Can we have a real thread about speed density?

[twdorris]

Not to take this thread too far off topic but after talking with a homie I found something that shows Omnimap is repackaging stuff.

I thought everyone already knew that.

Big whoop if they had their design carved into the oem plastic.

If you like clean installs using unique products that nobody else has bothered to come out with, then their sensors are very nice. I like the idea and the clean fit. The guys on our site that have used them for their SD installations seem to like them as well. So they seem like a good choice for those that want the benefits and are willing/able to accept the downside (price). They are definitely pricey, no doubt about that.

Thomas Dorris
ECMTuning, Inc.

 

[twdorris]

Now lets say you have gone from a 2.0 with a small turbo, to a 2.3 with a big turbo. The 2.3 could potentially be at 0 (or limited) boost at the same g/rev that the 2.0 is at under boost.

You know, I meant to reply to this because it's actually an awesome point. I keep talking about airflow per rev and a known volume, but I haven't specifically mentioned what happens when the volume changes (as when you swap in a stroker motor).

It's actually a really simply answer. You just tell the ECU the volume isn't 2.0L as it assumes when calculating load. You tell it that it's actually 2.3L instead and everything works out exactly as it should.

In fact, we should stop saying "airflow per rev" because it's actually "airflow per volume" (which is why it has a direct association with pressure). I just tend to assume 2 liters for volume. But running more volume is simply an adjustment to the assumption inside the ECU.

Thomas Dorris
ECMTuning, Inc.

 

[twdorris]

For the ideal SD, fuel should be a part of the VE table all by itself

Yeah, that's a big part of our disagreement for sure. As Hal mentioned way back several posts ago...it's all about what you want from a tuning solution. You can certainly just have the VE table spit out injector pulsewidth if you want. I think that's what the PMS system did way back in the day. Or you can have the VE table spit out real airflow and use a fuel table to tune fuel. For every system I've seen, once you have airflow dialed in, the fuel and timing tables work exactly the same regardless of how the airflow is being metered.

I run the exact SAME EVO8 fuel and timing tables in my own EVO8 with an EVO8 MAF sensor as I do in my Talon running speed density or GM MAF or whatever else. My Talon has HKS 272 cams, modified exhaust, runs e85, FIC 850s, etc., etc. It's pretty different from my bone stock EVO8, yet everything drops in.

I've also sent those exact same tables out to several other people running our system and they all report similar results... Stuff just drops in and works. Sure they can fine tune for their personal preference or fuel type or whatever else they want. That's just part of the fun. But at least when I'm looking at their log files and comparing their load to my load, I know it has a similar meaning regardless of their setup.

Argue g/rev to the grave, it works for oem, that's awesome. Putting SD on top of it, just isn't a great idea.

It's a fantastic idea. And it works beautifully.

Thomas Dorris
ECMTuning, Inc.

 

[juanpac0]

It's a shame that whenever we have these nice SD threads it always ends up being you two arguing over points that, to someone like me mean nothing.

I assume this thread was created in part to educate people who know nothing of SD about it. I know nothing of how SD works, just a basic idea, would it be possible to steer this thread more toward the "this is what SD is and how it works and this is how you can get started". Instead this is somewhat of a Jackal vs. Link debate over things that most people who are trying to get into SD know nothing about. Aren't we trying to demystify SD vs making it seem more complicated?

 

[Hal]

Speed density is simple in concept.

If you know the air pressure and it's temperature, then you can calculate it's mass. That's right, it's weight (or density).

Now that we know the mass, we still need to know how much of that mass made it into the engine. This is the volumetric efficiency. Unless some mechanical component (cams, manifold, etc.) is changed, the v/e for an engine will be constant.

Now if the rpm point for which were are trying to tune has a v/e of 100% then we would add fuel based upon 100% of the calculated air mass. If it's 80%, then we would add fuel for 80% of the calculated air mass.

It doesn't matter what the boost pressure is, because at the 80% point the engine will always take in 80% of the available intake charge. It doesn't matter how much fuel or air is in the charge, the engine will only take in 80% of the charge.

