Supra Forums banner

1 - 20 of 133 Posts

·
Registered
Joined
·
1,192 Posts
Discussion Starter · #1 · (Edited)
Ok... this is the dyno tuning section, but we rarely discuss the mechanics of tuning. All we really ever do is pretty much post dyno sheets and list what the car mods are, but nobody ever posts a fuel or timing map as a means of educating or learning from others. Well, I'm hoping this thread can help change that.

I'd like to start a discussion that informs and encourages contributions from those that have experience tuning. I for one, have learned a lot over the years, and much of what I know comes from the body of work that's out there. However, I've also learned a lot on my own. Consequently, I can't take full credit for what skills I have. Additionally, I hope to learn from you, too - the members of SupraForums. There are many of you I respect and look to for advice, and now's your chance to educate others.

In the following posts, I will lay my cards on the table and try my best to describe what is the culmination of about 6 years of experience with the AEM. Unfortunately, I have only ever tuned my car; so I hope others will offer suggestions on how I can get better. After all, that's the benefit for me. Additionally, I'm by no means a professional tuner; however, I feel confident enough with the following information that I believe others will learn something, And that's what this thread is about - education. If you have positive, constructive feedback - then by all means contribute! I'm hoping we can all learn from one another. So, lets get started. First, I'll start with the fuel map.

The most important aspect of fueling using the AEM is utilizing the Boost Fuel Correct Table. This table enables easy tuning of boost, IMO. Basically, it allows tuning of boost at lower boost levels that translate into accurate fueling at higher boost with liitle or no work (based on motor VE). The table provides proportionally more fuel as a linear function of load. Consequently, as boost goes up, fuel is added proportionally. If fuel is correct at say 12 psi, then it's good all the way to the extents of your MAP sensor. The image below is my Boost Fuel Correct Table.



Notice that it is simply a linear plot of fuel as a function of load. As load increases, fuel is added proportionally. So, what does that do to the Fuel Map?.... Well, it means that fuel is constant as load increases at constant RPM. Therefore, moving up the fuel map vertically at constant RPM results in exactly the same value in the base fuel map. The real benefit is making pulls on low boost on the street to tune the map. Once it's set, fuel at high boost is academic; little or no fuel adjustments are necessary.

The figure below is my fuel map; however, I direct your attention to the portion of the map from 0 psi to 60 psi; that is the boosted section of the map. For now, disregard the vacuum section; I'll address that later.



As you can see, fuel values in the cells running bottom to top vertically in the map are constant above 0 psi. This type of tuning has been called "boost comp" - meaning fuel is compensated by boost through the Boost Fuel Correct Table.

While still focusing on the boosted section of the map, you'll notice that fuel changes as a function of RPM for constant boost; those are the cells that run across the map from left to right horizontally. The fuel "hump" at 6100 RPMs is where peak torque or maximum load occurs on the motor. Consequently, more fuel is required when delivering more torque. Of course, injector duty is increasing as RPM goes up resulting in more fuel delivery; however, more fuel is proportionally required based on the torque the motor makes. Therefore, a motor that has good VE should have a fuel map that mimics the torque curve if boost comp is used for fueling. In my case, that's exactly what's going on here. The motor is working efficiently based on the flow characteristics of the exhaust manifold, IC and piping, intake plenum, and exhaust.

Below is an AEM log of a recent pull made on the dyno at high boost. You'll notice that the fuel is almost perfectly flat through all the critical phases of fueling. FWIW, this was an 843 whp pull, all turbo, on a stock GE head with 272 cams.



I should note that O2 feedback is applied; however, I'm only using a +- 3.5% correction for the entire fuel map. I've included the O2 feedback parameter in the log file. That's very low based on other maps I've seen. Most fuel maps use anywhere from +- 10 to 15 percent correction over the base fuel map. Consequently, the fuel map above stands on it's own without relying on the O2 sensor to make up for inconsistencies in fueling off the base map. Additionally, if the O2 sensor fails, I can easily make it home without it. ;)

Now, some might argue that a narrow fuel correction from O2 feedback is not a good idea because you're more likely to encounter an issue for minor fluctuations in fuel. However, I've learned that problems with the fuel system will manifest themselves well before making boost allowing fast diagnosis of a fueling problem. It's happened to me twice, and both times I noticed problems with fuel delivery before ever making a pound of boost. It's better to go lean in the vacuum section of the map than in the boosted section any day, IMO.:agreed:

