Tuning Tips using the TMO Datalogger

 

Proper engine tuning can unleash more power from your car than probably most bolt-ons alone. Tuning however is not near as easy as bolting on an exhaust system. Some consider it to be a "black art". Everyone will have somewhat different ideas about how to tune their car, we'll just cover some of the basic techneques you can use the TMO Datalogger for.

NOTE: If you have some kind of piggyback fuel computer (VPC, AFC, PMS, MASC etc.) that gives a false signal to the ECU, that false signal will be reported by the Datalogger to you. Keep that in mind when interpreting data.

We won't look at all the different variables but we will look at the more important ones.

ENGINE SPEED

It is normal to see some ripples in this curve at higher RPMs. It seems particularly noticeable when letting off the throttle after a WOT run.  This seems to be an artifact of either the ECU or the engine sensors.  It is not real. This curve shows artificial spikes in the RPM curve just after releasing the throttle (38.5 seconds).  It also shows oscillations at the low RPM, just after opening the throttle (31 seconds).  I don't know if those are real, but it does seem to be normal.

FUEL TRIMS and adjustments

Remember the old days of a tune-up where you would have to play with things like carburetor jetting?  Well, the ECU does that for you now. All cars will change a bit over their lifetimes.  The sensor calibration will drift, the injectors will get a little clogged, etc, etc.  Fuel trim is the ECU's attempt at tuning your car.

As you can see in the program, there are three levels of fuel trim - low, medium, and high.  These three levels correspond to three levels of air flow. Low is for idling, medium is roughly up to 50 MPH (I'm guessing, will be different depending on upgrades to your engine) and high is even higher. The range of adjustment is 80 to 120% for 90's and 60 to 140% for 91' and newer. 100% is the default set and relates to a stoichiometeric air/fuel ratio (which is 14.7:1).

These trims are adjusted while the ECU has the engine in full feedback with the oxygen sensor.  They only get adjusted if everything on the car has been relatively steady for at least 5-10 seconds.  The ECU takes the amount of air entering the engine, multiplies by 14.7:1 ratio, and calculates the fuel needed.  By using the O2 sensor, it can check to see if the final result of this calculation is correct.  If the O2 reads lean, the ECU increases its gain term a bit.  If it O2 reads rich, it decreases the gain term a bit.  If the car is stable, it starts to bleed this gain term into the current fuel trim value.  These fuel trim values are battery backed-up so that the ECU will be ready to apply them the next time the car starts.

You do NOT have to strive to keep the trims at 100%. You can actually run higher fuel trims and in my case I get better gas mileage. All three of my fuel trims were at 140% and the car ran fine and got better mileage because I had increased the airflow by gutting the MAF sensor which caused me to run leaner than stock (100%) so the ECU was adding as much fuel as it could (140%) to try and get the trim value down to stoichiometric again. I try and run my trims aroudn 110% now to get the best mileage and throttle response.

WARNING!!! You can run too lean at idle/part throttle and the car will not idle well and your drivability will suffer. EVERY MAF SENSOR IS DIFFERENT but most people have been able to run with the lower honeycomb removed, screw backed out all the way until it's flush and even trimming and filing the leading edges of the lower passage to a smooth "funnel" like flow surface and still run fine.

INJECTORS

Duty cycle is roughly defined as the ratio of the amount of time a signal is "active" to the amount of time available for the signal to be active. A couple examples. In the examples that follow, the amount of time available for the signal is 20ms (milliseconds).

Now, for the question of injector duty cycle.  Well, the logger spits out how long the injectors are being opened in ms.  What we need is the time available to the injector.

One might think the max amount of fuel an injector can squirt into the cylinder would be limited to the time that the intake valve is open. But that is a very short time relative to the entire four-stroke time that is really available.  The injector can squirt fuel onto the back of the injector while it is closed and let it pile up there.  This is actually done almost all the time at RPMs over 5000 (maybe lower, too) and is the reason why the intake valves get all crudded up over time with an accumulation of carbon.  This is also why you should check and clean your intake valves occasionally.

Anyway, if we decide to use all four cycles of the engine and call that the max time period, we need to find the period of four cycles.  That is simply equal to 1/RPMs times two, because it takes two rotations of the crankshaft to complete four cycles of the engine (1/RPMs because period is the inverse of frequency).

An example - what is the time available for the injector at 6000 RPM? Now, charting the time in terms of minutes just isn't convenient, so we first convert RPMs to rotations per second.

Inverting and multiplying by two, we get

So, at 6000 RPMs, the time available to your injectors is only 20ms. But that isn't the whole story - it gets worse. There is a thing called injector dead time.  This is the time it takes for your injector to respond to an "open-sesame" command from the ECU. It gets complicated because this time varies with the voltage at the battery.  The ECU has a lookup table of dead-time vs. battery voltage and really is the main reason the ECU looks at battery voltage in the first place (there are other less important reasons).  This time can range from 0.65ms to 2.1ms.  Anyway, you have to subtract this from the time available to get the true time available to the injector.

