 octane, 550+ cubes with speed density EFI and have a fun toy you can actually drive on the street.

Don't Expect: Gas Mileage, tire life.

Do Expect: Lots of speeding tickets

;-)

 

Now that's just insane and awesome!!!

I don't know anything about GM EFI systems but I do know that the Ford speed density systems are very touchy...the issues are almost always with idling and a "lumpy cam". That and the fact that the ECU's can't register/read anything above .996 VE, and that is a bit of an issue with a power adder.

So what are the differences in the parameters between the two (Ford and GM speed density)? I ran mine with speed density, for some time, and I could never get the ECU to work/regonize load tables above 100%, no matter what changes I made. The MAP sensor had the potential to output higher frequencies, but the ECU couldn't do anything with it.

 

It was fun to sit in as a passenger. With the first "good" BIN I thought I was going to inhale my laptop

Speed density and lumpy cam's aren't usually a good combination because the lumpier the cam, the bouncier the map sensor's reading of the vacuum, because well, the vacuum is all over the place. My buddy's truck is no exception. The "trick" was to make reasonable guesses as to what the vacuum at idle would be if instead of sampling random values we sampled say, 20 readings and average them together, then flatten the fuel table at the lower RPMs to provide that amount of fuel. The end result is the map sensor providing seemingly random values, and the computer using that to provide randomly selected fuel rate values, which by flattening the table provides about the same amount of fuel regardless of what the computer *thinks* is happening.

The parameters are very similar, and personally, the Ford EEC is quite a bit more sophisticated in this vintage.

But I prefer the GM ECM stuff because on the removable chip (removable once you solder in the ZIF at least), is this - the table data AND the machine assembly language are in the same chip. Since the GM's of this vintage were motorola 68HC processor based, using free tools one can over time reverse engineer which "port" on the processor reads what sensor, what memory locations represent injectors (one bit per inj = 8 bits ) and how that's masked for 4,6,8 cyl applications.

GM MAP sensors provide a 0-5V voltage for it's scale. So a naturally aspirated engine map sensor will give you 0-5V to represent vacuum, as there isn't a situation where a naturally aspirated engine will be 100% efficient anyway.

a 3 bar Gm map sensor will give you 0-1.something V for vacuum, and 1.something volts to 5V for boost. You then have to reprogram the fuel tables (and others) to make the relationship between map sensor and injector pulse width to make sense.

The drawback of 2 and 3 bar map sensors is you lose resolution of measuring vacuum - instead of using the entire 5V swing to represent vacuum, you're only using a little more than volt of it, out of 5V.

This aggrevates the idle problem even more. But one can compensate for that like I said above - mathematically figure out what fuel flow at idle should be based on displacement, rpm, and so on, and basically redo the injector pulse width table to reflect that, so no matter what "random" value the map sensor provides, the computer very stupidily will give your value because they're in multiple locations.

As I mentioned above, the GM ECM's of this vintage have both the assembly language code and the data on the same PROM chips. Since this was slated for my crewcab, which is a manual tranny crewcab, I removed all of the automatic transmission code from the PROM as it's obviously unnecessary. This left space for other code.

Also, *ALL* of the GM ECM's of this vintage (P4 designator) have the same processor, just that the inputs and outputs of that processor are wired differently, pinned out different, and used in different ways (i.e. different surrounding logic) but since the processors are the same, that means if one digs into the code deep enough on many different ECM's, you get to the point where you can use the 3-bar code from one ECM, the automatic tranny code of another, the wider fuel mapping table from yet another ECM, then merge them all together, compile it and generate a valid checksum and slap that into your EEPROM and plug into the ZIF socket

That's how you can get things that aren't supposed to work together, to work together

I imagine one can do the same thing with Ford EEC, but my experience with hacking into EEC's is minimal as compared to my GM ECM work. So I can't say for sure what can be done or not, but I do know that most of the editing software and tuners that I've seen for Ford EEC generally just plays with the tables - not the actual code the internal processor is using.

Once something is dissected down to the code level, then that particular "shiny box" can be used for just about any application you want.

I know a guy who uses a GM 1227730 ECM to run his two cylinder riding tractor. I know another who an EFI weedwacker. And another who uses a GM ECM as an alarm system for his house.

Strange applications but it should demonstrate what an ECM/EEC is - a tiny, simple computer.

Once you can get into it enough to gain full access to the code that runs the "shiny box", you can now do anything you want with it.

