EFI Boxer

option 3

 

The stock lid is very light when all of the other needed parts you wont make are removed. I happen to have a BB lid as wall art and they are interchangeable especially if you use a 79 to 81 version that has the same latches at the BBi. I would change the lid to line the stacks up with the openings and hang the BBi one on the wall. That said, the cabin shut line is usually altered to fit the lid but it can be done the other way around to avoid painting the car.

 

Status update.

- Pretty much finished with the wiring on the 7-12 bank. I ended up combining the coil and injector wires for each cylinder in braided sheath, then sealing the ends with Kapton tape and SCL heat shrink to keep it from moving / fraying. This is quasi-permanent, but could be undone (with a little patience) if I want to reuse the harness for a different layout (eg, ITB). I have the injector wires going in under the coil connector and I'm using 90 deg boots to try and keep them from sticking out too much beyond the valve covers. Maybe there are better ways I could go about this, but I think this is good compromise – all things considered. I should be able to get the 1-6 bank finished in the next few days. I'll probably do something about the Denso stickers on top of the coils. They're visually a bit loud, even if the stickers are all that indicates they are legit, and not some cheap knock-off. To be fair, the pics are focused on that area, but when you're looking at the whole engine, it doesn't scream "EFI conversion" or "restomod", etc..

- Calibrated the ECU for the new injectors. I was mistaken before, these injectors were used on the Ferrari 430 and 599 (not the 550). That really had no bearing on their selection, just a happy coincidence. Anyway, the ECU is programmed to take into consider an individual injector's flow at different voltages and pressures. Typically, you send an injector to Motec (in AU), who then analyzes it, and send back the data – which can take weeks. My Motec dealer happened to have this data from another customer, who was kind enough to share the raw data in exchange for me providing the calibration file that works for a Motec ECU. It's basically data entry of hundreds of numbers. So now I can confidently use these injectors with this ECU.

- Swapped the 3" intake for a 6". This brings the filter in line with the edge of the plenum. Curious to see the intake temps when pulling air from this area.

- Ordered 90 deg 22mm fuel / oil hose to see if I can increase the height of the valve cover oil crossover. I only need a few millimeters to improve the clearances at cylinder 12 (it's a bit cramped between the hose, fuel rail, and coil). Should arrive next week. Not a big deal if it doesn't work out, but I figured I'd try and clean up that area a bit.

- Finally received the last batch of Denso coils. Sourcing legit Denso "pencil" coils is a bit of a challenge. Once injection is good to go, I'm unblocked from immediately jumping into ignition.

- Planning on tapping the plate I used to block off the old CIS cold-start injector with 1/8" NPT threads. The air intake temp sensor will go there, and will have a similar look (sticks out).

Image Unavailable, Please Login

 

Looking really sharp! I suspect those are about 24 or 30lb injectors? What you're doing and how you're doing it does nothing but good for this car IMO. You'll notice a bump in power, fuel economy, cold-starting, everything.

 

Thanks, I'm trying!

These are 35lb injectors. At 50% duty cycle, they should be good for 500 HP.

Agreed. This engine is criminally detuned. Being able to fine-tune A/F and ignition throughout the RPM (and vacuum) would allow for optimizing the existing internals. Very curious to see the dyno numbers. I think the real fun happens after upgrading the internals (lighter crank, lighter pistons, more aggressive cams).

 

only when the ECU is programmed right - and this will take a while. or is this MOTEC a self learning ECU?

 

I've yet to EFI a Ferrari engine but I have helped out a number of Porsche people. I saw a dead stock 2.7L 911 engine from a base 911 pick up a full 16bhp at the wheels from swapping to EFI. No other changes, just dumping Bosch injection and using the single standard throttle body. Huge gain for a small engine. That was an AEM Infinity ECU which has a similar feature that Link ECUs have in that it YOU input exactly what air fuel ratio you want and a given throttle % and with each data logging run it will tell you exactly what changes to the fuel table need to be made. I had it road tuned in maybe 2 hours, then dyno tuning at WOT took maybe another 2 hours. Once it's running it's not very difficult to tune IMO. Some ECUs now will make the changes automatically based on monitoring you while driving, pretty neat.

 

Not "self-learning". To be honest, I think that term is a little misleading (more of a marketing gimmick). Most ECU that "self-learn" only do this in a narrow context. IMO, definitely not a substitute for proper tuning.

The best way to get a proper tune is to get on a load bearing dyno and "steady state tune". This allows you to hold the engine at various RPM / load, and tune for that specific region before moving on to the next region (as opposed to a ramp run at wide open throttle).

 

The Motec will do this as well. It's called "closed loop" trimming. The computer will compare target A/F with actual (from the O2 sensor) and then add/subtract fuel to match the target A/F. It's not ideal, because you're essentially playing catch-up all the time, and being reactive as opposed to proactively tuning the car in a way that doesn't require trimming.

Not to be confused with the sub systems used to trim fuel / ignition based on cold-start, cold run, idle control, etc.. Eg, the ECU will handle idle by retarding / advancing the ignition to smooth out idle.

