The Multi Air Project

Sandgroper

True Classic
Continued from (here) so as to not spoil the SOHC/ K20 discussion thread.

The goal of this thread is to help build an understanding and build a list of what is required to adapt the Fiat 500 ( Dodge Dart) Muli Air power-train to fit and run in the engine bay of a Fiat X 1/9. Contributions of ridding the mechanical dependencies of the engine and manual gearbox as a package of the infernal CANBUS factory system and employing an aftermarket engine/gearbox electronic management system are most welcome. I would hope we could keep discussions to how and if this is possible not if it is 'economical'.

The desire for a gearbox Downunder (and elsewhere) that can reliably take more punishment that the X's 5 Speed to 'give it the beans' off the line and have at least a torque biasing diff ( LSD ) is a carrot that is just dangling in front of me. The 5speed that came wit the X 1/9 is both few and now VERY old, it was never a strong box and the crown wheel and pinion is weak as a many trans-axel makes of that era.

Multi Air

Fiat Group and Fiat Power-train Technology unveiled its air management technology 'Multiair' at the 2009 Geneva Motor Show. Multiair is an electro-hydraulic system of engine valves for dynamic and direct control of air and combustion, cylinder by cylinder and stroke by stroke without a throttle valve. 1.4 L engine is also produced also at the Global Engine Manufacturing Alliance in Dundee, Michigan, United States.

The Hydraulic Intake Valve Solenoid

The Multi Air system can be employed to achieve five defined profiles controlling the intake valve/in respect to the physical camshaft lobe profile. The exhaust is a traditional non phased mechanical camshaft.

Profiles
One = Locked on or rigid opening and closing angles similar to the exhaust cam lobes to utilize the full Camshaft lobe profile (Full Lift - full engine power).
Two = Predictive to anticipate the intake valve closing time (EIVC or early intake valve closing - medium/low RPM engine torque + engine partial load reduces pumping losses / optimize volume efficiency )
Three = Delayed or retarded intake valve opening time ( LIVO or retarded intake valve opening time - for start up & idle, to speed up intake velocity therefore optimizing combustion)
Four = Double lift intake valve opening in a single cylinder stroke (Multi Lift or 'double' intake valve opening at low loads or city idling/stop & go mode)
Five = Closed intake valve time ( or no intake vale opening - after run / decelerate) .


The actuating piston is connected to the intake valve via a hydraulic chamber, which is controlled by a solenoid valve. The movement of this piston is controlled by a mechanical intake camshaft. When the
solenoid valve is closed, the oil in the hydraulic chamber behaves like a solid body and transmits the lift schedule to the intake valves. The lift schedule is imposed by the mechanical intake camshaft.
When the solenoid valve is open on the other hand, the hydraulic chamber and the intake valves are de-coupled; the intake valves do not follow the intake camshaft anymore and close under the valve spring
action. The final part of the valve closing stroke is controlled by a dedicated hydraulic brake, to ensure a soft and regular landing phase in any engine operating conditions. Through solenoid valve opening
and closing time control, a wide range of optimum intake valve opening schedules can be easily obtained.

To help improve power, the solenoid valve is always closed and full valve opening is achieved following completely the mechanical camshaft, which is specifically designed to maximize power at high engine
speed (long opening time). For low rpm torque, the solenoid valve is opened near the end of the camshaft profile, leading to early intake valve closing. This eliminates unwanted back-flow into the manifold
and maximizes the air mass trapped in the cylinders.

In engine part-load, the solenoid valve is opened earlier, causing partial valve openings to control the trapped air mass as a function of the required torque. Alternatively the intake valves can be partially
opened by closing the solenoid valve once the mechanical camshaft action has already started. In this case the air stream into the cylinder is faster and results in higher in-cylinder turbulence.


Questions


Do you need all five profiles in a performance oriented engine swap into an X 1/9 or is the entire 'economy' fuel cycle of improved milage desirable. If it is just performance can one do away with profile numbers four and five. One for full power, Two for torque at mid RPM and Three for engine warm up/start phase. Or is just profile One negating the whole mapping exercise altogether and just mechanically locking the solenoid package 'on'. This would negate the requirement for inputs from engine / wheel speed, etc.

