A guy I know that builds Supras was telling me that when you tune your air/fuel misture on the dyno (dynojet), the car will run leaner on the street. This is understandable since the load on the is more dependant on vehicle weight on the street versus the variable load on the rollers of the dyno. Does this make sense? What is everyone's thought about this?

 

i think that there are a ton of different factors that the dyno cannon simulate that you will se on the street, so your tuning could be a little bit off.. but nothing major that i know of

 

I would like more info also.....but then again the car is not magically weightless when you put it on the dyno so...ANY resitance against the wheels would be more than that if you were driving on the street....i assume thats what the correction id for...

 

it is my understanding that in order to obtain street results, one would use a load bearing dyno for tuining. I don't think that a normal roller dyno applies any type of resistance that causes the motor to work as if it were propelling you on the street. The AF ratio you would get at the roller dyno would be close to street, but then again you've still got humidity, temp, etc. to throw these things off on a street application every day.

 

Generally we find that cars _do_ tend to run leaner in real world conditions. This can be for a variety of reasons including:

- Air flow changes - even at 60 mph you're generating far more airflow than pretty much any dyno fan setup can create and what really counts here is air _pressure_ which is very hard to duplicate on a static dyno. The better positioned/designed your intake, the leaner you'll run on the street/track vs. the dyno.

- Intercooler efficiency - turbo cars with intercoolers are even more susceptible to dyno vs. road variations because of the air flow changes above. Intercoolers demand air pressure to create cooling flow through the fins and you just can't simulate much above 40-50 mph (with some really awesome fans) on the dyno.

- Load time - as mentioned, engines behave differently with different load times. Ideally you want to simulate the worst case conditions an engine will see on the track/road. This can only be done if the dyno is load controlled, not inertial only. In the case of all motor drag cars, we make very short pulls that duplicate the engine loading seen in 4th gear at the top of the track. With mapping for multiple gears we'd make pulls of varying loads and optimize each gear.

There are a variety of other reasons as well. You can generally expect on road A/F maps to read at least a tenth or two leaner on most cars. Some will vary more, some less. That's why unless you're tuning a race car to the limits, its always good to back off a half point on the mixture. The extra 1-2% power you gain with the leaner mixture (if you even gain that much) generally isn't worth the risk to your expensive parts under the hood.

C_A_T

 

So to be on the safe side, you can just add a bit of fuel too all rpm range after you get your ideal A/F ratio on the dyno, and it should be good for street?

 

That's not a bad methodology. That's essentially what we do, but not quite in that manner. We know where we want to run the motor in terms of max A/F ratio and we simply tune the fuel curve to be a little richer than that "ideal" ratio. Sometimes a customer will ask me to show them what it would do leaner, perhaps to simply have "dyno bragging rights".

The nice thing is, we try and get the fuel curve very, very flat. Within +/- 0.1 AFR is possible (and almost always within +/- 0.2). So if someone wants to run leaner, they can pull a few psi out of the FPR and do so without worrying about one section of the curve going leaner than desired, but then I don't have to take responsibility for the consequences.

C_A_T

 

Thats a good explanation. So would it be better to run more fuel pressure or add fuel by VAFC? VAFC will change the injector pulse, but Fuel pressure will add more fuel per pulse. Which is better?

 

a dyno is only one of the tools you use to fully tune a car. you need a wideband in the car to do street tuning. use the dyno to find the afr where the engine makes most power, then get on the street and log the afr's and adjust the fuel to where you found it likes to be on the dyno.

 

<TABLE WIDTH="90%" CELLSPACING=0 CELLPADDING=0 ALIGN=CENTER><TR><TD>Quote, originally posted by C_A_T &raquo;</TD></TR><TR><TD CLASS="quote">With mapping for multiple gears we'd make pulls of varying loads and optimize each gear.
</TD></TR></TABLE>

What engine management allows you to have maps for each gear? Motec?