To Romano's point though, it all does need to be setup properly for this to work well. My Motec dealer (and I believe most Motec dealers) offer a service where they'll remote desktop into your laptop, and help you setup the car. I am going to try to get it as close as I can on my own, but I might have my Motec dealer jump in if I get stuck.

 

Really cool that this tech is where it's at now. Getting EFI to behave has never been more simple. Wide open throttle tuning is the only thing you might want to make certain is done on a dyno, but I've been extremely close just street tuning for a couple hours old school with a laptop. I think swapping carburetor jets/chokes/tubes/adjusting timing and data logging is far more laborious than EFI tuning.

 

Agreed on the first part, disagree on the second but to each their own.

 

Yeah, I've heard similar. Helps to have good logging, as you can look back and see what the engine was doing (eg, A/F trimming), and make tweaks. I'll likely street tune to start, and then get on a dyno for fine tuning.

 

that is right. but the triject I have in my competiton has a very wide context. sure, the basic you have to program.

 

Status update. She runs (idles)!

My original plan was to run EFI injection with the original mechanical ignition to limit variables (ease troubleshooting). However, getting the ECU to play nice with the original ignition ended up being more hassle than it was worth, so I just went for it. The wiring was already in place for the coils, so I swapped the dizzy cap for a 3D printed cap that has a Hall effect sensor in it. The original rotor arm is replaced with a 3D printed version with a magnet that the sensor reads every full rotation of the camshaft. This is the "sync" value. Combined with a manually determined reference (offset) value, you now know cylinder 1 TDC; allowing for sequential ignition / injection.

A little trial and error later determining the reference / offset, the car started and was idling. You use a timing light on the flywheel similarly to how you would normally. The difference is that instead of rotating the distributor, you're pressing up and down keys on the laptop. I spent yesterday checking for fuel leaks, temperature, tuning idle (the ECU uses ignition timing to maintain a target idle), and tweaking closed loop fuel control (let the ECU add / subtract fuel based on the O2 sensor value). Never really got above idle.

Side note / funny story, I removed the coil, but figured I'd leave the Dinoplex removal for another day (it's not in the way, and there's a long list of TODO items). I forgot that I had the original sensor wires still attached to the distributor sensor (along with the wires going to the ECU); which had the side effect of continuing to push the RPM signal to the dash tach. So I have a rare multi-thousand dollar mechanical ignition box working as a glorified tach. Just thought that was funny. I still plan on pulling it, and packing it away. The ECU has a tach output; which was the original plan.

Pretty crazy not seeing all the ignition wires coming out from the distributor. They feed in from the pulley side along with the injector wires – which I thought would be a cleaner look. Also, you can see the air filter hanging out under the plenum from this angle.

Hoping to get a basic tune going this week. Still lots to do, but it's nice to hear the car running again.

Image Unavailable, Please Login

 

Status update.

No drive yet, but I have her revving to 5k!

I had to do two things before I could get back to tuning. Both of which fall in the category of "don't do what I did".

1. I originally simply tapped the fuel rails; when I really should have spent the $200 for the proper 6AN ORB drill bit. So I went back and bought that, then added the ORB chamfer. After a few heat cycles, no leaks and solid pressure. For anyone interested, here's what the fuel rail process looks like:

1. Purchased raw 24" rails from Ross Machine Racing.
2. Cut to length. I 3D scanned the intake runners while installed on the car to determine total length and injector gaps (95mm)
3. 3D printed a jig to mark injector holes.
4. Used a special injector drill bit from Ross Machine Racing to drill the 6 holes.
5. Drill the inlet and outlet ORB chamfer using a special "porting tool" drill bit from SCT.
6. Add the 9/16-18 UNF threads. I 3D printed a jig to keep the tap straight.
7. 3D printed a jig to mark holes for the rail brackets.
8. Drill two M6 holes.

You end up having 13 opportunities to turn your fuel rail into trash.

2. I originally left the mechanical advance in the distributor. I did this because I wasn't exactly sure how I wanted to handle the mechanical advance. Idle uses the static advance, and since I only planned on idling, I'd let the ECU handle the base timing. Well, that worked, but it was obviously a short-term hack. My solution was to 3D print an insert that "locks out" the distributor (I used PET-CF). Easier to show with pictures than to explain, but with this in-place, the distributor is completely locked out, and the ECU handles all timing. I was able to stick to my "nothing destructive" rule as well. I currently have the ECU programmed with 10 deg idle timing, ramping up to 29 deg by 5k RPM (basically, the user manual's specs). Once I get on a dyno, I can see to what degree the engine is "knock limited", and play with timing.

One thing I did to aid in tuning was to join the plenums. I completely closed the 7-12 idle set screw, and I have a vacuum hose connecting the plenums where the old AAV connections were. This helps smooth out differences in vacuum that can throw off the fuel calculation. Fuel is still two separate systems (with separate regulators set to the same PSI).