Now, let's take a look at fuel under vacuum. Interestingly, I treat the vacuum section of the map completely different than the boosted section. Under boost, my target AFRs are 11:1; however, that's not what I want for the vacuum section of the map. While crusing, I want to run stoich or 14.7:1. At idle with big cams, fuel needs to be ~12.5:1 to maintain steady idle performance. Unfortunately, the Boost Fuel Correct Table is primarily used for constant fueling unless corrections are made in the base Fuel Map. By design, that's exactly how I tune the vacuum section of the map. From 0 psi to -14.7 psi I tune cell-by-cell; this has been referred to as "hybrid boost comp". I haven't found any other combination of fuel corrections to get the variability in fueling required for idle and cruise.

Fortunately, the Boost Fuel Correct Table helps tuning idle because you want constant fuel to maintain constant idle inspite of changes in load or RPM. That's why I created a fuel "Idle plateau". The Boost Fuel Correct Table guarantees proportional fuel as a function of load; fortunately, that works under vacuum, too. From -14.7 psi to -3 psi and from 0 to 1650 RPM, fueling is maintained at a constant 12.5:1 AFR. Having constant fuel is absolutely critical for maintaining steady idle. Otherwise, you'll get oscillations in idle or "idle hunt" that I've seen on so many cars. Idle hunt is cause by a number of things, but they all come down to fuel and timing corrections that are occuring at idle. My philosophy is to minimize fuel and timing corrections at idle - fewer is better. In fact, I turn off O2 feedback below 1500 RPMs to ensure it doesn't contribute to idle hunt.

The other feature in the vacuum section of the map is the cruise "bowl" located next to idle. The difficulty in maintaining a rich idle and stoich cruise is that both are separated by just a few hundred RPMs. Unfortunately, the Throttle Inject Correct Table isn't sensative enough to fueling differences separated by a couple of hundred RPMs. However, tuning cell-by-cell makes obtaining proper fueling easy. However, it does require setting up the resolution in the map where it's needed to make it work.

The figure below shows a plot of the RPM Breakpoints to alter map resolution. This table allows the tuner to establish resolution in the map where it's required the most - namely at places where the map changes rapidly. In my case, I use it around cruise and peak torque.



In the case of idle and cruise, I needed a lot of resolution over a few hundred RPMs to get a rich idle and a stoich cruise. The table above performs exactly that task. By adding resolution to the right of idle, I have plenty of cells to tune fuel accurately for cruise. Now, it's easy to get great gas mileage and consistent idle. I should also note that the same technique is applied around peak torque. However, it isn't necessarily for fueling; I added resolution instead for timing. Specifically, to pull timing gracefully around peak torque. I'll talk about that in another post.

I'll make one more comment about putting resolution in the map where it's needed. For example, I also use the Load Breakpoint Table to skew resolution where it's needed based on boost. It's shown below.



I'm running an AEM 5bar MAP sensor. It measures boost from -14.7 to 60 psi. Unfortunately, my little 'ole 74mm turbo only works efficiently below 40psi. You can see the "Load resolution" in the base Fuel Map above. Consequently, anything above 40 psi is just an overboost condition that I'll (hopefully) never see. More likely, I'll run safely up to about 37 psi; however, I'd like a good cushion in the map from 37 to 41.5 psi. After 41.5 lbs of boost, it's a "don't care" - it should never happen. That's why I set up the Load Breakpoint Table for the large gap between 41 and 60 psi. To coin a phrase from the movie "Contact", it's "just a waste of space". :lol:

That's all I have for now. When I have time, I'll discuss timing using one map for both race and pump gas; and after that, I'll talk about tuning idle.

I look forward to an open discussion with constructive feedback and comments. If you've got a better mousetrap, then let's hear it! I'm also interested in learning new techniques. Even though I've worked with the AEM for a number of years, I admittedly only know a fraction of its capabilities. Consequently, I encourage open exchange of ideas and tuning knowledge.:wavey:

Thanks,
-scott
 

·
Registered
Joined
·
139 Posts
Nice post, I have been checking this section for a while to see something like this and finally it's here, hopefully you keep going with it.
 