How is this actuating signal achieved, is it analogue voltage signal. Voltage and Impedance required of the ECU signal driver is what value?
The pressure to fed the solenoids looks to be ~40Bar, is it fed by the main oil circuit or is there an additional dedicated pump to drive.


Vacum Pump -Most likely for Brake booster can it be deleted ( the X has/needs no brake booster )


Turbocharger control - Electronic control but the English Forge aftermarket guys are already doing BOV's

ECU Inputs The Body Control Module indeed has inputs to the Engine Control Unit ( Module )

What sensor inputs does the manual gearbox require/ provide the ECU.










Aussies build great ECU's perhaps candidate can be found here.
haltech, autronic, motec, wolf, ems, adaptronic, hawk

An oldie but a goodie amongst DIY racers
http://www.autronic.com.au/Downloads.php

The expensive one ( because it is fully featured and damn good )
http://www.motec.com.au/home
 
Last edited:
I would think the factory ECU would be best given the unique nature of this motor.

The question then, can we emulate the missing pieces in software? If the BCM has no critical functions, can we emulate it on the CANBUS? If it DOES have some important interaction with the ECU, can we emulate the missing pieces going into the BCM instead?

Either way it would require snooping the communication on a running 500 first.
 
Vacum Pump - purpose?
I would assume it is for the brake booster. Multiair engines regulate airflow by manipulating the valve timing and lift in addition to a throttle valve, so there isn't always any manifold vacuum to tap into.
 
Last edited:
I can only offer one thing... well maybe two...

As this stuff is way beyond my comprehension, or what I feel I or most others would even need to comprehend. I commend you though for your knowledge and ambition.

1. Vacuum Pumps?

My only experience with them is what I have seen and read... as they reduce crankcase pressure to the extent that sometimes another 100 hp is found on 700+ hp V8's. Insane stuff... but I believe they can be found mostly on long-service NASCAR engines...

2. WHY?

Cost aside as you said... what are you trying to accomplish here. You could also install a J47 GE Jet Aircraft Engine given the time and passion to do so. It appears to me to be something that SHOULD be easily done, but with all the complexities for the gain in hp... there are so many other similar solutions out there... that make 50% to 100% more!

Even if you spent the time and effort to facilitate a "bolt-in" conversion... just how many others would have the passion (time and money) to bother with it?

In this world... I find that Picture Puzzles are a ridiculous waste in almost every aspect one can imagine. Even when complete, very few folks will bother to glue them down and frame them for posterity. I have found that they do have one purpose, and that is CONSUMING TIME... especially good in surgery waiting rooms... I fear though, the cell phone may have replaced them these days.

Anyhow... there may be a GOOD reason you want to pursue this, at least in your own mind and life... but I'm almost positive its not shared by too many others. I do encourage you to follow your dream though, as long as any women and children are not neglected or negatively affected.

As always, I have found that MEN usually do things usually for TWO reasons, a GOOD reason and the REAL reason. Most of the time they don't come close to coinciding! HA!

Continue having fun!
 
What is really needed is a GM "Tech 2" type solution. Where you can turn off the bit's in the ECU that aren't needed.

I'm sure Fiat has these and it is possible most likely.
Again,Like Tony.. I have no clue...lol
 
Easier to emulate missing pieces than start from scratch...

I would think the factory ECU would be best given the unique nature of this motor.

The question then, can we emulate the missing pieces in software? If the BCM has no critical functions, can we emulate it on the CANBUS? If it DOES have some important interaction with the ECU, can we emulate the missing pieces going into the BCM instead?

Either way it would require snooping the communication on a running 500 first.

I'd agree with Steve on this one. Given the complexity of the MultiAir system (which effectively modifies the cam profile on the basis of a number of factors) I think trying to use an after market ECU would be extremely difficult. Most after market ECUs handle fuel and ignition timing, and even with just these two parameters tuning them well can be very time consuming.