I am a firm believer that timing and fuel should be slightly different for the same MAP/rpm point if the gear is different.

 

Yes, there are very few systems that allows such detailed tuning. I'll bet you could pull it off with an AEM EMS (I've only tuned a couple, and they were street car tunes) and I've got one in the shop I've been playing with. Not multiple maps per se, but the AEM does offer some gear dependent outputs and I think there are a couple ways to tie things in at least from a fuel perspective.

Otherwise, you just have to tune for the worst case scenario which is generally the highest gear you expect to pull a significant load in. Better to get that right and be a little off in 1st or 2nd gear vs. blowing up the motor.

Even then, you still have to be prepared to alter things at the track. Be it road racing or drag racing, a number of my more successful customers take it upon themselves to make small changes in fuel pressure at the track to optimize performance. This requires knowledge, and I know some people that have screwed things up in doing so, but if you're seeking that last tenth, you do what you have to.

C_A_T

 

the ems does have a user selectable fuel trim. you can set that up to have a correction for vehicle speed, there is also an input for gear voltage... prolly only usefull with automatic tho...??, then use the o2 feedback to keep afr's right when your at cruise on the highway. O2 feedback has throttle, rpm and load settings that are used to tell the computer when to use the feedback.

ill prolly try this on the car im working on now... i have notices slight leaning in 4th and 5th gear pulls.

 

ttt for some great info

 

just to add what shawn said, most dyno tuner (or at least when I tune peoples street cars with any programmable ems) get a good assessment of the load maps used in 4th gear, basically you *clean* the rest of the table map based on this, most programmable EMS systems have graph features which tends to help clean up any blips, etc. this all helps prepare for the street tuning which obviously needs cleaning up in real time driving while the tuner cleans the maps. The load/brake type dynos like what shawn uses helps a lot better than the dyno jet.

another approach is for the users who have previous datalog files saved so when you go to the dyno you clean up the exact maps used for each gear.

Obviously after all of this said and done, the a/f ratio is night and day comparison, we're just experiencing this with two of our cars where on the dynot he car made 215hp @ wheels with 12.3 to 12.5 a/f ratios, at the track the wideband output from the logger was showing ratios of 13.1/13.3's a lot leaner by almost a whole point. So, all said and done, you do need to cater for adjustment.


Greg

 

How about tuning done on Dynopack?
Would it be more accurate when compare the A/F on dyno vs street?

 

That's what I use. If we get the load matched perfectly to what is being seen on street/track, the ratios are pretty close. But to be honest, most customers don't have the money, nor the data, to allow us to spend the time to match the load perfectly (my high end race customers do, we often use in car video to make sure - I'm sharing too much now ). So we have to estimate a load and go with it. But the cars will run a little leaner on the street.

The nice thing is, most Honda NA engines are _relatively_ insensitive to mixture from a _power_ standpoint if you get the mixture near the optimum level. IOW, chaning a full point, plus or minus, from the optimum has a small effect on power, so its fine to go a little richer on the dyno in anticipation of leaning on the road.

C_A_T

 

<TABLE WIDTH="90%" CELLSPACING=0 CELLPADDING=0 ALIGN=CENTER><TR><TD>Quote, originally posted by C_A_T &raquo;</TD></TR><TR><TD CLASS="quote">T, chaning a full point, plus or minus, from the optimum has a small effect on power, so its fine to go a little richer on the dyno in anticipation of leaning on the road.
</TD></TR></TABLE>

I agree ...

Greg

 

Then what is a good A/F to tune for on a dynopack on a very very mild built N/A 1.8 L engine?

 

There really isn't a good A/F. You usually tune to what the motor likes. Each motor is different and with an NA, you have much more leeway as far as mistakes in A/F.

 

CAT,

Have you seen that cams with more overlap should be run a little richer as indicated by the wideband? Conceptually some raw gas will be drawn thru the combustion chamber and out the exhaust during overlapp if the header is scavenging well. I've seen more overlap make more power because the burnt gases are purged from the motor this way.

