Speed density vs VE, '88 electronics, thoughts..

[RoadWarrior222]

Hi folks,

Been getting out of the rut and reading what helps other motors. Now those folks who have had the opportunity to play with both MAP (speed density) and MAF (Mass air flow) based systems on the same engine/drivetrain/chassis have figured out something that it's probably taken 3.0 folks a long time to come round to.... speed density systems suck for engine mods. Now I had heard this before, with no qualification, and assumed it was maybe ford or chev specific, maybe the MAF systems had better computers I thought. But now I understand the reason why it is so. A MAP system always assumes the stock volumetric efficiency for the engine. It has to, all it knows is that at X pressure and Y temperature, the motor takes in Z amount of air, by calculating it from pressure seen at the MAP, and temperature seen at the IAT or charge temperature sensor if equipped. So, since the VE is a constant in the calculation there, if you change it, the SMEC gets it's math wrong. A MAF system however, measures the actual amount of air going in, then figures how much fuel that that will need to stay stoichiometric(ish) it does not rely on the VE in the calculation, therefore get more air in, and the computer can compensate with more fuel.

Basically, if we do mods to the 3.0 to improve breathing, the SMEC has no frigging clue how much air is going in. It just assumes that everything is the same as when it came from the factory. It appears to out manoeuvre attempts to add more fuel to compensate, because it probably marks sensors as skewed or faulty and learns a compensation, because after all, VE doesn't normally change on the fly, why should it need more fuel...

Now, some early 3.0s have a charge temperature sensor mounted to the plenum. Why is this significant? Because it might be the "easy" way to "fix" things. I'm figuring the reason why they didn't bother past the first couple of years, is because at the plenum, by the time the SMEC is in closed loop, the air intake tract has warmed up quite a bit. Therefore the intake temp ends up being fairly constant, or within a smaller range. I figure that they probably started deriving some fudge factor from coolant temp instead, since the plenum is kinda loosely thermally coupled to the engine, this makes some sort of sense. This might explain the arguments over whether insulating plenum spacers actually work. If the car has IAT, it sees the cooler temps and compensates, more fuel and more power. If it doesn't, then it probably just runs leaner.

So, it's looking to me like the best plan is to keep my '88 electronics, then when I put them on the ported '92 project motor , that I should be able to fudge the IAT temperature reading so that the SMEC thinks there's more mass of air in the engine (which there should be). Now, those "eBay power chip" things do something similar. It's just a resistance bias of the IAT thermistor. Here's where people get confused though. Most ECs will tune out the "wrong" temperature signal over time, since their other sensors tell them they're using too much fuel. So it's a temporary boost on an unmodified motor. They may benefit MAP/speed-density system users who have put upgraded intakes and headers on more, since they might end up at a point where they've got more fuelling than stock, 'coz the other sensors are supporting the IAT readings. However, I'm figuring that going too rich through IAT will have the SMEC asking questions and might ignore it completely. Therefore, I think careful tuning of an IAT bias after major engine mods may be a good thing to help realise the power gains, and get the fuel needed for them out of the SMEC.

So basically, if you're getting 10% more air in your motor, you've got to figure out how much colder you should tell the SMEC the intake air is, such that the mass would be the same for the same volume, because it doesn't expect to have more air volume in there.

I dunno what you can do about all this if you have later electronics though, it's guessing the temp, it's expecting the VE to be the same, if you fudge the coolant temp too much it will drop out of closed loop.... You've either got to measure your new V.E. and stick it in a new SMEC/SBEC callibration or switch to a MAF/megasquirt setup.

I guess, for early electronics one might be able to rig something to handle mild boost with a boost biased IAT, dunno how much range there is in the tables though. If it was only intended to handle air down to say -30C, then that's the most extra mass you can pump in before tricking the IAT bottoms out.

Anyhoo, I seem to be satisfied now why just throwing more fuel at it via AFPR, upsized injectors etc doesn't tend to work. So this might be another thing to remember, if porting a 3.0 or adding stroke, bore, lift, duration or boost, either get a custom cal, use megasquirt, or use '88 electronics and fake out the IAT...

