MAF vs MAP sensor?

You can buy them assembled for about $400....but they are not really very good. You get what you pay for. They don't have anywhere near enough load/rpm points to tune the engien well, particularly if boost is involved. They have 1 iginition channel, so you need other boxes if you want waste spark ignition.

Don't be afraid of the tuning, it's pretty straight forward to get aftermarket ECUs up and running. I'm playing with a Motec now, and it's about the least user freindly ECU I've experienced, very flexable and will do anything, but a lot of stuff to set-up because of that. The Haltechs and FAST ECUs I've played with are much quicker and easier to set-up (and install) but they get that ease by limiting you to more common set-ups.

I think your first suggestion of finding someone who can re-map the stock ECU is the best option for low cost. A naturally aspirated engine will run just fine using TPS vs. rpm maps, that makes it like a very good carb set-up. If you can add a MAP sensor open to atmosphere then it will be about perfect and is the perfered set-up for a performance NA engine. The stock ECU should handle that, but you need to be able to re-map it as the maps the factory put in are only correct with the stock intake and exhaust system. The ECU does make corrections on it's own using the O2 sensor, but only at low power settings

 

Well I am a good part replacer but do not know what all the parts actually do. What I do kwon is a friend who is an ex microsoft programer. He has looked at all the systems and was going to do motec and instead did MS. He has all the $ in the world as a retired at 30 y/o guy but chose MS because it works. He has no issues and is able to use it to be SCCA solo prepped at the national level while everyone else is spending big bucks for motec. He said it was so easy I think he got his whole thing installed and running in a day and tweeked to run at the national level without a dyno in a weekend of real life testing. He is a smart guy I'm a dumb guy so I figure maybe it would take me a week? I say don't underestimate the simple. MS has a huge following and they seem to do everything a motec unit can do as far as be a competent unit that can pass a sniffer test in Cali and work on wide band 02 sensor. Of course no aftermarket programable ecu is smog legal however but it does have enough steps to pass.

 

That doesn't surprise me...MS is an open source system, that means your microsoft guy can mess with the programing. The MS does work, but you end up with a patch work of boxes, and a slow processor and very limited set-up options unless you can write the source code to make it do what you want. MS in kind of the ECU equivalent of linex.

When you buy a major brand ECU you get code that works and a single box to worry about. Much simpler. I looked at them because I don't have car budget to waste....but I bought a motec because I know it will work.

 

One thing I like about MS it is so cheap that if I can't make it work I can throw it away. The motec is like $5k or something like that with just the most very basic parts of their system. Also the MS guys are all there to help you on the web just like FChat someone always knows what to do. Motec is right down the street from me in Huntington Beach and I can't get them to give me the time of day even before I have bought the product. They think they are Ferrari. I can't even get decent presales support. I can't imagine what I will do with $5k worth of wires and questionable post sales support. I think microsoft when I deal with motec. I would seriously think electromotive if the MS did not work. They use a slightly different set of rules to make their ecu work and lots of racers run them successfully. I think the place where motec and EFI system is so good especially in racing is the data aquisition is that right? Did you really buy the motec because it has a map that can take 64 chunks of data or whatever vs MS that can do like 8? I do not know the terms but I don't see how the speed or number of chucks is so important. Isn't it kind of like having a high hz o-scope to loot at spark plug firing that can be seen just find with a ghetto 20hz unit? I really don't understand that part can you explain?

 

Sorry for the length and I’m not a real tech- geek, so I won’t be offend if someone jumps in and corrects anything I say.

What you’re saying makes sense and a race only car is MUCH easier to run than a street car. It doesn’t seem like that should be true, but it is. Race motorcycle carbs never even used to have idle circuits in them – just extra parts, and it’s kind of the same thing with EFI. At ½-2/3 throttle and up the maps for fuel and ignition tend to be very smooth, I’ll bet you could get away with an rpm points every 2000 rpm and load points ½, ¾, open throttle and only be giving up a couple % on hp at any place in between. To make an engine start and idle nicely though, you need 5 or 6 load and rpm points, off idle throttle response, 2 or 3 more rpm points and another 3 or 4 load points….so at least 7 rpm points and 8 load points just to get it started and rolling nicely. I think at least 4 more rpm point and load points to get to ½ throttle and at least 3 more from there to redline. That means you need at least 14 rpm points and 15 load points with no margin for error and that is only for natural aspiration, for boost at least another 4 load points are required. My finding when I raced was that I could race with a bike that had bottom end tuning issues, but it’s sooo much easier when it doesn’t…best to have at least the minimum good running rpm and load points.