Your input will be much appreciated.

 

This is very interesting... I've been tuning my motor (on the street) with my digital A/F guage (it's a voltometer style) using the stock O2 sensor. (1-wire) Basically I've been trying to keep the A/F ratio around 14:1 during WOT (depending on the RPM, during vtec it's about 13.8-13.6:1), and these leaner settings seem to give better results than when it was running much richer in the past... Or perhaps my butt dyno is off.

Basically I'm asking if a 14:1 A/F ratio is dangerously lean for my N/A Z6/Y8 motor (running on a PM6 ECU, stock Y8 cam)...

 

Conceptually I understand your point, but frankly I've never really looked into trying to pinpoint mixture differences based upon cam profiles.

The big problem is that everyone comes in with a different combination, especially when they've gone to really big cams. They usually have headwork from one of a dozen porters who all have their own philosophy on what works best (what I/E flow ratio, valve sizes, port sizes, etc.)

Ideally you'd conduct an experiment to find out. Conceptually, you'd have to ask if the improved scavenging doesn't pull more exhaust gas out as well (as you mentioned) which means the overall makeup of what the O2 sensor reads doesn't change. Also, A/F ratios are determined by how much oxygen is in the mixture (not fuel, which is how one sensor can read gas, methanol, etc.), so if you have unburnt fuel mixed in, especially in small quantities, it may not even affect the wideband reading.

I can tell you that the way I tune is to find an optimal mixture and ignition timing solution first. Then I begin to optimize cam timing. After a cam timing change, all I have to do is look at the A/F chart on a fuel injected engine to tell if it was beneficial. If it got leaner, I found more airflow and I'll find even more power when I add in more fuel to compensate.

C_A_T

 

You're undertaking a futile task. A factory O2 sensor is only accurate right around 14.7:1. Its essentially a trip switch for the ECU. If above a certain voltage, take out fuel, below a certain voltage, add fuel and it keeps doing that in closed loop. More modern factory O2 sensors are moving toward wideband capabilities (and some actually are widebands), but not the old 1-wires. To compound your problem, the A/F meters you see are pretty coarse in their resolution as well.

I'd suggest investing in a wideband if you really want to tune yourself on the street. Its a relatively small investment these days.

C_A_T

 

That sounds like a good idea, since my guage was specifically designed for wideband use anyways... What are some good places to look at to buy any wideband components?

 

<TABLE WIDTH="90%" CELLSPACING=0 CELLPADDING=0 ALIGN=CENTER><TR><TD>Quote, originally posted by C_A_T &raquo;</TD></TR><TR><TD CLASS="quote">Conceptually I understand your point, but frankly I've never really looked into trying to pinpoint mixture differences based upon cam profiles.

The big problem is that everyone comes in with a different combination, especially when they've gone to really big cams. They usually have headwork from one of a dozen porters who all have their own philosophy on what works best (what I/E flow ratio, valve sizes, port sizes, etc.)

Ideally you'd conduct an experiment to find out. Conceptually, you'd have to ask if the improved scavenging doesn't pull more exhaust gas out as well (as you mentioned) which means the overall makeup of what the O2 sensor reads doesn't change. Also, A/F ratios are determined by how much oxygen is in the mixture (not fuel, which is how one sensor can read gas, methanol, etc.), so if you have unburnt fuel mixed in, especially in small quantities, it may not even affect the wideband reading.

I can tell you that the way I tune is to find an optimal mixture and ignition timing solution first. Then I begin to optimize cam timing. After a cam timing change, all I have to do is look at the A/F chart on a fuel injected engine to tell if it was beneficial. If it got leaner, I found more airflow and I'll find even more power when I add in more fuel to compensate.

C_A_T</TD></TR></TABLE>


AWESOME info. i will remember the cam timing/ af curves when doing my own car