Tuning Help

[Jim]

  macka said:

Thanks Jim,

Sounds like the trick is to find MTB timing through the entire RPM range.  Am I right to assume that as the rpm increases the MTB timing advance would decrease.

Appreciate you imput, Jim 

thanks

macka

<{POST_SNAPBACK}>

Macka, Not necessarily. As rpm increases, the time from the spark plug firing to when you want peak pressure in the cylinder to occur (in the Falcon it's about 15° ATDC) decreases. I don't think I explained that very well. Try this, if you are firing the spark plug at 20°BTDC at 2000 rpm then at 4000 rpm you've only got half the time available to get the charge ignited and burning to create the peak cylinder pressure in the correct place. If the burn time was the same, then at 4000 rpm you would need to fire the spark plug earlier (more advance) to give you enough time. The reason engines fire the spark plug before top dead centre is because it takes time for the air & fuel to ignite and burn, the trick is to get it to build up to it's peak cylinder pressure at a time where you can get the most work from it.

In real life you would need to find MBT at all engine speeds and all engine loads, and an engine dyno is really the best place to do this.

Getting back to your first question of "too many load points?" Sometimes the engine response is linear and if this is the case then you only need very few points as most (all?) engine management systems interpolate their tables, however most engines are non-linear due to manifold tuning effects, cam profiles etc. If the engine is non-linear in a small range and linear over a large range then you would need to put more tuning points in the non-linear region and you could get away with fewer where it is linear. It really needs investigation of the particular engine you're tuning, and the engine management needs to be flexible to accomodate it.

[macka'sxr6t]

Thanks Jim,

You explained it OK. I was not thinking to straight. what you say makes logical sense.

In your OP how accurate are Wideband O2 sensors for tuning an engine for max power using data logging in the absense of a dyno?

macka

[Jim]

  macka said:

Thanks  Jim,

You explained it OK.  I was not thinking to straight.  what you say makes logical sense.

In your OP how accurate are Wideband O2 sensors for tuning an engine for max power using data logging in the absense of a dyno?

macka

<{POST_SNAPBACK}>

Wide band O2 sensors can be quite accurate, but only at measuring your air/fuel ratio (AFR), they do not give you any feedback on the power you are producing. The best you can do is tune to a rule of thumb "best AFR for power" however this may not be the best in all cases. One of the reasons is that if your engine is detonation limited (ie BDL is less than MBT), going richer than the "ideal" can give you some detonation relief and there may be an opportunity to put some more spark in and get closer to MBT timing. In this case you would produce more power going slightly richer with more spark than staying leaner with less. However if you are not detonation limited anywhere then tuning to a rule of thumb AFR would be okay, though this rare in modern engines where car companies are chasing fuel economy. Naturally aspirated engines usually aren't detonation limited everywhere, just usually around peak torque, though in turbocharged engines you can reach detonation by putting in more boost, so the slightly richer, more spark theory of tuning would be more applicable, and used for more of the rpm range, in the XR6T than say the XR8.

Essentially, if you want to tune for maximum power and torque, you really need direct feedback on the torque you are producing, otherwise you are just guessing. You may get close but you could never optimise the tune, particularly the ignition timing.

[macka'sxr6t]

Thanks Jim,

What is considered a safe AFR @ WOT? for general street use and the occassional drag race.

I've heard some say mid 11's and others say its OK to be lean as 12.6

macka

[Jim]

For durability I wouldn't be tuning the WOT fuelling to be much leaner than mid 11's. I'd pick the actual AFR number based on exhaust temperature, as this is a major factor in what determines the durability of the turbo and other engine components, eg exhaust valves. You may be able to get more power running leaner to 12.6 but only if you could put the spark into it. The pre-turbine exhaust temperature would probably be very high too, perhaps over 1000°C. This would certainly reduce the life of the turbine bearing carrier! As I mentioned previously, if you run it a bit richer you may be able to gain more that you lost by being able to increase the ignition advance. Advancing the spark towards MBT also has the advantage of reducing your exhaust temperatures because more of the energy is being put into the piston and less is being put into the exhaust.

On a bit of an aside and an example of spark advance and its influence on exhaust energy, I spoke to the guys from ION a couple of years ago, when they were supplying their 4 speed transmission to Maseratti (3200GT I think). As is typical of most modern cars, their transmission controller was calling for a reduction in torque on gear shifts, which is done through ignition retarding. However because of the spark retard, and the consequential increase in exhaust energy, boost was going through the roof on gear changes. The engine wasn't producing any more torque, just boost. In the end they had to make some software changes to control the boost on gear shifts otherwise they got too much wheel spin when the shift was over and they let the engine have its spark back.

The car manufacturers pick the fuelling pretty much by the exhaust temperature (though power has an influence) as they have the responsibility to ensure that the engine and the exhaust catalytic converter last for the life of the car (emissions components are required to last over 80,000 km and still meet the emissions requirements, Euro 2 or Euro 3 depending of if you have a BA or BF respectively). It's temperature that degrades catalysts, and if you go above about 950°C in the catalyst it starts to degrade very quickly.

If it is starting to look like WOT fuelling and spark is a bit of a juggling act, well it is. However if you've got a dyno getting the peak power isn't too difficult.

[macka'sxr6t]

Thanks for the informative way you reply to my questions, I really appreciate the time you are taking to answer them.

From reading your replies I get the feeling that the way to properly tune an engine would be to hold your engine at a specified load point and play with fuel and ignition maps to try and find your MTB.

As access to an engine dyno is limited.

Could you do it with a chassis dyno?

How accurate would the tune be?