The MS 1 has 8x8, clearly not enough, I would not even consider using one those ever carbs would be superior in every way. It will work at the track sort of, but the low end will be more than sketchy. The MS II is 16x16 I believe and the rpm and load points are user definable (that’s key when working with reduced numbers), so it will work. In comparison, a haltech is 32x32 user definable, so there is basically no way you will run out of points. The Motec says it’s 40 rpm points (way way more than plenty) and 21 load points, but the latest software allows 24 load points…I think they realized 21 is kind of marginal for boosted engines.

The process speed and number of bits. The number of bits and clock speed are very much related. Basically, a computer does an operation every clock pulse. One operation could be moving information to or from memory or doing arithmetic of some sort. Computers work in 8 bit chunks, so an 8 bit processor moves 8 bits every clock pulse. A 32 bit process moves 32, or four groups of 8, so if the 2 chips are running at the same clock speed, the 32 bit chip is 4 times faster. A chip can also have a math co-processor which handles the arithmetic while the main processor is moving data in and out. For math intense applications (like EFI systems) this can effectively double to operating speed of the processor. For basic multiport fuel injection, a plain old 8 bit chip is probably fine. If you want to have sequential fuel injection and run the ignition, it’s not fine or even adequate, an MS 1 is right out. The MS 2 has a 16 bit processor with no math co-processor (I think), so IMO it’s ok for injection, but still cutting it close if you add ignition. It’s probably fine if you run multiport injection and ignition (remember it only has 1 ignition output, so you need a separate processor to run anything except a single distributor). For a race engine, multiport injection is just fine, sequential really only helps at idle and low power and is most helpful when the injectors are very big, like on a boosted engine.

When you don’t have enough processor speed, you notice it mainly in 2 places. In the lower gears, particularly first where you build rpm fast, but the ECU is using maps the wrong maps because it didn’t have time to get the right ones maing your fuel and timing worng and the engine sluggish. It also shows up as poor throttle-on performance because it’s using the wrong load maps because it didn’t have time to load the right ones which can cause detonation because higher loads often what less timing. The units with 32 bits and a math co-processor will run sequential injection, waste-spark ignition (they still want a separate processor for direct fire over 6 cylinder, that should tell you something) and never be on the wrong map. You also get idle control, outputs for electric fan control, warning lights, shift light, what ever, you get 8 or 12 of then to do with as you please (I’ve found uses for all of them). You get 2 sets of maps (you can pull the power back in first gear to help with launch) or the motec will also do launch and traction control. And you get cylinder trim for fuel and ignition to fix any across the board imbalances or the motec will let you map individual cylinders to fix local imbalances.

I ran a haltech E6k on my car which was a 16 bit ECU, but I think faster than a MS II and had 22 load points and 32 rpm points. I ran multiport injection and waste spark ignition and it worked pretty well. When I switched the newer faster 32 bit haltech E11, there was a very noticeable improvement throttle response and idle improved with the sequential injection. My E11 is a v1, which I got really cheap because haltech recalled them because it has a few problems and I bought one of the ones they got back. I’ve struggle a bit with even mixture at low power (I have very big fuel injectors), and 1st and 2nd gear are very hard to control. I decided to switch to a motec to solve both problems. The cylinder mapping feature will cure my low power imbalances (caused by very slight mis-matches in very large injectors) and the traction control will make it safe for me to let others drive my car and easier for me to handle…and the launch control will be really nice leaving the line at the autocross.

My recommendations are:
1. motec. It is the best unit on the market and will do anything you ever want to do. The down side is cost…they charge for freakin everything. They are much cheaper if you order from Australia. I paid $2800 for an M800 ($2470), flying lead harness ($173) and programming cable ($135). I also bought the traction control for another $375. I passed on the data logging ($190) and the wide band O2 ($680). Options like the traction control, logging, O2, ect are all just buying a code number so you can make upgrades anytime. Ignition modules are extra, figure $400 to run your 4 coils.