However, at some point there will be real fuel supply problems and bigger injectors or more fuel pressure will help... but I'm pretty sure the IAT needs biasing too. Seems like the SMEC is quick to call the O2 sensor a liar, and go more on it's MAP readings. But if an IAT reading supports the O2, then it just figures it's in Anchorage instead of Miami.

RW222

[tkelly27]

Mass air sucks really bad. For nearly every high performance MAF/MAS car community out there, there is a kit to get it over to speed density. I know of no kits to turn a speed density system into a mass air system, unless you work really really hard at megasquirt. Xratti on here is switching from a ford 2.3 computer to a TD computer to make it easier to tune. If you're boosting, there isn't much worse than MAS, trust me. At it's best, it's a crappy guess, at it's worse, you've got a tiny boost leak and a nuked engine.

It sucks for you because you can't dig into the guts of your computer and tweak it. If you confuse your engine into thinking it's colder, you'll be limiting yourself to cold engine spark maps.

It would help if I knew what you were trying to do, are you boosted? If so, there's no amount of CTS or IAT mods you can do to make up for that. Maybe look into an AFC also has the spark issue, but it's a bit more tuneable. Another beneficial idea would be an AFPR so you can add in small amounts without screwing with what the ECU sees.

[RoadWarrior222]

Hmmm, as far as I understood it, it was previously desireable to "revert" to speed density because MAF was poorly understood and the oversized MAFs and other supporting parts either weren't out there or were really expensive. Now most performance guys have "got the hang of" MAF systems they are seen as superior. At least that what I'm getting from recently written stuff. The aftermarket conversion kits for speed density are a lot cheaper than the MAF kits, thus more popular. It probably won't be that way forever. They're mainly intended to go from carb to EFI though I thought.

However, this is for N/A for me and I can see what you're saying about speed density being more reliable for boost. For the 2.2 and 2.5 folks with their SBEC/SMEC callibrations being well understood, I can see no reason why they'd want to use a MAF setup. It works, why change. Also I'm thinking that measuring boost on a MAP sensor is a bit different from measuring vacuum, and the faults of the method in vacuum don't really affect the top end. Since basically a large proportion of the overall VE is coming from the boost, the errors in engine n/a VE are probably small enough that it makes less difference. The 3.0 cals are poorly understood thus far, I'd be more or less on my own with it. While I'd still be interested in digging into those to get a cal that works perfectly, I'd want something that might run halfway decent in the first place, hence trying to figure out what's going on.

I'm not thinking of messing with the CTS at the moment, pretty sure I've got the IAT on mine, but if the spark map gets a little off, there's still the base advance on the dizzy to play with, to find somewhere that's okayish.

I'm looking at building the '92 motor to go in the '88 with 11:1-12:1 compression, porting the heads, with inserts and porting in the exhaust, combustion chamber work and possibly custom pistons. Ported and spaced plenum, ported lower, 52mm TB, "cold air" induction. I'm actually aiming at making power lean, shaping things to avoid detonation, and polishing things to keep the heat out of the heads, so won't want/need too much extra fuel I hope. I want to keep peak torque low and try to make most of my power where I can use it. I'm taking onboard some conventional wisdom, but deliberately ignoring a lot more, 'coz I think I've got some tricks that suit what I want. I'm shooting for 225HP at the crank, well actually, I'm building it like it would be 250 and hoping that I at least get a useable 225. I just wanna do a lot of stuff different, 'coz most N/A are getting hung up below 190, so obviously I'm going to have to do something radical to "bust out". When it's running I'm thinking of going to custom upper intakes and dual throttle bodies, but that's just extra complication, I'll get it running good on the rest first.

Oh... and other targets are, 1/4 in low 14s @ 100mph on street tires and 30mpg. (I know it's a brick, it'll have some aerodynamic help for that)

anyhoo, whether it will all work together or not is something I'll just have to find out.

RW222

[Ondonti]

Quote:

Originally Posted by RoadWarrior222 Hmmm, as far as I understood it, it was previously desireable to "revert" to speed density because MAF was poorly understood and the oversized MAFs and other supporting parts either weren't out there or were really expensive. Now most performance guys have "got the hang of" MAF systems they are seen as superior. At least that what I'm getting from recently written stuff. RW222

And thats a bunch of BS you are reading because someone is trying to sell a product (that coincidentally is more expensive....hmmmmm....hmmmmm...)