·
Registered
Joined
·
6,785 Posts
Subscribed, hopefully someone will chime in and explain more on how timing values are set(obviously not at peak power) but more importantly(for me) lower in the curve. Such as cruising, etc..
 

·
Registered
Joined
·
165 Posts
I also found my fuel map mirrored my torque curve. On my setup, BL61 aka Precision T61, I make peak torque at approximately 4800 RPM and my fuel map indicates at that engine speed it's the highest pulsewidth.



I'm curious to hear what your found on tuning timing. I'm still on my GE VVTi block and to be safe I'm running very conservative timing since the VVTi internals aren't too forgiving. I'm also on a 2.5 mm headgasket which change the squish of the motor. If I was to add a degree or two of timing, my knock sensors would pick up about 50 ms of knock. Without the additional timing, the knock sensors would occasional see about 20-30 ms of knock. This is occuring at very high engine speeds (5000+ RPMs). So the big question is this knock or engine noise? To be safe, I would rather run about 2 degrees less timing and be down about 15 rwhp than walk on the border of a safe tune. When do you stop relying on your knock sensors?
 

·
Registered
Joined
·
2,189 Posts
Great effort, Scott!

I'll say one thing about your load breakpoints. You don't have to put that last load point (16) at 60psi. You can just put it at 45 and your graph won't have that big block at the top. When you view your fuel or ignition tables they'll just go up to 45 instead of 60. It really makes no difference, but that's an option you have. I found this out when I did the MAP sensor wizard and it automatically reset my breakpoints!
 

·
The "Tactileneck"
Joined
·
1,168 Posts
Great post!!! I'm in with slither though, i would definately be interested to see what your input is on timing and how you would go about setting those values up.
 

·
Papichulin
Joined
·
305 Posts
I have never been MORE subscribed to a thread in my life... Fucking A :rockon: :cool: !
 

·
Banned
Joined
·
1,984 Posts
Hybrid boost comp!

I haven't heard those words in some time.
Scott you've summarized some good stuff here, and hopefully we'll all learn a thing or too.
 

·
Registered
Joined
·
1,192 Posts
Discussion Starter · #17 ·
Cool... a sticky! Looks like we have a good response; thanks for the enthusiastic "subscribes".

I'll try to address questions I can answer, but like I said, there's so much I don't know. Consequently, I hope we'll get contributions from others on the preferred way of tuning. First, however, I'd like to cover the basics. Fuel was first; now, I'll look at timing.

The timing map is actually pretty straight forward. It's just a matter of setting ignition as a function of RPM and load. The issues that make it hard, however, are ramping timing into cruise, transitioning gracefully through peak torque as boost levels off, and creating a map that suffices for race gas and pump gas. Before discussing each of those areas, lets first take a look at the map.



In general, timing benefits from RPM and Load Breakpoints. As mentioned in my first post, breakpoints allow you to add resolution where the map is most dynamic. That makes it easier to make fine adjustments where the trajectory of the map is changing rapidly. It's no different for timing. I'll get into the use of breakpoints for timing, but let's start from the lower left of the map at low RPM and load and work our way to the upper right.

In the lower left of the map is where idle resides. Basically, we're talking high vacuum and low RPM. What should look familiar is a "flat spot" for idle. Unlike fuel, where a "plateau" was required, we now have a "floor". However, the same principle applies to maintaining steady, consistent idle without oscillations - namely constant timing inspite of variations in load or RPMs. That's why there's a flat spot in the timing map. In my case, I set timing at about 15 degrees for the same cells called out in the fuel map discussion for idle. Maintaining an oscillation free idle is pretty difficult; consequently, I'll leave that for a later post.

Just as in the fuel case where a fat idle resides next to a leaner cruise, we have very little timing for idle sitting next to alot of timing for cruise. Fortunately, the RPM breakpoints help here; however, the increase in timing over RPMs is mostly linear; so, the changes in timing aren't nearly as complex as the "bowl" that was created for cruise in the fuel map. The only real decision here is how much timing is right for cruise?... Well, that's a good question, and I'm not sure I can answer it. I typically run between 29 and 36 degrees of timing on the highway, and the timing map peaks at 46 degrees of timing above 4650 RPMs. However, I've never cruised at that speed; so, I can't tell you if that's correct or not. All I'm really concerned about is getting off the idle floor as quickly as possible once the car is underway. After all, there isn't much load on the motor in vacuum while driving around; so you can add a lot of timing quickly. That's why there's a cruise "ramp" in the timing map.