2. Haltech E11v2 or Electromotive. They are both good units. The haltech can be had for $1600 with the wiring harness, but you need ignition modules, again figure $400 for basic stuff. Wide band O2 (5 wire)is an extra $300 including the sensor, I think it does 4 wire for free. The electromotive uses there own coil packs and I think runs about $2500, I don’t know as much about these.

3. Haltech E6x. It’s a good unit and runs about $1200 with the harness, but needs to be run as multiport injection and only with high impedance injectors. Again you need $400 for ignition modules. This would be a very streetable set-up at a reasonable cost.

4. MS II with the motronic. Leave the stock motronic running the ignition and install the MS II to run the fuel. This would be a very simple install (you can trigger the MS from one of the coils, nothing else to install) and give you good control of the fuel. This is probably the cheapest way to go and should yield a good running car, but you are stuck with the factory ignition timming or spending money for a new chip for the motronic. . This could be done for $500-$1000 I’d guess – remember the $400 assembled MS II doesn’t have a wire harness (haltech and motec use tinned mil-spec wire that runs about 50 cents per foot), programming cables, ect). There are other piggy-back unit that could also be used and could also do some ignition tuning, but I’m honestly not that familiar with them.

5. MS II. It will work, but is better suited to the track than the street. The basic unit is about $400. You’ll need an ignition control and probably modules as well, I’d guess $1000 for the extra stuff unless you get the stuff from a salvage yard. You may need the coils to match you ignition control. You need to build a wire harness, or I guess you could cut-up the OEM one. I’d plan on not less than $1500 and would not consider it street worthy.

 

Uhh.. NO.. those "tables" aren't tableS, they are TABLE.

One table, inputs are adapted TPS angle (called Alpha) and engine RPM
(called N) hence the moniker "Alpha-N".

The output is a synthetic maf value, determined under standard conditions
on the dyno with the specific engine/vehicle/etc.

They don't affect "timing" at all.

Jim Conforti

PS: Simba: this is what I do for a living - I have a bit better grasp of it than
you might imagine.

 

BTW, guys.. Alpha-N has been used on race cars for years. See: BTCC/DTM/etc

The problem with MAP sensors on a vehicle with multi throttle bodies is that
it becomes very hard to get a representative MAP value from the engine.

Aka: define "manifold".

A car like the 328, with a common plenum behind a single TB makes obtaining
a proper MAP value easy. Something w/ 8 TB's makes it much more difficult
to get a stable MAP value.

Aftermarket ECUs use all sorts of kludges to bypass this.

Jim

 

And is the preferred set-up for high performance natural apsiration. MAF are restrictive on plenum type intakes and or impractical and restrictive on multi runner intakes.

MAP sensors are not real accurate at or near barometric pressure...like at wide open throttle. The problem with multi-runner is generally solved with a small pressure tap in each runner then T'd and sent to the MAP...still not perfect and still not accurate at WOT, but workable. The aftermarket ECUs generally offer a WOT map to so the MAP reading is ignored after a predetermined throttle position.

Do you mean piggy-backs? The stand alone units do really need to kludge anything, they have a clean sheet of paper to work from.

 

I guess it depends on the manufacturer. The are 3 pieces of information for each point, so it can be a single 2-D table with text in the boxes, a 3-d graph, or a series or 2-d graphs. I've seen them all used by one manufacture or another.

 

No Mark, you haven't as inside the ECU, the tables are all of similar design.

There are two pieces of independent information, and one piece of
dependent information MAF = f(Alpha, N)

That is a 2d data table, regardless of how you present it to "humans".

BTW, show me "text in the boxes" inside an ECU.

 

No, I mean stand alone units.

Which, regardless of a "clean sheet of paper" as you put it, still have to
follow the laws of physics.

Attaching a MAP sensor to multiple throttle bodies requires creating a
balance tube (physically) and attaching said tube to the intake port side
of all throttle bodies, with the sensor connected to the balance tube.

The volumes of the different entities come into play as to the response
of the system then, it (generally) especially sucks at idle.

Some standalone manufacturers kludge this by "blending" the use of Alpha-N at or
near idle with MAP/Speed-density off idle.

 

Congratulations Jim. In only 20 post your confrontational style has earned you a permanent membership on my ignore list…I think that’s a record. Good Bye Jim.