There are a few more items that I need to go back and cleanup before I can do a road tune. Overall though, the car is running really solid. Temps and pressures are fine, and both banks have the same AF ratio. At this point, I think I am at a point where the hard parts are past me, and now it's just cleanup, testing, and tuning.

In-case anyone is curious, I wouldn't say it sounds any different than a well tuned CIS car.

Image Unavailable, Please Login Image Unavailable, Please Login

 

I’m curious about the dyno tuning necessity.
I dyno tuned my Motec track car. When we stated running on track the dyno tune was at best a baseline. Aero drag temps load it’s all so different out in the wild.

maybe on road the opportunities are too limited to tune?

 

I think this topic probably warrants some context / proportion. I'm not trying to offend anyone on their preferred style of tuning or make claims that there's only one way to do it. So here are my 2 cents:

I could get a rough tune in my driveway to get it safe enough to get out on the street. Fuel maps (A/F) are pretty straightforward and I can set the timing to what is in the user manual. I'd drive down the street, look at the logs, see where I was lean / rich, and tweak. I could stop there, and almost assuredly have a better tuned car than I had with CIS and mechanical advance. Easy enough.

The next level of tuning is based on identifying the volumetric efficiency (VE) of the engine at the intersection of vacuum and RPM. VE is essentially the percentage of efficiency the engine has at moving air from the intake out the exhaust at any given time. This is a newer style of tuning that post-dates the "gold box" Motec ECU. This style of tuning requires holding the RPM, and changing the vacuum (throttle) while you measure torque, A/F ratio, and watch for knock. This is where you can really push the timing to extract the most torque (without knocking). Eg, a "steady-state" dyno can hold the car at 4k RPM while you change the throttle – which is wholly different than putting the car in 3rd gear and flooring it. This level of tuning fidelity is what will really unlock the possibilities of EFI – especially after the internals are upgraded (cams would greatly influence the VE).

There is an in-between, where you find a long stretch of road, and try to hold the RPM. You're limited in your ability to change vacuum while doing this, but it's better than a "rip down the street and see what happened" style of tuning.

Hope that helps explain where I am coming from on this. The above is written as fact, but it's really just my opinion. Caveating every sentence with "in my opinion" would just drag it out.

 

Having done EFI conversions on a variety of engines, unless its forced induction I dont dyno tune. You can spend a lot of time and money squeaking out a few more HP which will go mostly unnoticed when you drive it.

The Haltech systems I use learn as you drive based on the wideband O2 input after I set the initial table values. I get the fuel tables close based on experience, the ignition I'll pick the base timing and total timing at 3K. The fuel map re-writes itself in real time as I drive. I've watched the tables get tweaked based on the O2 input while I drive. I double check emissions with my gas analyzer as well to see the actual HC numbers. If I then dyno tuned it after that I might get 8 HP more? Or not. The last forced induction car I did I put a 3.3L whipple on a FordGT, used a canned tune then dyno tuned it. 751HP with the canned tune and 770HP after a half a day of dyno tuning.

 

I imagine most would agree that a day's worth of dyno time isn't worth a 3% bump in power unless it's a real track car where every HP counts and you're tuning more than just power.

I'm happy to be the guinea pig on this one. I'll do my best to street tune it, then I'll compare to what I can get on the dyno. I'll report as I go. Curious myself to see where the diminishing returns are.

 

Very interesting
Do you have any stock dyno numbers for your car. I know different days and temps would distort that anyway but would be interest to see what the straight efi conversion did in numbers.

in any event the difference in the way it drives will be apparent and that’s the important.part

 

Unfortunately, no. I was having CIS issues with the 7-12 bank; so a dyno run wouldn't have been a fair comparison (distributor probably needs a rebuild). I had wanted to resolve that issue first, then start the EFI conversion, but it was dragging on, so I switched my focus to the conversion.

Is there a general consensus on the BBi's peak torque / horsepower? I figured it was in the 325 hp range.

 

I believe mine made 312 on a mustang chassis dyno. Calculated HP, I forget the wheel HP but it would've made me even sadder so I think I mentally blocked it out. Power peaked just over 5K and began to fall after that and I blame the cams for it because a cam swap during a rebuild completely transformed that same car. Im doing another one now with the same cam profiles I used in mine, it'll really wake it up.

 

Wow, that's just sad. I don't blame you for not wanting to see the power at the wheels. You run the risk of seeing a number that starts with a `2`.

Really looking forward to seeing what cams, and lighter internals do for this car.

 

Based purely on feel I’d say hp def not more than the quoted 340 and likely 10% less

however Malay leery of peak hp vs power curve same with tq
BBs have a fairly broad power range, imo a bbi is flat and falling off slowly after 6200 rpm which is the biggest issue. Cams would fix that and positions can up the hp, efi can make it all work great

seems like a real 400-420 hp is possible carried to 7k and beyond. With itbs cams pistons maybe north of 450. What’s weak as you go past 7k rpm, oiling ?

 

Im doing 2 flat 12's and 3 V8's right now, the right internals transforms them. Ferrari left a lot on the table with these engines but they had more rules to follow as well.