Even though Matt has already looked into this, and I do not underestimate the complexity of the task, I'd imagine that fitting as much of the existing engine and associated wiring and sensors would be the way to go. For anything that could not be brought over to the X (for whatever reason) then one would need to either emulate the functionality, "trick" the ECU or (with enough inside knowledge) find a factory operating mode (or reset) to get the ECU to ignore that parameter.

So we simply need to find one of the Fiat engineers involved with the development of this engine and buy him a couple of beers..

DOm.
 
Very interesting article.

Ok, so going on the possibly false assumption that we can emulate or fake anything we need to for the ECM to be happy....

What fitment issues were there? Did the whole engine/trans fit without any/much cutting? Does the Abarth setup fit?

What magic is needed to adapt the shifter and linkage?

Other than a mess of electronic dependencies, are there any other barriers?

Was emulation of the missing electronics already deemed impossible by someone?
 
Emulation of signals

It's not impossible, just time consuming, and expensive from the equipment needed perspective, and the only people who would normally have the right equipment are people who perform the function for a living ($$).

As an example, I'm going to reference one of the (technically) simple conversions we performed a few years ago. The customer had a need to convert one of the industrial sized 3D stereolithography systems (3D model maker) from an older gas discharge UV laser to a more modern solid state UV laser system. Better laser lifetimes, higher peak power over lifetime of the laser, which translated to more work being performed, and lower utility costs. The cost analysis was obvious and in favor of the conversion, so $$ flowed into the project quite liberally. In this case, all of the I/O commands came through a simple RS232. Yeah, remember those, in use since at least 1982? The sniffer (recorder) for the signals was likely almost as old. RS232 uses either two or four lines of data I believe, again, not my specialty. I also don't think it multiplexes (sends MANY signals on the same line overlapping one another). One line, one signal at a time.

Now a modern multiprocessor control system like that found in a Fiat 500 or any other modern, complex product, is going to have dozens of I/O lines, with MANY signals multiplexed within each. Sure, it's a daunting project. But it absolutely can be done. Someone is going to have to outlay some serious cash for the first 'stand alone' conversion kit. Will it be worth it? Remember that line from the movie Jaws? You're going to need a (bigger) specialist.
 
It's not impossible, just time consuming, and expensive from the equipment needed perspective, and the only people who would normally have the right equipment are people who perform the function for a living ($$).

This would probably be best accomplished as a hobby project rather than commercial. I can see where it wouldn't be cost effective for MWB to invest in someone resolving all the electronics when they already have a product. As a hobby project though, this would be fun and the equipment (the kind a hobbyist would use) is inexpensive.

Ideally the way to do this is to buy a running 500, preferably a beat up but functional one. Then, just drive it and proceed to data log communication and attempt to emulate subsystems that won't be moving to the X. After having worked your way past the check engine lights and stuff you have a running solution ready to transplant.

Does the CANBUS have well defined bus addresses for the subsystems, or does every manufacturer set their own (i.e. are airbags are always at "5" and BCM is always at "8", etc)? I guess I should read up on CANBUS.
 
The long and winding road

Thanks Dom adds a bit to what I'm looking to add.


"Either way it would require snooping the communication on a running 500 first."

Perhaps someone already has done this!


Thanks Coupfan you raise some interesting points, but modern comms are not as scary as some fear. I had to live with a Citroen with a Com 2000 unit for a while PITA but worked it out. Break down the actual component and what it is looking for and you can reverse engineer anything. The Asian market has been doing this skunk works for decades. My sea faring days SATCOMS and link 11 were just as interesting, as uploading analysis of bore-scope results on a LM2500 gas turbine to the Defense contractor in NZ from the Indian ocean became a challenge. Nowadays it is a sinch with Network Centric Warfare and INMARSAT-B. At the end of it I remembered what an 'old salty' said to me when I was a Or'd seaman - "Think sideways".