The key to accurate fueling, then is building an accurate v/e map.

Unless you do that, you won't have a solid foundation to start from.

Different engine management systems implement s/d in different ways. They all work, but some offer more flexibility (with a potential increase in complexity).

The key is to find a system that meets your OVERALL needs. The ability to measure the air charge and supply the appropriate amount of fuel is only one piece. In my opinion, the least important piece (because they all do it one way or another), when comparing the different choices.

You have to determine if the system meets your requirements for tuning. If you want to turn 10k rpms, a system that lets you tune only to 7500 may not be the best choice. If you want to run 40 psi, a system that let's you tune only to 30 psi may not be the best choice. Remember, there's more to tuning that just the afr. Ignition timing is another factor. Make sure your solution permits timing adjustments in the same range as it does fuel adjustments.

Now compare the other features (nlts, impacts on emssions compliance, als, etc). Look at the logging capability. Does it record the items you need, how many samples a second?

How easy is it to work with? How good is the support?

Hal

 

[twdorris]

Aren't we trying to demystify SD vs making it seem more complicated?

The high level concept was covered on page one with this link.

SD 101

Once you get past the initial terms and concepts, the details get more complicated. That's more what this thread has become, IMO. And once you start talking details, opinions and varying viewpoints start coming into play. It's just a natural part of discussion I guess. If everyone agreed on everything, there'd be nothing to discuss.

In this particular case, we've gotten off a bit on "load" instead of pure SD discussion, but it's still related. Discussions go off when someone makes one statement that someone else doesn't buy. In my case, it was when Curtis made the following statement.

With having the map sensor right there just before basically the cylinders, it becomes the best indicator of accurate cylinder pressure.

I simply can't accept that. It goes against laws of physics IMO. I've gone over my reasoning above at least a couple different times now, but I don't think he or his buddies agree yet. I'm sure MAP-based load can be made to work just fine. Plenty of systems do it. But I absolutely do not think it's the "best indicator of accurate cylinder pressure".

So the discussion has changed a bit. Now we're talking about load, but it's still airflow and VE and MAP. So it's all still directly related to SD too and the various ways you can get airflow (or injector pulsewidth) out of SD. If it seems complicated, that may just be the nature of the beast.

Thomas Dorris
ECMTuning, Inc.

 

[dsm-onster]

This is gunna sound like a straight out attack on the individuals, but to put it perspectively a lot of people are idiots. I've watched videos of people that have 3065's running on a bone stock 2.0, stock exhaust, stock everything.... @ insane boost pressures, that is idiotic (especially when they expect it to hold together).

Who ever would push a stock 2.0L with a 3065 and stock everything else would do it with link OR jackel or anything that fits their ideal for their delusional mind. Having load based maps doesn't make one stupid . Stupid is as stupid does. Or maybe they were trying to make a "record". That's the trend now. Since us dsmer's and DSM's have been in the world a long time now.

It only takes a second to figure out how a g/revload based SD system works. Factory uses it not becuase it works with factory power, but because it's simple and effective. If AEM is building a solution based on that premise of morons that push their cars without support, then AEM is picking up the slack for the morons that you describe. Where's the learning curve? I know you know what it means to understand HOW the motor get's its flow. As you've said, and nobly, that educating a consumer makes everyone happy and gives the consumer the best shot at his goal.

. . .BTW, don't get me wrong. I've used your solution before. It's great. I'm jsut being realistic here. So don't take this as an assualt or even an intention to offend. Carry on with your work!

 

[tkelly27]

You know, I meant to reply to this because it's actually an awesome point. I keep talking about airflow per rev and a known volume, but I haven't specifically mentioned what happens when the volume changes (as when you swap in a stroker motor).

It's actually a really simply answer. You just tell the ECU the volume isn't 2.0L as it assumes when calculating load. You tell it that it's actually 2.3L instead and everything works out exactly as it should.