 

That makes sense. I remember a small company (Performance Specialties I believe) went back and forth with a few GEN 1 Lightning owners, to get the ECU to read/recognize higher frequency levels for boosted applications. But the problem always led back to the ECU not being able to do anything with it. The 0-5 volt ECU's are alot like DeviceNet systems, so the obstacle was getting the ECU to calculate load levels above 1.00 VE from a frequency input from the MAP sensor...I guess nobody was ever able to crack the code.

What you have done is awesome...keep up the good work!

 

Frequency, voltage, doesn't really matter.

The Ford map sensor provides a frequency output based on vacuum, and offers a linear correlation.

80hz for 0 InHG, 160hz for 32 InHg.

So a 2-bar map sensor would have to rescale that, and your tables would have to compensate.

80Hz for 14.7psi, 120hz for 0 InHg, 160hz for 32 InHg.

The principle is the same regardless of what the sensor itself outputs, voltage of frequency. Essentially the sensor's output that represents a vacuum gets "scrunched" leaving room for a scrunched representation of boost.

The computer itself just takes the values of various sensors, and extropolates a injector pulse width off a calculation using those sensor values, so one has to edit the various tables to provide what you need.

Let's pretend we have a one-sensor EFI system to make the concept easier to grasp for everyone.

At idle, you need "X" fuel. Idle on our imaginary engine produces 22 InHg. So a 1 bar map sensor will give 145 hz output.

The EFI computer sees this 145hz output and knows to inject fuel with a 5% pulse width (duty cycle), meaning that for each RPM the injector is open for 5% of the time, and closed for 95% of the time

With the same imaginary single sensor EFI and same imaginary engine, if we're using a 2-bar map sensor the output for the same conditions will be 155hz, and we've edited the map sensor table so that 155hz now gives a 5% duty cycle to the injectors.

Of course the EEC is much more complicated than my imaginary one sensor EEC.

The EEC IV for example has many functions:

load scaling
MAF transfer
WOT spark advance vs RPM
WOT spark advance vs ECT
WOT spark advance vs ACT
accelerator enrichment
WOT fuel miltiplier vs RPM
WOT fuel miltiplier vs TP
part throttle spark advance vs ACT
open loop fuel vs ACT
closed throttle open loop fuel multiplier
spark advance vs BAP
spark advance rate
dwell
altitude fuel adjustment
cranking fuel vs ECT
injector adjustment for low battery
dashpot clip and decrement rate
transmission TV pressure vs TP
torque convertor lockup vs TP
upshift speed vs TP
downshift speed vs TP
idle airflow

EEC Scalars:

injector size
injector slope
minimum injector pulse width
accelerator pump multiplier
open loop fuel multiplier
part throttle timing adder
dwell minimum
dwell maximum
ACT minimum for adaptive control
ACT maximum for adaptive control
minimum ECT for deceleration fuel shutoff
minimum RPM for deceleration fuel shutoff
minimum load (MAP) for closed loop
hi-load timeout to open loop
idle speed neutral
idle speed drive
CID
number HEGO sensors
WOT TPS value
EGR multiplier
EGR type
PIP filter
half fuel rev limit
speed limit
maximum spark retard
cooling fan ECT hi/lo/hysteresis
intake manifold volume
thermactor presence

EEC Tables:

accelerator enrichment (lb/min)
startup fuel (A:F ratio)
base fuel (A:F ratio)
injector timing (crank degrees)
injector firing order
base spark (deg BTDC)
limp mode spark (deg BTDC)
injector output port
borderline detonation spark
borderline compensation vs ECT
17 11/20/97
borderline compensation vs ACT
borderline compensation vs lambda
acceleration fuel time constant
exhaust pulse delay
HEGO amplitude
HEGO bias
engine torque
engine frictional torque


The GM ECM stuff from the P4 family have much less parameters to play with...

Anyway, from the above you can see EFI tuning can get very complicated very quickly, because what values you change in one place, may significantly effect what the values should be in other places, resulting in what may initially seem to be a never-ending "value editing spree".

You start off by adjusting the scalar values to match the hardware you're using, then adjust the tables to meet your fuel requirements under a variety of circumstances. A wideband o2 meter and a dyno can be really, really helpful for this tuning.

My friend's truck that I described above probably has quite a bit more power to give except without it being strapped to a dyno I really can't dial in things that precisely because using how hard my **** plants into the passenger seat is not a consistant, accurate measuring device. The wideband o2 meter I have does offer many clues, but that's only one measurement out of the huge list above.