Papa Tony good questions I have added it to the 1st post as I will any findings. I don't for a second think that the 1.4 Multi Air Turbo is only capable of 250 HP, the Old 8V Uno Turbo could be make these numbers but it came back to how much air the head could process so the brick wall came up about there. The extra 8 valves in the 16 is just the beginning, as every 16 valver I have messed with flow begins where the 8v became too expensive to develop further. Once that happens then a 16 valve head becomes more economical to improve unless you are constrained to 8v by class regs etc. Nice to have someone 'blow sunshine up my a*$e' but I don't feel worthy of the praise. Once mods become a reality then folks are more likely to be interested- actually do them. Bob Browns excellent work is a shining example. Just that I like to drive like the old salty said, sideways, way reminds me that most problems have been attacked by someone before, therefor finding that path trodden by others (am I lazy ?) and gather the lessons. It makes ones task that much easier, I guess that is what I am trying to attempt here. Doing it on my own will be frustrating, sharing the task and making the results a GNU XWeb open project will benefit anyone with the great chassis that the X has, under-powered she has always been. The engine and the gearbox are in no way outperforming it, with around 200 Hp I have not felt it was beyond its ability. The gearbox I have always felt has been a limiting factor and the unavailability of an affordable LSD ( the Quafe Uno unit did not fit from what I could see) makes this many horsies a Porsche like 930 Turbo Widow-maker out of an X on a wet apex - ask me how I know.:eek:

"The REAL reason" is that all these 500's will be in scrapyards after someone with more enthusiasm than skill ( or just damn unlucky) crashes them. This pegs my curiosity - are they an opportunity to upgrade with an engine/GB that does NOT need rebuilding because it is soooo old. Its not that I want to build it because I can more that the opportunity for a choice power train seems to be in the offering. I am just curious, and trying to understand how the 500 Turbo works I would probably do it chassis I have decent gearboxes I do not.

Having put a modern 16valver in a Peugeot 205 with a Quafe LSD is just parallel - that thing just came alive. With a 16vTurbo, it was just damn insane, it is just a pity it was not mine. But no matter a FWD with 300 Hp is a recipe for disaster without TC as the front end is just way too light - but I think you get the drift.

I applaud Matt for his excellent work and as I said in North America it makes sense, I know he has invested a LOT of time and effort in the K20 conversion and the aftermarket crowd has a LOT of parts to make a hell of a bunch more ponies. I just have personal reservations about the cutting and shutting. The X chassis is way over engineered so I have full faith that it is a strong as it needs to be. I guess I am just lazy in that I would chose to spend my effort on ones and zeros as I am comfortable with them. Micro processor control is just as a tough nut to crack for me as performing the mechanical details. So I am willing to have a crack at it. I have no intention of sitting down and building a Megasquirt project on my lonesome, I'd grab an ECU off the shelf and play the inputs and AFR's, that's why I am asking what the 500Turbo needs for inputs. That would make a shopping list to exclude the off the shelf ECU's that won't cut it. I guess that folks need to have a look at the Euro DTM race series etc and how Bosch, Motec and other high end off the shelf ECUs have grown in step with all the electronic camshaft phasers, ABS etc. That are expensive because they are so damn complex. The software is not, sensor list is the issue.

My time with the Uno / Punto Turbo's taught me that a decent ECU costs money but in the end it is worth it as Dyno's are out there with tuners that work every day with them, finding a good one is like finding a good doctor or dentist. Some I hold in higher regard than some doctors I have had the misfortune to experience. So far I guess I have spent around 30 hours on them since the time I blew up the Uno T series one and learnt that you need decent Air Fuel Ratio maps and boost controllers. Good maps cost money - once. Once there they can be shared, and are the new base maps so become the start of the other guys install, he improves and shares so do you. Perhaps that's why I'm a Linux / Android guy.

Anyway. So I took at look at what the North American Nuvo 500 aftermarket is up to.

Didn't take long to find at least the beginning of a reading list as I have been tagging Web-link favorites on the various forums world wide with interest over the last few days since I mused the question. That most car crowds, be they rice burner turbo lads or cubic inch muscle devotees will, have an after market 'tuner' segment, this is where those who like to extend the stock capabilities always exist.

What do we think of this proposition that someone may have done some of the work and has the question already been answered? I haven't looked that hard.

But have a look here, these look like 'piggy back' systems which talk to the Fiat ECU. I may have taken the 'mick' previously about all the trinkets the 500 lads accessorize their cars with but there is some serious stuff going on at places like this.