In fact, we should stop saying "airflow per rev" because it's actually "airflow per volume" (which is why it has a direct association with pressure). I just tend to assume 2 liters for volume. But running more volume is simply an adjustment to the assumption inside the ECU.

Thomas Dorris
ECMTuning, Inc.

So your system automatically adjusts so that if you were spooling up at 1g/rev you will then be tuning your spool up at 1*(2.3/2.0)=1.15g/rev and the target AFR value there is what you'll get? Same with the timing? If your peak load value was 2.0g/rev, it adjusts this up automatically to 2.3g/rev for you, and keeps the AFR target that was in the 2.0g/rev cell?

That's nice, but again, superfluous. In the map based loading you adjust the load cells by the 2.3/2.0 ratio in the fuel table, and the spark events occur at the same positions because cylinder pressure will remain the same assuming VE is the same in the 2.3 engine due to the ideal gas law. This is essentially what you are emulating by indexing the tables.

If g/rev indicates cylinder pressure better, why does it have to be indexed when you change engine volume?

Two ways to get the same result, one requires software calculations, assumptions and an extra table, the other requires 1 change to one table. Aside from being able to correlate it to a mass air car, I'm failing to see how this is beneficial.

Tuning wise, I don't even use a logger. Partially because ours is just about useless at this point (5 samples a second), and partially because I haven't needed to. My tuning map isn't based on a calculated value generated from the ECU, so I can use a boost gauge, tach, and the O2 gauge to quickly make adjustments without ever loading a log. I don't have to ask if link can do that, I know this isn't possible with g/rev. I would like to know how one gets the VE table nailed down.

Since I can't use the software without buying it, I'm guessing you take a guess at what the max g/rev you'll be getting and set up your load table based on that. Then you take a guess at what g/rev you will spool up at and setup more resolution there (I start spooling at 0psi). Then you would set all those values to a safe value, then you can start taking logged runs to adjust the VE to match it?

I guess you would look at the VE cell you were at and adjust that until you are at the target afr, then you know you are getting the right g/rev, then you can tune the g/rev table?

Again, not seeing the convenience there. Especially when the VE of the engine is changed.

 

[dsm-onster]

In the map based loading you adjust the load cells by the 2.3/2.0 ratio in the fuel table, and the spark events occur at the same positions because cylinder pressure will remain the same assuming VE is the same in the 2.3 engine due to the ideal gas law.

Just asking this? How does a 2.3 motor with the same VE have the same cylinder pressure as a 2.0 with the same VE. What's the point of stroking if the cylinder pressure is the same??????

OR are you talking about total ve vs engine ve again?

Regardless, a faster piston is the main reason for timing changes for a stroker, NOT cylinder pressure. You make timing match the ideal cylinder pressure. It's timing. Flame front speed is just a minor part of the deal when stroking a motor.

 

[tkelly27]

P*V=nRT. At the same inlet pressure and temp with the same VE the only thing changed is V. The point of stroking is to increase the V in the ideal gas law equation such that a lower P can give you the same n, or the same p can give you more n, which is what you really want. Right?

You are right, the VE should change based on piston speed, but in both systems the VE will need to be changed to account for that, same with the spark. When you index the g/rev load tables you are keeping them at the same inlet pressure, just like the fuel tables. The difference would be, when you increase the VE at a given boost, you are now at a different g/rev load level.

 

[dsm-onster]

You are right, the VE should change based on piston speed, but in both systems the VE will need to be changed to account for that, same with the spark. When you index the g/rev load tables you are keeping them at the same inlet pressure, just like the fuel tables. The difference would be, when you increase the VE at a given boost, you are now at a different g/rev load level.

Ideal gas law is what I sleep with every night

So in both systems VE will need to be changed. What's the difference then? I'm not being sarcastic. What IS the advantage of boost as load? I hope you're not suggesting that you can go on to infinite g/rev with the same boost. Because ds-map's target market is not going to have that option. At some point they are going to have to up the boost for bigger power.