I feel that it would'nt be the most cost effective way but would it be a better way than just doing 5-6 power runs @ WOT like most seem to?

I you did this type of tuning say every 1000rpm and let the ECU calculate what it needs to between these points

Would the tune feel like it has flat spots or surging?

If it did, would it then be as stated in another reply just a matter of adding another load point to the tables? to tune these characteristics out.

macka

[Jim]

Yep that's how you do it. It's called doing a spark loop. Basically you pick an AFR and engine speed and load and then you sweep the spark through a range looking for the peak torque - and that gives you your MBT.

Tuning WOT would go something like,

1. Pick a speed you want to tune.

2. Pick an AFR you want to check.

3. Make sure the engine conditions are known and fixed - that is make sure your engine coolant temperature, engine oil temperature and air intake temperature are the same for every point you tune. This is very important so that you know that any torque changes are from fuel and spark - not something else

4. Do a spark loop at those conditions to find MBT - make sure the engine does not detonate as you're advancing the spark! Remember you may get to detonation before your torque reaches its peak.

5. Go back to step 2 and check a difference AFR. Do this until you are happy that you've covered the range of AFRs you want to check.

6. Pick the best AFR and spark advance that gave you the best torque for that speed and load.

7. Go back to step 1 and do it again at another speed.

This sort of work can easily be done on an chassis dyno if you are only tuning WOT. To do a complete MAP of the engine at all speeds and loads, takes a lot more time though can be done on the chassis rolls. You just need a very stable foot so that you can hold constant loads though out the rev range. The accuracy of the tune really depends on the operator, with some patience you could get good results.

Tuning every 1000 rpm would work okay, though you may want to have more points around the point where the turbo starts coming on and the engine conditions start changing quickly. You may not be able to feel a flat spot or surging just because you don't have enough points around a certain condition - so its good practice to check in between the points once you've mapped then to see if the engine management is doing what you expect. You could even have a quick play around with spark to see if that helps you out or not. If it doesn't get any better then you don't need a point there so leave it and let the interpolation do its thing.

If someone's tuned the same basic engine many times then it becomes second nature after a while - particularly if they've tuned the exact combination of components before and have a calibration file handy. An experienced tuner may only need several WOT runs to pick any engine to engine variance and fix it very quickly. My understanding is that most of the aftermarket tuning of the XR6T is just a WOT tweek of boost, AFR and spark anyhow (along with turning off some of the engine, driveline and emission component protection).

On another aside (I seem to be doing a bit of this lately) I find it a bit amusing when people do a power curve on their car and the whole curve takes a few seconds. Car manufacturers do a dyno curve by holding the engine stable for at least 10 seconds at each point on the power curve, and using an average torque reading over that 10 seconds. The whole curve takes minutes The dynamic power curve always gives you an optimistic reading compared to a stablised one. However that said, when you are accelerating a car, everything is very dynamic. Oh yeah, you should check out the tune that you have just done (under stablised conditions) under dynamic conditions just to make sure that the engine management doesn't do funny things when everything is moving around.

Here's something to think about, so far the tuning has only been for AFR and spark, can you imagine how big the task would grow to if you try and optimise the tune if you threw the variables of cam timing and cam overlap in there also (like the XR6T engine has)?

[macka'sxr6t]

Thanks again Jim,

If proper engine mapping would provide the best way to find MBT. Why do so many tuners just tune using the WOT method?

I've just had a thought about mapping the engine properly by holding it at it's various rpm load points. Instead of having to have a steady foot would'nt you be able to use the cruise control.

By that I mean. Hold the car in 3rd in performance for Auto's 4th for manuals. Bring the car up to the desired load point rpm note then set the cruise control to hold the engine steady while you play with fuel and ignition maps to find your MBT. This would also give you a chance to play with your camshaft timing if your engine hardware and tuning software allows it.

Do you know if the Nissan CA18DET engines have the ability for independent camshaft timing?

macka

[Jim]

I presume that the tuners would only use the WOT method as they are only changing the WOT area of the engine map. No need to change it anywhere else (eg at part throttle) as this is the same as the factory tune. It would also be something that they can show the customer. If people are paying money for a tune then they want to see two lines on a graph - with the 'after' tune line higher that the 'before' tune one.

Using the cruise control on and engine dynamometer would work only if the dyno was set up to control on load only and not speed. Cruise control is a speed controller and if the dyno was also set up as a speed controller the two would fight each other. With the dyno set on load control it would control the load, the cruise control would control the speed - perfect!

My understanding is that CA18DET engines don't have variable camshaft timing of any kind, though I happy to be proven wrong as I'm not that familiar with Nissan engines. I'm pretty sure that the device on the front of the inlet cam is the camshaft position sensor.

[macka'sxr6t]

  Jim said:

I presume that the tuners would only use the WOT method as they are only changing the WOT area of the engine map.  No need to change it anywhere else (eg at part throttle) as this is the same as the factory tune. 

Using the cruise control on and engine dynamometer would work only if the dyno was set up to control on load only and not speed.  Cruise control is a speed controller and if the dyno was also set up as a speed controller the two would fight each other.  With the dyno set on load control it would control the load, the cruise control would control the speed - perfect!

<{POST_SNAPBACK}>

Surely part throttle response, is just as important in normal driving conditions as WOT.

You should'nt have to floor it everytime to see that your car has more power.

Tuning to find MBT in all throttle postions or load points, should show some big improvement in throttle response over the stock tune? or the WOT tuning method, right through the entire rpm range.

Or am I missing the point?

Do you know if any dyno's can be set up just to control load only?

macka