500 Forums
http://www.fiat500owners.com/forum/...1521-tmc-motorsport-magnetic-marelli-ecu.html


Madness 500 in North America ( does this gear require the BCM or can they program their unit to disregard it? )
http://shop.500madness.com/index.php?route=product/product&path=319&product_id=2669

Their Magneti Marelli ECU
http://shop.500madness.com/index.php?route=product/product&product_id=1055

The Forge BOV
http://shop.500madness.com/index.php?route=product/product&path=319&product_id=1842

Harness
http://shop.500madness.com/index.php?route=product/product&path=319&product_id=2382

Ported TB
http://shop.500madness.com/index.php?route=product/product&path=324&product_id=2073

And this looks lke the major competitor in the UK
TMC Motorsport
http://www.tmcmotorsport.com/abarth_engine.aspx

Road Road Motorsports in Santa Fe
http://roadracemotorsports.com/

I like the look of the Madness ECU whilst the maps are their own hard won and earned proprietary information, how hard would they find it to remap for an X 1/9. Folks like this and RRM with extensive background in Japanese Turbo cars I think not hard at all. The pedal box and loom look like 'bingo' someone is paving the way thru this jungle - follow. The throttle is I am guessing is fly by wire so that helps.

From what I can tell so far RRM stopped selling the Marelli unit because they offer their own (based of TMC unit I believe) The differences are hotly debated, with neither side showing any proof really.

The gearbox is now tugging at my curiosity, where does the selector enter the casing or is it too fly by wire.

WRT the gearbox I know there is a manual version, is this a clutch less direct shift or a real manual unit? If there is a manual Abarth unit then the whole question about inputs back to the ECU ( or ECM as the 500 guys call it ) are so much less involved. The speed sensor is in most likelihood working off one of the ABS sensors as most modern cars do. Can a Punto / 500 CV or a chopper wheel be adapted to an X CV? Probably, it's really a just a question of knowing a good machinest. (I do - Grahame loves a challenge, us old farts are like that)

So a reliable 240 HP is already here and for nuts like me that are not afraid of breaking things there seems to be room for more AFR tweaking on E85- Hhhmmmmm. New learning curve of Multi Air AFRS with a bigger snail surely someone is crazy enough to have tried it. As has been so eloquently said - you have enough horsepower when you can leave two black lines from one corner to the next. Then you have a handling issue.

All I get is the feeling that I should have not let that Abarth pass in at Auction, because I thought "nnnaaarr too damn hard" I'd never get that in an X anyway.

Happy to hear that I am wrong as this is just a discussion about what might be possible, no one gets killed and no egos are to be harmed in the process. From my casual back of the cigarette packet measurements it looks like the power-train would probably fit, but until a engine bay mule is offered up a power-train we can't know 100%.

For the price of 2 rebuilt 5 speeds Downunder that Abarth could have been in my yard. The next one will NOT slip thru may hands if it goes for a similar price.

Anyway - discuss. I am off to the Fiat 500 interwebs to see what the heck they are up to. When I come back I will report what sort of shenanigans they are up to and if it looks promising.


Sandy
Share the knowledge
 
I *thought* the Madness ECU piggy back unit was made by TMC. Any case, it is a pre-programmed multi setting tuning box. Just like the Magenti.

I have it on the Abarth along with the go-pedal.
 
I think with all that you said Sandy...

Your REAL reason actually closely matches a GOOD reason.

Now to execute it, simply and cost effectively...

Thanks for your response... and I don't blow wind (figuratively)... I try to give compliments where and when they are deserved.
 
Does it work

Thank you Tim

I am just a bambino WRT the topic, so actually owning one yourself you are more likely to be right.

Did the bits you added make a 'big' or moderate increase to the engine power. Perhaps you could share the experience. We are in the 'discussion section so I guess the Moderator do not find it contradictory to the rules.

Any pointers as where I should look for information would be appreciated.
 
Cheap I don't think

Thanks Tony

Not sure the first will be cheap, maybe once a few have trodden this path the price would come down as better, more cost effective ways are found. That's the nature of collaboration and development to me.