Go investigate motec, electromotive, haltech, AEM, etc.

Motecs cost about 6000-8000 dollars and are used by top racing teams.
Find me someone running MAF and Motec and Ill buy the argument.

[RoadWarrior222]

Seems the ford guys like it....
Electronic Fuel Injection Mass Flow vs. Speed Density- Car Craft Magazine
Speed Density to Mass Air Conversion. F150, Bronco, Mustang.

The only argument I'm really finding for speed density is familiarity, like the reason some people still hand write letters is because they can work a pen and paper but find word processing software scary. I'm seeing BS arguments about lack of top end resolution, but duh, use a bigger MAF, and I bet motors that need that, are not streetable with a speed-density setup, crappy high idle and lumpy all over the bottom end, whereas with a correctly sized MAF setup, they could be. Though I guess it must be said that if you're making north of 1000 WHP, then speed-density fits all sizes upwards, whereas sourcing big enough MAF units gets tricky around 600WHP. I guess anyone putting 600+WHP in a FWD platform isn't too worried about whether it's streetable or not.

But anyhoo, the only point in bringing that debate up was to point out that we need new "maps" (Yes I know it's not really a map in the SMEC) to go anywhere with the stock computer and that the V6 code does not seem to be as easily played with as the 4 banger code, so we can try working round it, or use something else that's better documented. Was just contrasting the things you can do with MAF setups to compensate for mods without re-mapping and how it copes with mods better without remapping.

It further occurs to me that the reason MAF kits might be a hard sell and priced high is because generally, you can just pick a car with the same number of cylinders and approx same displacement, get a set of injectors and matched MAF for the power you want to make, steal the computer and harness from the donor and wire it up in yours with a fair chance of it working okay, due to the MAF and injector size matching being about the critical part. Not only that but also, a lot of stock MAF computers have a fair amount known about them, quite a few tools for them, and often don't need modding to be re-flashed, some models of EEC-VI for instance. Therefore, stock systems can be retasked for practically any other engine you're likely to find in a wrecker. It would be when you start building up high power motors that you'd need to spend money, I guess if you know that only your system will do for MAF on a $15,000 engine build, you tend to ask a little more. F1 teams do their own electronics... and have been using MAF.

[Ondonti]

car craft sells things....

[RoadWarrior222]

Well whatever....

All I was trying to say in the first place is that without remapping the SMEC it's the V.E. constant in the map calculations that we're fighting when trying to do anything with the heads, compression or cams for N/A. I was just observing that IAT if you've got it, was about the only place where a fudge factor might make any difference when breathing and V.E. is improved or changed. After all when the O2 starts screaming "LEAN LEAN LEAN LEAN" and stays there after a couple of steps of enrichment, the SMEC probably says "Screw it, the sucker has failed, I'll ignore it now..." Whereas if it's reading lean, AND the IAT is saying, "real cold, heavy air" then the SMEC might figure on trying a few more steps of enrichment, or switching to a "map" for greater air mass, before it gives up.

As far as the fuel is concerned, when you're using enough to hit 85% duty in the injectors before you run out of pedal, then you can probably tune an AFPR and a variable resistor IAT offset together on the dyno or track for more fuel and get the power to stay, since it's V.E. factor is producing the "right" results as seen by the O2 and it's IAT+MAP reading. When it isn't seeing confirmation from more than just the O2, it looks like it dials it back and runs off stock engine fallback values, which are now wrong. You might wonder how it can be set "right" without IAT involvement and then go off and lean it back. Well the thing is, it's not always satisfied with how it's running, so at cruise in vacuum it will start playing around a little to lean the motor back a bit, for fuel economy reasons. That's when things start to go amiss, the changes it thinks it's making don't jibe with what it's seeing at the O2 sensor and it probably starts hunting around, after a bit of this it probably flags the O2 as unreliable and backs everything off. Ergo, no more juicy tune.

Anyhoo, didn't want to just "give up" at sub 200 HP levels like a lot of people do with this motor in chrysler applications (Hmmm I wonder why mitsu 12V applications can make more power with bolts ons, could it be a MAF system and no VE constant to fight? No of course not, what a silly thought, must be the JDM pixie dust.) and went looking for a reason why stuff doesn't deliver power like it should, or why some stuff appears to deliver power for some people and not others.

