Street (vs dyno) tuning -- timing, etc.

[caramall2]

If boost is steady and controlled and AFRs are good (and of course, checking for knock and looking at learning views)...curious on next step (or is this the next step? I realize boost, AFR, and timing are actually all interrelated (e.g. more boost increases flame propagation, may have to retard timing, etc....)...but...).

 

How does one "street tune" for timing? How do you effectively set timing without a dyno to show you what's happening at each range, load, etc.? Advance until knock and then back off? That doesn't seem to guarantee max torque, etc., or does it (and could damage one's engine?)? Can you really use dyno info from a log accurately (doesn't seem that precise)?

 

Also, some say it's better to lean out some and retard timing a little giving the same power but less fuel consumption (not sure what that does to EGT). Is that a step to take as well?

 

Thanks.

[BrandonspecB]

On the street all you really can do is find the knock limit with the knock sensor and then back it off. This isn't ideal. I find road dynos don't give accurate enough results to see if you are 1 degree away from MBT. They work really good to see fairly large changes in power. I advance until I start seeing knock counts then back it off 2 or 3 degrees. Then as ambient temps change I make sure my compensations are adding or pulling enough timing without getting knock.

 

My opinion on rich vs lean vs advanced or retarded timing...leaner is hotter, retarded timing means higher EGTs from starting combustion later in the cycle. People get so caught up with knock, but heat is what kills engines. Detonation is a by product of excessive heat. I prefer rich with advanced timing. Keeps things cooler. This is also another reason I run water/meth injection.

 

I read in a Ford magazine where they interviewed the engineers and calibrators that designed and setup the new 5.0L V8 "Coyote" engine. They added piston oil squirters to keep the pistons cool to prevent detonation. If I remember right, one of the things the engineers wanted to add to the old 4.6L, but couldn't because of the bean counters, was oil squirters. The 4g63 in the Evos and DSMs also has oil squirters. Keeping things cool goes a long way to making power on pump gas while adding durability.

 

So, as far as running leaner to increase mileage...WOT gives poor economy no matter what the AFR. The difference between a 10.5 or 11 won't be noticed at all. The difference between a 10.5 and 12.5 won't be noticed unless you're road racing for extended periods of time. Staying out of the throttle results in the best economy gains.

[BrandonspecB]

Here's the link to the 5.0 article.

 

http://www.mustang50magazine.com/techarticles/m5lp_1003_2011_ford_mustang_gt_50_coyote_engine/viewall.html

 

Here's the quote about piston oil jets. I was wrong about why the engineers didn't do it in the 4.6L

 

"You may also think "weak" when viewing the Coyote's racy-looking but hypereutectic pistons. But there's a twist: oil-cooling jets. A fine mist of oil is squirted continuously from jets in the block's main webs. This oil sprays directly on the underside of the piston, at the vulnerable piston boss and bottom of the crown. The engineers sold the expense of oil jets to management by telling them it speeds engine warm-up (which is true), but the real reason was for piston cooling, hence longevity. This means the lighter, quieter, tighter-fitting, less-expensive hypereutectic piston can be run in this demanding high-rpm, high-load application.

Benefits of the squirters are extensive. Testing shows the crankshaft runs 25 degrees cooler with them, and they help with octane sensitivity. Combined with the heads superior water-jacketing they are one reason the high-compression Coyote can feed on 87-octane gasoline. Interestingly, adding piston-cooling oil jets was one thing engineers on the original Four-Valve modular-the 280hp Lincoln Mark VIII's 4.6-told us they would do if asked to increase performance. That was 17 years ago, so it's been a long wait for this fundamental improvement."

 

Something else to consider is that when you really change how an engine moves air, it can take months to get all of the tune dialed in right. WOT is relatively easy. The rest takes many, many hours of logging under all kinds of conditions. It takes me months of logging and tuning until I'm finally comfortable enough to leave my laptop at home.

[caramall2]

Great info and response! Thanks much.

 

It makes my head spin a little to think about tuning the entire range of cells in addition to WOT. Makes some sense though--I guess you're trying to optimize across the entire range. Are you again setting AFR and just trying to maximize timing for that load & rpm? Doesn't the ECU do some of that for us in CL mode or not really that much?

[BrandonspecB]

I'll share with you some of the methods and theories I use to tune my own engine. I'm not a pro, but I've been doing this for almost 10 years on a few different cars. I've never tuned anybody elses car, so take everything with a grain of salt.

 

First thing I do is dial in everything that affects flame speed. That would be boost, AFR, and cam advance. I do timing last because the right amount of timing depends on how fast the flame travels and the piston speed. All the things I mentioned before affect how fast of a flame you have.

 

AFR: I have a couple theories and reasons for doing what I do here. I'll try to explain them and give credit where it's due to where I learned it.

 

There is an AFR that makes the most amount of torque with gasoline. It isn't 14.7. If I remember correctly, according to Greg Banish in his book Engine Management: Advanced Tuning (which if you really like this stuff I seriously recommend getting it) NA engines make the most torque at 13.x AFR and turbo engines at 12.x AFR. I can't tell you the exact AFR because I left the book at my mom's house.