 

Mark thanks for taking the time for the long reply. You are helping me to learn and I thank you. Tell me more about the piggly back idea. Lets say we run a standard motronic ecu in the 348. You put in some new headers and bigger intake but you need more fuel to make the changes really work. So you leave the motronics intacted as is and you overlay the MSII to dump in more fuel so it will work? Is this as simple as buying an MSII and hardwiring right on top of the motronic and them programing extra fuel via the fuel injectors with your laptop?

Or I could use the MSII as a fuel controler to run extra fuel
to the intake when I run nitrous into the plenum, thereby avoiding having to flow a wet mixture of fuel and nitrous in a dry manifold???? If yes please tell us more!!!!

 

Mark,

Get real - you are passing on information that is false/misleading to people.

Let's elaborate on that, shall we?

I'm talking about the internals of OEM engine ECUs.

The display of the data on a PC can come in various forms.

Example, you can connect a Bosch ECU over the K line to SAMS2000
or the any number of ETAS calibration tools from VS1000 to INCA VME
or you can run an emulator device in the ECU.

In every case the data will "look" different on a screen, but the underlying
data is the same in the ECU.

As someone who's WRITTEN Alpha-N conversions for various cars, I have quite
a handle on this stuff.

The original statement which I made and you tried to 'correct' was that the
Alpha-N default mode effects timing and also is "pig rich".

In fact OEM calibrationists are bound to do their best to make sure that the
default modes meet emissions as closely as possible. This means that if
you unplug a MAF, you should get very similar fueling and timing to what
you would consider "normal operation".

So to close this thought (as my ears are still ringing from 3rd row seats
at the Who tonite in SLC ) here are some conversions you can make:

1) Alpha-N - generally only requiring minor software mods.

2) MAP sensor - requiring mechanical modifications (addition of a properly
sized/volumed balance tube - if needed on a multi-TB engine), a MAP sensor
and software mods

3) Scrapping the OEM ECU for an aftermarket one.

Because I edit/modify OEM ecus, I rarely recommend the 3rd option.

Except for something like Megasquirt, they are far too expensive for what
they do - unless your car doesn't have an ECU. Still, adopting an OEM
ecu from another car can make a lot of sense in that case because all
of the parts will be available over-the-counter for repairs, and at "less
than Ferrari" prices.

Jim

 

A piggyback unit installs between the factory ECU and the injectors/coils or between the sensors and the ECU or both. It takes the output from your senors or ecu and adds or subtracts fuel and timing based on what you program into it. They are very popular with the import crowd because they are the cheapest way to make the car run.

I buddy of mine installs “power commander” units on H-Ds, which is a piggyback that I believe works by altering the sensor inputs to the ECU. He’s able to get them tuned pretty well, but the problem there is that you can’t change the injectors and once you tell it it’s at WOT there is no more fuel to be had. He’s trying to mess with variable pressure regulators to go with it, but is anxiously awaiting the release of a stand-a-lone that will fit in the available space so he can quit spending so much time screwing around and get on with tuning.

I talked buddy of mine at work who had a piggyback unit on his supercharged Honda. He said it worked, but it seemed like any time there was a problem, it turned out to be the piggyback and he wouldn’t do it that way again. He felt a stand-a-lone ECU would definitely be worth the money over the piggyback.

Jim at Norwood kind of told me the same thing. I’m pretty sure they used to use both haltech and electromotive before switching to motec. He felt that the motec was worth the extra money because with the motec he knows he can deliver a car that runs good with no quirks. He said making hp is never the problem, no one ever comes back complaining about hp….they complain about drivability issues. You know, “when I do this, it does that” stuff. I don’t want to put words in his mouth, but that was my take on the conversation….and that is exactly why I now own a motec. I got sick of having a 500+ hp Ferrari that didn’t ever seem to run quite right. Piggyback units tend to have a lot more of this type of little quick type issues.

I think the better option is to mount the MS in parallel to the motronic. That will probably mean a second TPS, map (I think it’s built in to the MS), and temp. You can trigger it of one of the coils. Then just unplug the stock injector harness and wire them to the MS and you’ve got complete control of the fuel. Now remember, if you have the MAFs in place, the fuel will be right anyway, it’s the TPS or MAP controled systems that get confused by changes you make to exhaust and such.

A MS could supply extra fuel for nitrous through separate injectors, but I don’t know if it has an input to trigger enrichment though the primary injectors. I know that haltech, electromotive, and motec all do. I also know the motec and haltech give you a few options on how you handle it an di think electromotive does as well. You can trigger it to trim the primary fuel/air maps for more fuel/less timing, you can switch to secondary maps tuned specifically for the nitrous, or you can fire secondary injectors and retard the timing.