[c2xejk]

Quote:

Originally Posted by RoadWarrior222 After all when the O2 starts screaming "LEAN LEAN LEAN LEAN" and stays there after a couple of steps of enrichment, the SMEC probably says "Screw it, the sucker has failed, I'll ignore it now..."

Quote:

As far as the fuel is concerned, when you're using enough to hit 85% duty in the injectors before you run out of pedal, then you can probably tune an .

Can you please site your sources. I do not believe your facts are correct.

[Ondonti]

why exactly do you think chrysler 12v's dont make more power with boltons?
i would like to know those sources too.

Robs 3.0 3 speed van stock made 112whp
Ed's 3.0 5 speed van with lots of goodies made 150+whp.
peak HP @ 5000 rpms is not the problem area in the VE curve so you arent going to see more peak hp by adding fuel above peak HP, just increasing power in the RPM's above peak hp.

If you have a setup that makes peak hp higher then 5000 rpms then you have a whole different issue at hand.

When a car runs for crap in the low RPM's thats because of the cam, not because of the fuel system. Big cams are made to have a shorter rpm range for good power but make more hp in that shorter RPM range.
They run for crap at 1000 rpms because there is too much overlap.
MAF cant do anything about that.

[RoadWarrior222]

Sources for what? That O2 sensors fail lean? That SMEC ignores bad sensors?

The 85% thing, I used it clumsily, just realised that unless the SMEC is getting generous with the fuel all of a sudden it won't try to keep up and will be lean all the way to full throttle, which is where it will hit ~85%, not before. Didn't think about that sentence too carefully.

Ummm, not sure you were following me, I didn't say chrysler applications don't make any power, with bolt ons, I said mitsu applications seem to make more power, meaning with similar stuff done. If you missed it, I was previously referring to a ceiling of around 190 crank HP that chrysler applications seem to top out at, so I'm not entirely sure why reminding me that Ed's van made 150+whp is relevant, unless you're implying that he had 60HP worth of transmission loss or something.

[tkelly27]

Those articles are for people who can't tune their systems. MAF is kinda cool if you can't tune.

MAF Translator

Notice the flagship product has speed density in it?

Also, in the DSM circle the tuning solution DSMLink now has a speed density conversion built into it.

I don't know of a single race car running a MAF system. I also know of millions of speed density systems running on the street.

Back to getting more fuel into your engine, there is little known about what actually happens when you do XYZ to the V6 calibrations. I'd get a logger and start experimenting. I'd also decide where you want to go now. You know dicking around with resistors and crap aren't going to get you where you need to be. What was it on the dodge garage, even if you disconnect your CTS from a turbo 2.2/2.5 you can only get an extra 1/2lb of boost. That's not a 10% increase in airflow...

You can get an AFC for $100ish and then actually be able to modify things from inside your car. You can get a MS assembled for $250 (you'll also note that MS, while able to do mass air, comes standard with a MAP sensor installed INSIDE the unit...).

You could also do everyone a huge favor and get a team together and disassemble the V6 code. Then we could have seriously beastly turbo V6's for super cheap.

[Ondonti]

Quote:

Originally Posted by RoadWarrior222 Sources for what? That O2 sensors fail lean? That SMEC ignores bad sensors? The 85% thing, I used it clumsily, just realised that unless the SMEC is getting generous with the fuel all of a sudden it won't try to keep up and will be lean all the way to full throttle, which is where it will hit ~85%, not before. Didn't think about that sentence too carefully. Ummm, not sure you were following me, I didn't say chrysler applications don't make any power, with bolt ons, I said mitsu applications seem to make more power, meaning with similar stuff done. If you missed it, I was previously referring to a ceiling of around 190 crank HP that chrysler applications seem to top out at, so I'm not entirely sure why reminding me that Ed's van made 150+whp is relevant, unless you're implying that he had 60HP worth of transmission loss or something.

Where are you getting this mitsu information.

What chrysler 12v motor has top'd out @ 190 crank hp and which mitsu 12v has not top'ed out @ 190crank hp?