 

The reason these richer than stoich AFRs make more torque is because there is a better chance of fuel molecules meeting up with oxygen molecules. Leaner AFRs things get pretty spread out. Problem is, on pump gas at high loads you can't run in the mid 12s because of the heat. So, we pour in extra fuel in the range of 10.5-11 AFR just to cool things off and keep the detonation away. In fact, in Greg's book he has a dyno graph comparing a lean and a rich AFR on the same setup. There's hardly any difference in power. But, the richer AFR would be cooler.

 

Now, with that in mind, what I do is leave the car in closed loop until a little over 1.0 load. Closed loop is where the ECU targets a stoichiometric 14.7:1 AFR because this is where you achieve the best balance of economy, power, emissions, and smoothness. You want more power and smoother power delivery? Go richer than 14.7, but economy goes down and emissions go up. Go leaner than 14.7? Power and smoothness decrease, emissions and fuel economy increase. 14.7 gives the best balance and is ideal for just cruising around.

 

At around 1 load though is when the engine starts seeing positive pressure. At this point I like to get out of closed loop and target a slightly richer AFR in open loop mode to make more torque and smooth things out. I don't give a damn about economy. I want torque and smooth power delivery. I shoot for 13s a little over 1 load and get it into the 12s until load really starts to go up. Then I gradually go richer as load goes up until I'm at my WOT target AFR right before peak load occurs. I then carry that AFR out to redline. Pump gas, moderate boost levels I like mid to high 10s. Just depends on the setup. Better intercoolers you can go leaner. Your experience may be different. This is with pump gas only.

 

Once fuel is how I want it and I have a nice smooth transition from closed to open loop and a nice transition that gets gradually richer as load goes up, I'll start tinkering with timing.

 

There are a couple rules of thumb with timing. As load increases timing should decrease. As RPM increases, timing should increase.

 

With that in mind I start dialing in WOT, making changes as I see knock and making sure I have smooth changes from cell to cell. Part throttle timing takes a little longer. I mostly drive around with my laptop open and keep an eye on knock. I then put the car through different part throttle acceleration scenarios and if I see a large amount of knock that is repeatable I'll go in and make changes to those cells and then make sure everything is smoothed out and blended. This takes a long time, but I do it every morning on my way to work and every night on the way home.

 

I'll post up some of my maps so you can see what things look like.

 

You can see how things get gradually richer. In the AFR graph, the red line is the AFR from my wide band and the pink line is the target from my open loop table. They match fairly close. You can see when the meth kicks in. I target almost a 13:1 AFR at high load because of the meth injection. You can see in the graph that my AFRs are in the 11.5-12 range at high load/high RPM. I'm only injecting enough gasoline for around a 13:1, but the methanol richens things up a lot. Meth and water do a lot better job at cooling than gasoline does.

 

Timing drops as load increases and then it comes up as load decrease and RPM increases. I've found that timing usually ends up almost being a mirror image of load on a graph.

 

The way I know my tune is pretty dialed in for a street tune is if it is smooth, pulls hard, and doesn't knock under different conditions. One way you can tell if you have your timing dialed is if you add 5 degrees of timing to the whole map and start seeing pretty consistent 2-3 degrees of knock retard in almost all driving conditions and then if you pull 5 degrees out you notice a decrease in engine response and power. Don't split hairs on a few degrees of timing or .2 AFR. You won't be able to tell. Engines ran well for a long time without near as much technology as we have now. If things are close, it pulls hard, and doesn't knock, consider it done.

 

Hope that helps some. I'm sure if I've informed you wrong someone will e-kick my ass.

[BrandonspecB]

Great info and response! Thanks much.

 

It makes my head spin a little to think about tuning the entire range of cells in addition to WOT. Makes some sense though--I guess you're trying to optimize across the entire range. Are you again setting AFR and just trying to maximize timing for that load & rpm? Doesn't the ECU do some of that for us in CL mode or not really that much?

 

It definitely is overwhelming. Lots to do and change, especially if your car is heavily modded.

 

I find using the Airboy spreadsheet really helps straighten out all the data.

[McDowell Performance Tunin]

This will help with the general concepts of Subaru tuning. It is a bit old, but has a lot of information that applies. It does have a lot to take in, but a good read.

 

Click Ver 1.0

 

 

https://sites.google.com/site/asubienewbietuningguide/

[caramall2]

BKatter, Stellar write up--one of the best I've seen! I think it should be stickied. Thanks much.

 

What do you do for cam tuning (AVCS I assume)? I know there are lots of theories on that as well.

 

Logan, thanks for the link to the guide.

[BrandonspecB]

BKatter, Stellar write up--one of the best I've seen! I think it should be stickied. Thanks much.

 

What do you do for cam tuning (AVCS I assume)? I know there are lots of theories on that as well.

 

Logan, thanks for the link to the guide.

 

Here's some info that I use to set my cam timing.