I still maintain that you have to alter timing tables for a stroker motor even if you're using boost as "load". That is the point of my post. It's about timing to make cylinder pressure not ve tables and load references. So there's really no advantage as I see it. I'm wanting to be educated: please do that. Though I have just bought V3, I'm definately not arguing for one side or another.

 

[tkelly27]

I'm learning a lot too : D

If there is no g/rev limit in link, then both systems would handle as much power as you could throw at it. I was under the impression that the bit value of airflow limited this to a maximum. It has in the past, but they say it is not the case now. I can only take their word for it.

I only use engine VE, so this would be the same at any boost level.

Using made up numbers, lets say you go from 50% VE to 100%VE, your airflow should double at a given boost pressure. Lets say you were at the 1g/rev loading condition and you go up to the 2g/rev condition.

Assuming you want to keep the same fuel ratio, when you change a map only system, you will change the cells in that table 2x and then you will have the same fuel ratio. With g/rev, you will have to modify the VE table so that it is now reading the 2g/rev value, then you tune the 2g/rev cell to the target you want.

Maybe I mistook what they were saying, but it seemed to me that the claim to fame for g/rev is that all engines want the same fuel and spark at equal g/rev, which isn't true.

I'm not saying g/rev loading doesn't work, it's just clumsy to switch between the three tables with different loading conditions. It's also a very capable system, I hope you enjoy it.

 

[Hal]

SD has the biggest side effect benefit of removing the maf from the intake system and freeing up potential power, but the most beneficial is how unbelievably ROCK SOLID it is all the time on the vehicle it was tuned for.

Yeah, until your MAP sensor fails, or to a lesser effect, the IAT fails.

Oh and how about what happens to your turbo when you have a boost leak that S/D works well in spite of. How long before you over spin it to the point of damage, all because you didn't realize you had a leak.

S/D isn't the magical cure you make it out to be. All choices have good points and bad.

You don't like MAF, great... now that 'link has S/D you don't like the way they went about implementing it... great. So much for "free thinking". If it doesn't conform to your idea of how it should work, it's wrong. Either that or you just plain don't care for 'link, regardless of the discussion.

 

[dsm-onster]

If I take my g/rev tuned car, and disconnect any pipe past the maf, the system is totally crippled; and don't act like it ever hasn't happened. Secondly if you a: boost leak, or vacuum leak your g/rev tuned application is either screwed up now, or screwed up after you fix the problem.

SD has the biggest side effect benefit of removing the maf from the intake system and freeing up potential power, but the most beneficial is how unbelievably ROCK SOLID it is all the time on the vehicle it was tuned for.

Wahuh?

I was talking about speed density. Maybe I'm behind what you all are saying but I thought that Tom was building/built a speed density system with grams/rev vs. rpm as the table and then after that put in the fuel table and timing table. I am lousy at understanding the logic of computers because I am not logical. But What differnece does it make to use Toms SPEED DENSITY platform over yours?

. . .Gread discussion BTW. I'm thinking I'm getting it. But maybe not ???

 

[dsm-onster]

Maybe I mistook what they were saying, but it seemed to me that the claim to fame for g/rev is that all engines want the same fuel and spark at equal g/rev, which isn't true.

I'm not saying g/rev loading doesn't work, it's just clumsy to switch between the three tables with different loading conditions. It's also a very capable system, I hope you enjoy it.

No. I was just saying that if you use jackel or link and then you change to a 2.3l motor, no matter what you have to go in and edit the timing tables for your different piston speed so that you can get best power. Period. Still have to go through tables and do out the the tune FOR YOUR SETUP. No matter what solution you use. It doesn't matter how you control the fuel flow. That's all I was saying. I hope everyone enjoys ever dime they put into their dsm. . . (is THAT wishful thinking or what guys?)

 

[dsm-onster]

Oh and how about what happens to your turbo when you have a boost leak that S/D works well in spite of. How long before you over spin it to the point of damage, all because you didn't realize you had a leak.

Yes, having a failsafe like Boostest really helps. If things don't look right, then you can correct it.