Now, let's get to the good stuff... making power! As mentioned above, RPM breakpoints were used to add cell resolution around peak torque. In this case, it really helps for setting timing around peak torque. Why?... well, that's where the engine is under its greatest load, and you can use the resolution around peak torque to pull timing as needed without making significant changes to timing. I've learned that very subtle changes in timing while making power will absolutely manifest itself as inconsistencies in torque and horsepower. Consequently, changes in timing must be smooth and gradual. For example, the dyno sheet below shows where a bump in torque and hp occured (which coincidentally resulted in max power!). Interestingly, this was caused by an inadvertent increase in timing.



After spending some time on the dyno, I finally got the timing smooth through peak torque. The hard part is traversing the timing map making subtle adjustments to ensure timing is consistent. Additionally, I learned that pulling timing around peak torque must occur gradually well before and after peak torque. Once peak torque is past, timing can be increased gradually to flatten out the torque curve. The log file below shows just how subtle timing should be to produce a classic torque curve.



Believe it or not, it took a good hour or so traversing the timing trajectory on the timing map using the log file as a guide to determine how to get a smooth transition through peak torque as shown in the log above.

Lastly, I'd like to address the use of one map for both race gas and pump gas. I'm a firm believer it can be done, and it's not too hard. Here's my timing map again for reference.



In the upper right part of the map, I've labled a timing "mound" and a timing "valley". They are for race gas and pump gas, respectively. For pump gas, I run about 19psi of boost; on race gas, I'm anywhere between 25 and 35 lbs of boost. Consequently, I pull timing around 19 psi with about a 3 psi buffer that ends up looking like a "valley" in the timing map. Anything above that is fair game for race gas. Therefore, timing for race gas looks like a "mound" because timing is being added. It's just that simple. I don't see any need for two separate maps. If you're on the hairy edge of power, then ok... maybe two maps are required. However, I don't suggest ever being on the hairy edge especially with pump gas; it's just not worth it, IMO.

Well, that's all I have in this installment. I'm not sure when I'll post up next, but at least the basics are out there for you to read over. Again, I'm all for constructive comments and suggestions for making this a worth while sticky. So, contribute your experiences! There are plenty of you that tune your own car and can add value to this thread. Let's all learn from one another!

Thanks,
-scott
 

·
Premium Member
Joined
·
2,887 Posts
Great writeup Scott, probably one of the best I have seen laid out on the subject.

Couple of points I will add. First, the only real good way to tune your timing at part throttle is a good load bearing dyno that features real time torque readout. You can then tune for maximum torque at a given load under lighter loads. I have found there is a window of at least 5 degrees (sometimes 10 on certain cars) where adding more timing really doesn't gain anything for torque at light throttle. For cars that I tune were I do not have access to such a dyno (or the owner doesn't want to pay for the dyno time, I simply use a timing map I know is close for part throttle, and focus on the WOT stuff). Your description of timing for WOT is great.

Second thing to not overlook in the tuning process, is injection timing. It is tuned somewhat like ignition timing, on a load bearing dyno (and is most critical at light loads). You can gain both fuel efficiency and torque at the wheels by optimizing this.

Sharing the same timing map for both pump gas and race gas is something I caution against. The AEM only has one boost based fuel cut, so if you are doing both fuels with one map (I run into the same problem when doing a pump/race map switch), you have to set the boost cut above the level of high (race) boost. If you were to get a boost spike on pump, you are going to spike straight into race gas timing numbers and the effect will not be good if you don't catch it fast. When I do a pump/race switch, I simply set the main timing map numbers VERY conservative, that way if someone gets a boost spike on pump gas, it will be able to survive it. For your own car it may work just fine, that I am not taking away from you. But for others who may not pay as close of attention to their car, the "what ifs" become all the more important.

Lastly, while the car will certainly idle smooth at 12.5:1, it is very feasible to get them idling around 14.7:1 (except some of the most radical 1600cc, cam'd cars and even then you can rough it in the ballpark) and this will help conserve fuel, and reduce plug fouling/fuel smell.

Again, great writeup and thank you for your efforts!
 
1 - 20 of 133 Posts
Top