 

Quote from Greg Banish's book:

 

"Advancing or retarding the camshaft relative to crankshaft angle can shift the tuning characteristics of the engine to allow greater efficiency at higher or lower speeds, respectively.

 

Advancing the camshaft generally gives better low-end and midrange torque by closing the intake valve sooner, trapping the charge in the cylinder before it gets a chance to revert back into the intake port as the piston begins to ascend. Since most systems exhibit relatively little exhaust scavenging effect at lower speeds, opening the exhaust valve earlier also reduces static pumping losses by venting the high pressure more immediately to the exhaust system.

 

Likewise, retarding cam timing allows the intake valve to remain open later where the ram tuning effect can continue to pack air into the cylinder at the end of the intake stroke at higher speeds, making more power. As speed increase and scavenging efficiency increases in the exhaust, larger volumes of gases can be extracted from the cylinder with a later valve closing. This leaves more empty volume for the intake charge to fill on the following cycle."

 

Here's a really good paper on how valve timing affects an engine:

 

http://www.mechadyne-int.com/vva-reference/papers/the-impact-of-variable-valve-actuation-on-engine-performance-and-emissions.pdf

 

So, at low engine speeds, when the air entering the engine doesn't have a lot of inertia, you want to close the intake valve sooner to keep what you have from going back into the intake manifold as the piston starts ascending on the compression stroke. Closing the valve sooner also increases dynamic compression.

 

At high engine speeds you want to keep the intake valve open longer. The air has more inertia and will continue to fill the chamber even as the piston starts ascending on the compression stroke.

 

So, for my car, I did some testing. I unfortunately don't have the graphs anymore, otherwise I'd post them up.

 

I did WOT pulls from 2000 RPMs to redline with cam timing at 40, 35, 30, 25, 20, 15, 10, 5, and 0.

 

Then I graphed out airflow and MAP to see how cam advance affected both. I used the valve timing that gave me the most pressure at low RPMs and the most airflow at high RPMs. Then I blended things and ended up with what I have now.

 

I also left the large 40 degree area from the stock map in place. That's there for EGR.

 

I added a copy of my AVCS map to give you an idea of what I do. Probably not a good idea to copy it though.

[BlazeRex]

BKatter is pretty close on what he is saying. AVCS isn't quite lined up with my experience. Also, on anything over 93 octane fuel, knock cannot be used to tell if there is a 'safe' amount of timing.. You can blow up the motor before it even knocks.

[caramall2]

Thanks all. Another nice write-up.

 

BKatter, what's the AVCS "cruise" vs "non-cruise" (we don't have two tables on the 2005)? "Non-cruise" sounds more like WOT. Can you post a shot of that map?

 

I'm still a little leery of using knock events to show excess timing vs using torque :-0 Probably just a little chicken (and have replaced one engine already due to cracked ringland ;-)

[BrandonspecB]

My cruise and non-cruise maps are all the same.

 

Using torque is the right way to determine if the tune is dialed. It's just more expensive if you don't have easy access to a dyno. Also, being able to hold the engine in one load cell to see the difference in torque from small changes would sure be helpful.

 

If I were independently wealthy I'd for sure have my own dyno

[caramall2]

Yes, wouldn't we all love a dyno in the back yard ;-)

 

1) I'm not sure if this is the stock map, but it's Cobb's old Econ map for 2005. It doesn't have the low load 40 deg cells (and there are a lot less columns too). There's also less advance at higher RPMs (2005 version tapers off a lot quicker as RPM increases) than yours. Maybe comparing apples to oranges (or at least different types of apples ;-) since different model years. Thoughts?

 

2) What happens as you put in too much or too little cam advance? Or, should I say, what's the worst that happens? I'd specifically be curious what advancing it in the low range would do.

 

 

3) Blaze, what was your experience?

AVCS isn't quite lined up with my experience.

[need bb]

Remember that the 07+ have different heads so the timing and AVCS tables are different.

[BlazeRex]

Yes, wouldn't we all love a dyno in the back yard ;-)

 

1) I'm not sure if this is the stock map, but it's Cobb's old Econ map for 2005. It doesn't have the low load 40 deg cells (and there are a lot less columns too). There's also less advance at higher RPMs (2005 version tapers off a lot quicker as RPM increases) than yours. Maybe comparing apples to oranges (or at least different types of apples ;-) since different model years. Thoughts?

 

2) What happens as you put in too much or too little cam advance? Or, should I say, what's the worst that happens? I'd specifically be curious what advancing it in the low range would do.

 

 

3) Blaze, what was your experience?

1) That map looks pretty close to stock IIRC.

 

2) Nothing really, just a decrease in Torque/VE. Its not dangerous.

 

3) I used to put that high amount of cam advance down low until I did more testing. Throttle response suffered and fuel mileage was actually slightly lower. Now I have much lower values down low. Also, I have more cam advance up in higher loads.

[BrandonspecB]

Remember that the 07+ have different heads so the timing and AVCS tables are different.

 

This.

 

Plus turbo size affects VE as well.