And what type of dyno are you comparing. what kind of corrections.
I dont really think you are looking at any real numbers to make those statements....and the chrysler's have different intake manifolds, different exhaust manifolds, and different cams (in some applications).....AND it is a proven fact that the mitsubishi air meters are the worst MAF style devices in existence. just having it installed hurts power.


Also my point about making peak HP @ 5000 rpms. There is NO VE problem @ 5000 rpms so there is NO peak hp loss because of some VE problems near 6000 rpms.

Here is the proof that speed density vs MAF has nothing to do with the peak hp differences in 12v motors.

I can tune a chrysler 12v motor to run 12.5 a/f on a wideband @ 5000 rpms for at least a short period of time (adjuting fuel pressure).This air/fuel ratio should give me about the maximum power possible. Just because it leans out after that RPM doesnt mean peak HP is changed. Therefore The MAF/SD has no part of this issue because I can get 12.5 a/f @ 5000 rpms regardless of the type of air meter system being used.

There is no such thing as a "better" 12.5 a/f ratio. SD or MAF there is no difference when you are looking at one point it the RPM range where both have the same a/f ratio. We can also change the base timing on the distributor so both chrysler and mitsu engines run the exact same timing advance @ 5000 rpms regardless of the metering system used.

[c2xejk]

Quote:

Originally Posted by RoadWarrior222 Sources for what? That O2 sensors fail lean? That SMEC ignores bad sensors?

If the controller can not get the O2 to sweep even at idle, then there is a problem... The controller can adapt, though there are limitations.

Quote:

The 85% thing, I used it clumsily, just realised that unless the SMEC is getting generous with the fuel all of a sudden it won't try to keep up and will be lean all the way to full throttle, which is where it will hit ~85%, not before. Didn't think about that sentence too carefully.

Are you saying the controller will not go over an 85% duty cycle for the injectors?

Quote:

Ummm, not sure you were following me, I didn't say chrysler applications don't make any power, with bolt ons, I said mitsu applications seem to make more power, meaning with similar stuff done. If you missed it, I was previously referring to a ceiling of around 190 crank HP that chrysler applications seem to top out at, so I'm not entirely sure why reminding me that Ed's van made 150+whp is relevant, unless you're implying that he had 60HP worth of transmission loss or something.

Why would I need 60hp worth of transmissions losses to be over 190 crank hp??? Last time I dyned at 156whp@5200 RPM, they equates to roughly 183 crank hp... Doesn't mean that I can't/won't get over 190 crank hp... I am still making tweaks/improvements.

The setup of the Mitsu engine is physically a little different.

[RoadWarrior222]

Jeebus, I know the damn thing should be good for more, I'm not saying there's a hard limit, it's just that approaching that point it seems everything has less and less effect, stuff that you'd figure would be +20HP on any other motors gets like 3-5 or so. Ed you're even saying that, you're at about 183 crank and you're having to fight for every extra HP tooth and nail now right ? All I know about the mitsus is what I skim off places like 3si, where I'm not a member so don't seem to have a search function to pull it all up again. But SOHC 12V guys there are claiming like 20-25WHP dynoed gain from K&N intake and an exhaust. See figures like 167WHP just from that. The internals are the same. I've seen mention of folks there building 12V SOHCs that "hang with" VR4s on the track, in the high 13s, can't find build lists for them, figure they have to be over 200HP though. Okay so those guys start with a different plenum and bigger TB, but after we've done all we can to those on our motors, which only seems to get us 15HP to their 20 (141ish stock vs 161ish stock, haven't seen anyone claiming much above 150ish with just plenum and TB mods) we still don't seem to get 20WHP with an exhaust and "intake" best figures I've seen anywhere for those on ours seem to be about 15 again. Just saying that the overall impression I get is that we get only about 75% of what they do for same power adders.

I see what you're saying Brent about setting the fuel up for 12:1 on the dyno at 5000, I presume you've seen figures from that? What are those? Of course I know that under a very limited range of conditions you can make the motors behave. I don't beleive there aren't VE problems with that though, it would seem to me that it would be pig rich at idle, and your adjustments would get tuned out pretty damn quick if you left it idling very long.

Anyhoo, if you guys knew of anyone making over 200 crank verifiable and repeatable, on an untouched SMEC/SBEC you'd just point to that, and not have me wasting my time digging out every example in the world of people failing to make it.