 

[Hal]

Yes, you have to change the underlying v/e table to match the new v/e and you may need to adjust timing because of the new physical properties of the engine.

This where you begin to move from S/D as an air metering system to S/D as a tuning methodology.

 

[turboglenn]

It's a shame that whenever we have these nice SD threads it always ends up being you two arguing over points that, to someone like me mean nothing.

I assume this thread was created in part to educate people who know nothing of SD about it. I know nothing of how SD works, just a basic idea, would it be possible to steer this thread more toward the "this is what SD is and how it works and this is how you can get started". Instead this is somewhat of a Jackal vs. Link debate over things that most people who are trying to get into SD know nothing about. Aren't we trying to demystify SD vs making it seem more complicated?

exactly why i have not posted but 2 time in this entire thread... sucks, real discussion would be nice

 

[Rob10_99]

everyone around here talks in pressures, ya mang @ 15 psi its knocking but by 20 its cool. I run 10psi, blah blah blah. really super lame.

/superkewltrollvoiceoff

Yeah exactly, 20psi does not equal 20psi from any other car.
Wholy hell this jackal/link thing is out of hand, I see a showdown coming!

Seriously, lets hear some idea's as to how-to showdown the differences/pro/con's and benefits of both/all systems of SD.


EVEN A TECH ARTICLE OR FOUR IF YOUR UP TO IT!!!!!¿ ffs.

 

[LandoAWD]

I'm so sick of this crap. I know you have been spoken to about how you behave on here, and you just don't seem to absorb this.

Feel free to edit the childish, personal crap out. It would be nice to have one discussion that isn't ruined by dick measuring.

 

[habitatguy187]

I have no idea what you guys are talking about, nor do I care, what I want to hear about is people who run speed density personally, what they think about it, how it compares to their old mas system, and how easy it is to tune/reliable.

 

[Hal]

I've run MAF (OEM, GM-MAF, MAFT, PMS and AFC) over the years and I've also run speed density via ECMLink.

I beta tested ECMLink's speed density back to back against a 3.5" GM MAF.

Let's address each of your requests.

Reliability:

Both systems are reliable up until something fails. Let's take the most common failure, the blown I/C pipe and then assume that it can't be fixed. Both systems will still get us home. If it's a MAF based car, unplug the MAF and drive home in "limp" mode. If it's S/D you can drive home and simply keep your foot out of the throttle so that you don't risk over spinning the turbo charger.

Comparison to MAF:

Generally speaking, if some one had a poorly operating MAF system then they really aren't in a position to provide a valid comparison after they switched to S/D. I've been running a GM MAF based system ever since the MAFT came out. I even beta tested some of the versions. I've always been able to make it work to my satisfaction so I can provide a valid basis for comparison.

I beta tested S/D for ECMLink, and like wise I was able to make it operate to my satisfaction.

Repeated drive cycles over the same routes, back to back, under the same conditions. Looking at the MAF based data and the S/D data. They both got the job done.

Easy to tune:

This is where the rubber hits the road. It can be as hard or as easy as you want to make it. Different systems have different methodologies but they all have one thing in common. You must build a base table that represents the V/E of your setup and as your setup changes you must adjust this table.

Some systems offer various features that build upon the base table in order to make the overall tuning experience more user friendly.

How easy something is to use is pretty subjective. Speaking from an ECMLink perspective, once the v/e table is built, you don't even know you're tuning a s/d system.

I prefer a s/d solution built around the oem ecu functionality (closed loop, obd2, etc) for my 2g.

Why? Because if I build the v/e table properly and my fuel delivery system is working properly my car will pass emissions, idle and drive like stock, start when it's cold out side, and still make good power when I put the pedal to the floor.

The ECMLink implementation of S/D is about as close to plug and play as it gets. I've put it in both 1g and 2g cars and it's worked flawlessly every time.

Hal

 

[the_mork]

What really helped me understand SD was a post from l2r99gst over on evolutionm in a discussion about throttle plate logic vs MAP logic.