I think they call it the bleeding edge when buying new technology or doing this sort of project.

As in the blood sweat and tears or the money bleeding from ones wallet, I think the tears are more of a problem at this stage.


No doubtsome one somewhere is musing the same thoughts.
 
I think the FIRST question to ask is what does it take to fit the drivetrain physically into the X. All the rest of this is probably pointless if it's not easier to fit than other options available in the US.

Ignoring everything else, what is involved in mounting the engine and transmission in the engine bay, connecting it to the wheels, and connecting a shifter?

I'd really be interested if it's possible to come up with a kit for an average shadetree mechanic to drop in one of the plentiful base model 500 drivetrains (again, ignoring the electronics challenges for now).

If Matt would be willing to share more details on what he found when they looked into this it would be very helpful.

On the other thread he said "the mounting is a little more simplified". Just "a little more" isn't encouraging.
 
I think the FIRST question to ask is what does it take to fit the drivetrain physically into the X. All the rest of this is probably pointless if it's not easier to fit than other options available in the US.

Ignoring everything else, what is involved in mounting the engine and transmission in the engine bay, connecting it to the wheels, and connecting a shifter?

I'd really be interested if it's possible to come up with a kit for an average shadetree mechanic to drop in one of the plentiful base model 500 drivetrains (again, ignoring the electronics challenges for now).

If Matt would be willing to share more details on what he found when they looked into this it would be very helpful.

On the other thread he said "the mounting is a little more simplified". Just "a little more" isn't encouraging.

This video was posted in the other thread (that almost got hijacked).
http://vid1276.photobucket.com/albu...rth-500-X19/ABARTH500-MOCK-UP_zps09f61131.mp4
might help with your questions on fitment
 
I must have missed that video. It looks like fitment is VERY good. I'd be looking, initially, at the more available non-turbo version so maybe it wouldn't require cutting the firewall out?

Yeah, I know the turbo version would be very desirable. I'm trying to look at it as, given a good supply of junkyard non-turbo cars, is a budget swap in the average garage doable if the electronics can be worked out?

Maybe a future turbo swap could build on the work, but targeting the base model has benefits; it's cheaper and easier to find a used daily driver testbed, potentially fewer subsystems to eliminate/emulate, cheaper and wider availability of donor cars when the work is done, and might provide an affordable option benefiting more people.

I'm curious what needs to be done for the shifter. Is it a cable shift, or dual rods?

On another subject, I notice that Wireshark actually has CANBUS support when running on Linux. I know Wireshark is fantastic for monitoring Ethernet.

This is starting to get my interest in the X again. After blowing the BVH 1300 and failing to get the Uno Turbo engine functional I've not really been bitten by the bug again until now.
 
I must have missed that video. It looks like fitment is VERY good. I'd be looking, initially, at the more available non-turbo version so maybe it wouldn't require cutting the firewall out?

Yeah, I know the turbo version would be very desirable. I'm trying to look at it as, given a good supply of junkyard non-turbo cars, is a budget swap in the average garage doable if the electronics can be worked out?

Maybe a future turbo swap could build on the work, but targeting the base model has benefits; it's cheaper and easier to find a used daily driver testbed, potentially fewer subsystems to eliminate/emulate, cheaper and wider availability of donor cars when the work is done, and might provide an affordable option benefiting more people.

I'm curious what needs to be done for the shifter. Is it a cable shift, or dual rods?

On another subject, I notice that Wireshark actually has CANBUS support when running on Linux. I know Wireshark is fantastic for monitoring Ethernet.

This is starting to get my interest in the X again. After blowing the BVH 1300 and failing to get the Uno Turbo engine functional I've not really been bitten by the bug again until now.

Significant differences between the stock multi air, the Turbo multi air (not too much) then the Abarth. The Abarth can make a LOT of HP and torque for reasonable $ - we are talking 240HP possibly. I don't think it would be worth the effort unless one went with the Abarth. The cool factor of installing a *real* abarth motor in an X on its own is enough for me ;)
 
Back
Top