For a street-use vehicle, "close enough" often is fine, whereas for competition you'd want things as precise and consistant as possible.

Like with any project, the annoying little details take up 90% of the time, whereas the big stuff usually takes about 10%.

 

I'm familiar with the scalers, functions and talbles when it comes to tuning. These are things I see when I open up my tuning software to make changes to my 94, and there are almost as many parameters for the E4OD as well. I decided to go the DIY tuning route so several years ago, so I could make the necessary changes and learn more about EFI systems.

The concept as to how the Ford MAP works with the ECU isn't the issue. I've been down the road and tried adjusting the scaler values so the load tables will reflect readings above .996 VE for fuel and spark, and even the software wont allow it. For my mass air database (the system I'm currently running), I can program values that reflect a boosted application and the mass air ECU responds quite well. But not for the speed density side. Several others I know have tried it with a different tuning software and they ran into the same problem.

This leads to believe that there is probably some code that could be cracked and modified to allow a Ford speed density ECU's to be tuned more accurately, on a boosted application.

I got it close using a WOT fuel multiplier and injector value changes, but I had to store two programs on my chip; a summer and a winter program so the fueling would be more accurate in these conditions.

That thing you're working with is definitely a monster. It sounds like you have it just about nailed down...nice work

 

Okay. Even though you have the depth for this maybe others can benefit from my post anyway. After all, FTE is more than you and I

Curious... which EEC are you using? Tuning Software? Interface or other method of injecting edited data into the ECC?

The combination I have is Paul Booth's EECEDITOR which I struggle with, a 96 EEC-V (the last one that supports the older V8's and a dizzy), and Moates.net J3 adapter and some related do-dads.

[/quote]This leads to believe that there is probably some code that could be cracked and modified to allow a Ford speed density ECU's to be tuned more accurately, on a boosted application.[/quote]

The processor in the EEC's is a derivative of the Intel 8051/52 so it's a matter of digging into that to reveal where the code is and what it looks like. Then on can remove limitations such as these, as I have with the GM ECM code. The source code must be available, usually through brute force methods.

My buddy's truck will probably require the same, especially since the idle is way off. While sounds pretty good in the audio clip know that the engine was at operating temperature. On cold starts, it runs like (insert fecal related expelative here).

It's a fire breathing dragon for sure, and I do like oddball projects like these. His "next steps" is to record a couple of hard dyno runs, as well as "gentle" runs, so we have graphs and paper to work with. This will be done with his laptop attached logging the ALDL output - i.e. sensor readings and so forth.

Then we'll have to revisit the tuning using this information, eliminating the "this feels/sounds right" perception.

And it's good practice for me, because once I get my 500cid twin turbo stroker going, I'll be going through this again, with 58cid less displacement.

 

If you don't mind me asking what kind of money dose he have in that motor after all of it was said and done. and that has to be one of the meanest motors i have ever heard! And once you get done with your project truck will you be able to run with him?

 

Honestly, I couldn't ballpark what he's spent on this mostly because I don't know every part that's in it, but I can say that none or very few of the parts were new. Most everything was purchased used, acquired in traide, bartered for, or he did "side jobs" to get some of these parts.

But if you want to see a high number... lol
The block lists @ $2800
heads list @ $2000 for the pair, "naked"
intake @ $500
The T70's list for $2000 each.
The injectors @ 70 each (eight needed) $560
The EFI system (ecm, custom harness, all sensors, et all) $200

Plus valves, rockers, lifters, pushrods, tinware, drysump oil system, howe 3-core radiator, fluidamper, arp bolts/nuts/studs, flywheel, starter, solid motor mounts, plumbing and duct work, large HKS blow-off, pair of external wastegates, silicone hoses, stainless headers/flangers, accessories, the alcohol carb he uses as a big-**** throttle body, a stainless "hat" for the carb, block and rocker girdles, etc.

Not to mention the countless hours machining, portmatching, porting, decking, and so on.

The major blockwork (line boring, cyl boring, crosshatching, etc) was done by a professional machine shop. Everything else he did, a little each night. All the porting and port-matching was done by hand with a pile of carbide burrs in various air grinders - strait and angled. He made the bungs and tig-welded them to the intake for EFI, he machined parts to graft the TPS sensor onto the carb's throttle control, he hand "adjusted" the chambers in the heads and so on.