For those who haven't had it click yet I'll try to explain it the way that he did because I can't find his post at this point. A lot of this has been covered already but it never hurts to read it twice.

At least in the evoSD there are two tuning tables that handle SD tuning. There is the RPM VE calibration table and a table for calibrating MAP vs Load Ve. I had a hard time finding a good image of the SD RPM VE table so you'll have to bear with me, it is the lower table in this image that has RPM as the top row.

You must be logged in to view this image or video.

This table has the very basic and express purpose of telling the computer how efficient your motor is at any given RPM range. Near the middle of the range the motor is generally more efficient than at either end and it doesn't matter if the motor is actually 100% efficient. 100% efficient only references the peak efficiency of your motor. If our cars did not have throttle bodies then this would be the only table that would need to be tuned but since we generally like the idea of being able to adjust the power output of our motors we need to add a system to inform the computer when the motor has a change in VE due to intake restriction such as a partially closed throttle body. For this we need a second table. For a while there was a discussion whether this was easier with throttle plate logic or MAP logic. Map logic was chosen and this table was created.

You must be logged in to view this image or video.

Please ignore that the table is duplicated in this image there is only one table. Basically what we find is that at 2500rpm the engine may be 98% efficient in relation to its intake charge. This is the mVE (motor volumetric efficiency), but despite the motor being generally equally efficient at 2500rpm no matter where the throttle plate is the total volumetric efficiency of the motor is significantly different because of the intake restriction. So we are cruising along the highway at 65mph and the throttle plate is held open only a few percent. Our intake charge is likely around 50kpa (vacuum conditions), but now we want to pass someone so we open the throttle plate to about 60% and suddenly our intake charge is around 120kpa with no change in RPM or mVE and over a very short period of time. What has changed is our tVE. As you can see in this example MAP is directly indexed to load but it is not necessarily perfectly linear. For an NA car this table could end at just slightly more than atmospheric pressure

So when we put it all together it works like this. The computer checks for motor RPM and determines what the mVE is for that range the next step is to check the MAP reading at that moment which determines the tVE after MAP is converted to a load number. tVE is multiplied by mVE as a percentage to provide a raw load number which is fed into the ram address for raw load. After this point it is up to the stock code to compensate the load for temperature. accel enrichment, cold start enrichment, lean spool enleaning, etc. Of course this process is radically simplified in my explanation. Here is a short worked simplified example.

The motor is at 1900rpm which might be 96% efficient so we check the map vs load table and the MAP is at 70KPa or 52.5 load. So we take the load and multiply it by .96 which puts our load at right around 50 and we reference fuel and timing tables normally from here on out. Now we open the throttle plate and our MAP jumps to 99KPa or 90 load, and we again multiply that by .96.

The other systems use different tables and different methods but it all comes down to determining tVE and then multiplying it by mVE in order to find your place on fuel and timing maps.

This probably isn't clear at all so ask some more questions.

 

[the_mork]

This warrants a completely separate post because I don't want it mixed in with the real SD discussion but hak you seem to be expecting a very large swing in boost control targeted by load or g/rev when in truth that is not the case at all. I don't know if you realized this but load targeted boost control is what Mitsubishi uses as factory boost control and it is the reason why you get so many answers when people ask what is stock boost. The truth is there isn't a stock boost in PSI for DSMs, Evos, or probably any Mitsubishi vehicle. Honestly boost fluctuates much more for a traditionally controlled turbo (MBC) you just don't see it because we use gauge pressure as our measure most of the time and not absolute pressure.

 

[Hal]

I think load based boost control would add a tremendous amount of flexibility.

Couple it with the right turbo selection (VNT anyone?) and you can actually tune your torque curve.

Just like the late model diesel pickup trucks

Variable boost pressure is one of the things not widely discussed.

Most people tune to a fixed pressure, but with controllers available today (such as TurboSmart) it's not a requirement.

So, why not adjust the boost curve to compliment the torque curve and traction requirements.