When it’s all said and done, you really get what you pay for. I can understand not wanting to put money into a civic, but I honestly can’t understand trying to cob a Ferrari together. I think the very minimum would be a haltech E6x, it will do everything you are asking for. They turn-up used, but new in the box on Ebay for $700-$1000 or new from a dealer who will answer all the questions you have during the install for $1200-$1400. I just don’t think you’re saving anywhere near enough money to live with the short-comings of anything less. I had the older version (e6k) and I know a couple other guys with them and everyone is happy, they work good. For an extra $500 an E11 is a bit better, but I don’t feel it’s require unless you need to run injectors over about 36 lb/hr (the E6x will limit you to 14 ohm injectors without adding an external drive box and then it’s just as cheap to buy the E11- or an E8 which is an E6 with extra drivers built in for an extra $200). The motec may or may not make the car run any better, in my case I believe it will solve a problem that has been making me crazy. The motec will work with all the OEM senors, where the haltech may want you to change the temp sensors (although I don’t think that is true with the latest versions). The electromotive is also good.

 

I’m thinking this discussion would make more sense in terms that most people are more familiar with.

The MS 1 system is band brakes and not suitable for use on anything that goes over 10mph

The MS 2 or piggyback set-ups are drum brakes. They are cheap and work fine on the rear of a minivan.

The E6x is stock disc brakes, good for most applications but may let you down in more extreme set-ups.

The E11 or electromotive are the big brake upgrade, really good most everywhere and all that most people would ever want.

The motec is the ceramic upgrade….most people will never notice the difference over big iron brakes, but if you need it, you need it.

Unplugging a sensor and hoping for the best is like having no brakes at all.

Does that help get the option into perspective?

 

I'd first ask Andy Hls if he stuck with Motronic, and if so, if you could have his custom Alph-n map. I can burn the chip for you (if he owns the software rights to it and is OK with it) if he emails you the .bin file.

I think that he went with 8 Jenvey tb's instead of 8 355 tb's, but he left the tb's open and he's getting considerable hp out of that setup.

Frankly, I'm pretty sure that I can make Motronic 2.7 work for you as is. I've got to get a little deeper into the mnemonics/interrupts, but the system itself will output as much fuel to air as you want (as well as let you adjust the spark advance).

I may go blind from reading Assembler, though!

 

You're building the 348 Competizione LM engine. 348 guts + 8 355 tb's + modified Motronic 2.7 tables.

It's a proven race-winner.

Of course, the cam pulleys were set to the Euro position and the exhaust was a bit more open, too.




When you are finished, you ought to be able to give the 348 Brotherhood a step by step of what to buy, what to fabricate, how to set the cams, and what chip to use to get ~~400hp.

Heck, I'll make the chips for free for the Brothers who are going to the trouble of installing the eight 355 tb's for this project.

 

How easy is it to make a chip? The required maps will depent on the exact set-up for the engine, so even with someone elses maps, if the set-up is not the same, the maps will not be right. It may take a few tries to get right....but once you get it you get it right it should work great.

 

It's no problem to modify the assembly code to handle larger maps, if needed, and there is **plenty** of unused room on the 348's chips. Modifying a fuel table itself is even less trouble than modifying the code, of course, once you know where it is stored on the chip and under what conditions (e.g. hot, cold, idle, partial throttle, WOT) it is used.

The hassle is walking through 32k of uncommented, undocumented, disassembled machine code the first time.

 

You bet! Mark Thanks for taking the time to explain.

OTOH, what was landshark trying to say? I have lots of respect for you but really at our level playing with OEM is out of the question. We don't have the compliler/decomplier codebreaking blah blah blah stuff to burn eprom chips to tweek our motronic ecus. So to me it seems that aftermarket ECU's that are laptop programable are the only way to go. But if you are teaching a class in how to tweek OEM units on the sly please sign me up.

 

Well I take that back maybe ND can do it. If you do this ND are you not fearful that you do not have a means of data collection. Therefore, your only clue to something being wrong is the burn valve? One of the advantages of most user programable systems is data aquisition to evaluate engine parameters. I am not sure you just dump a little extra fuel in it.