Next time I talk to him I'll ask if he has a ballpark as to what he spent, but I know that most of the "cost" was countless hours cleaning parts up, machining, and careful assembly.

---

As far as being able to "run" with him, I would have to say "probably not".

Our projects are similar in some aspects - crewcabs, big blocks, twin turbos, EFI. But they're being built for different purposes. His machine is really a serious street/race machine/hotrod/muscle car that just happens to look like a crewcab pickup.

Early on he joked around that he was building a 4-door, 4wd "viper killer".

But he's got more 57 more cubes, better parts all around (flatlander, blue thunder, eagle, scat, etc), significantly larger turbos, and what is really a hybrid street/race cam with more of an emphasis on race, bigger valves, massive amounts of portwork and port matching, etc.

In contrast, I'm using mostly junkyard parts (not necessarily Ford parts either), a RV/Towing cam, etc.

About the only things our engines have in common is the bell housing pattern and the valve lift

Even if I port the heck out of my cast iron "industrial" heads, they probably won't flow what his blue thunders flowed BEFORE he ported them

I'll have a lower compression but also more room for boost that way so there's my one advantage. That and I can tinker with the EFI endlessly.

He built "seriously" for fun/power/acceleration. My project is about towing power, and usefulness on the street, and also gas mileage. My first twin-turbo pickup which was built using many mismatched, mis-sized parts managed to hover around 17 mpg or so on the highway if I was gentle.

My goal this time is 20 mpg. No idea if I can hit it, but I'm sure as heck going to try.

 

I use SCT Advantage software and a 4 bank chip and the EEC calibration code is from the C3P family (5.8L, speed density w/E4OD).

Raymond Brantley was working on getting past the speed density's limitations but never got to the bottom of it. He, and a few others, ran into the same dilemma that I ran into no too long ago with the SD application; the ECU not registering anything above .996 VE or 99.6% load. I think they were using EECTuner and/or Tweecer. As you probably know, the SD ECU's have a set load table (volumetric efficiency) that affects fuel and spark; MAP vs RPM which gives a calculated engine load. The cells and scales can be changed, but it not for anything above 99.6% load.

 

That sounds insane!

Just one question, how on earth did he manage to floor it in second gear and stay on the road? With 550 cubes and 40 PSI... in a truck no less, I can't imagine having any sort of traction till 80+ mph.

 

There was tire squealing for sure but the recorder didn't pick it up because the engine/turbos were so much louder. What you hear is what we heard sitting in the cabin of the truck with the windows up. With the windows down, the recoder just clips and records massive distortion.

With 33x11.50's all around and 4wd engaged, it's got a fair amount of rubber to apply power to the asphalt, which the weight of the truck certainly helps - this is a 77 F350 crewcab, not a lightweight import obviously.

Remember that engine RPMs won't instantly jump from 2200 to 6000 within milliseconds of pushing the throttle - it does take time even with turbos.

Certainly it accelerated very quickly (by far quicker than stock of course) but remember all the stuff that has to accelerate too - heavy tires, heavy rims, drive axles on all four corners, front and rear diff gears, drive shafts, internal transmission parts, torque converter, flywheel, and of course the very large, very heavy rotating assembly inside the engine - crank, rods, pistons, wrist pins.

It's not instant, even though the boost skyrocketed much quicker. But yes, there was a lot of tire squeal and there was lines of rubber behind the truck.

 

Ahh, that makes a lot more sense then. Sounds sweet.

I need to get off my *** and get going with my turbo big block 4x4.

 

Me too. I have 7/8th of the parts I need to finish. Just not enough cash to have the new block machined, and it doesn't fit on my milling machine unfortunately.

A friend of mine has a bridgeport, I'm seriously considering something quite inappropriate and low-tech.

 

To whomever asked about the $$$ that's buried in this engine, I spoke to my pal last night and remembered to ask.

He said he's got about $4500-5000 CASH into it, and the difference was made up through bartering, ebaying, craigslisting, a little junkyarding, and so on.

The tranny is a very well built C6 that was put together for a bracket racing project that someone he knows had to bail on due to having another child, so he got the tranny for a "sick" amount of money - $500 with the torque converter. The TC was sold as the bite RPM was 4000 and installed something much more streetable. He couldn't remember what it was RPM wise, but it's a TCI unit.

Right now the truck is on his two-post lift so he can replace the ring gears in both axles... The front is cracked but held together, whereas the rear one became shrapnel.

I imagine he's going to have this problem often