Note: subtracted from this dead time is the time it takes to turn off the injector. But, when people talk of injector duty cycle, they never include this time.  Well, they do, but they say things like, "Never run an injector over 80% (or 90%)."  But such a thing is really kind of silly to say, as duty cycle is a ratio that doesn't take into account the times involved.

For example, would you want to run at 90% duty cycle at 2000 RPM?

Now, 54ms of fuel would probably be ridiculous for most applications. But I'm just using it as an extreme example.  54ms would probably also burn out the injector - I do not think you are supposed to keep them on that long.  They do have a limit.  You need to watch for that, too.

It just so happens that at 6000 RPM, 90% gives you 18ms, which leaves you 2ms of available "dead time", which is probably a good thing.  At 8000, you only 15ms available, so 90% gives you a paltry 13.5ms (dead time of 1.5ms still probably okay on a good battery).

Of course, what is missing from this is what does 13.5ms mean, in terms of fuel?  Not easy on the first generation, because 450cc/min injectors are further derated by the 1G low fuel pressure.  That 450 rating is at 42.7 psi, not the 36.3 psi that the 1G fuel regulator uses.  450cc/min * sqrt(36.3/42.7) = 415cc/min injectors!  Multiply this by 13.5ms = 0.093375cc of fuel.  You need to figure out if this is enough for the air you are pushing through the engine.

Multiply this number by, say 11:1 A/F mass ratio and the mass of 1cc of fuel, and you get the mass of the air this amount of fuel will support. Now, if I had a way to flow-bench the MAF, I could figure out what the actual air-mass/sec rate is, and I could log it.  And you could figure out if you are getting enough fuel into your engine to match the air.

Either that, or you could look at your O2 sensor readings!

BTW, by using the injector pulse width number along with the RPM value, I should be able to produce a nice duty cycle plot.  I will add this data type in the future.  But keep in mind what it really means. Personally, I find pulse width a hell of a lot more useful because it directly correlates to the amount of fuel entering the engine - duty cycle does not. -todd-

I have personally run my 450s on my 1g up to 27+ms and have a calculated duty cycle of 130% from 4K to 7500rpms. This seems confusing but it works. If you look at all the available charts and formulas we as DSMers may be running our injectors way too hard on PUMP gas.

This is using a B.S.F.C (Brake Specific Fuel Consumption) of .50lbs/hr/hp (that is .50lbs of fuel per hour for every horsepower it produces). It is recommended turbocharged engines run a B.S.F.C. of .60 or more which means even lower numbers. This is with PUMP gas. If you run race gas you can go with a lower B.S.F.C. but who knows how low, maybe .40 or .45. As you can see, if you are trying to run fast times off of 550's and PUMP gas you are probably running them too hard. Most people are probably running their 450's and 550s too hard on the street and are running them past 80% duty cycle. Is this a bad thing? Maybe, but you can't argue with success.

The one thing I will add is at WOT you want a value to be a smooth, straight curve. If the value is jumpy at all your injectors are not squirting a smooth pattern and you will get detonation from this. If you add fuel through your VPC, AFC etc. and you notice the injector open time did not increase at high rpms, you might be in need of bigger injectors as you're probably pushing the ones you have to the max.

OXY FEEDBACK TRIM

Driving along with this value displayed is great for seeing how the car runs under normal driving conditions. I try to keep mine switching from around 100% low valleys to 140% high peaks. This will show it as a little lean on the fuel trims and it gives me the best drivability and good gas mileage.

AIR FLOW Hz

This is the actual signal the Mass Air Flow Sensor sends to the ECU. One important note here is the max the Datalogger will read is 1606hz. This is NOT the max signal the MAS can send to the ECU. The MAS is capable of sending out a signal up to ~2000hz or so. What you want to do is actually have the signal read 1606hz for most of the rpm range while at WOT. If you could extend the MAS hz signal line beyond the 1606 limit, you would see the graph line almost exactly mimic the fuel injector pulse signal line. That is because the ECU uses the MAS signal to tell how much fuel to add.

If you have some kind of piggyback fuel computer (VPC, AFC, PMS, MASC etc.) that gives a false signal to the ECU, you don't want to have the MAS hz signal set too low, say 1400hz at WOT. What this would do is tell the ECU that less air is entering the engine and the ECU would open the injectors a less amount of time for less fuel. You could increase your fuel adjustments on your AFC etc. but you don't have near the control the ECU does for the entire fuel curve under all the given conditions. What this means is, it's better to let the ECU control the fuel and deliver the most fuel it can from it's fuel tables and then if it's not enough add more by the way of your controller.

Now VPC users are even worse off. The gain knob is just that, a total fuel curve gain knob. You are just broadening the entire range and not adding specific amounts of fuel at specific rpms. This is why you need a GCC or AFC also to control individual rpm ranges.

IMPORTANT! If you run the stock MAS you should watch for MAS overrun. This is when there is more air flowing through the MAS than it can "count" and the signal starts to break down. It's very easy to see with a logger. You will see the air flow hz go up to 1606hz and stay there (normal) but above some rpm range (5500, 6000, 6500, every MAS is different) the signal will start to fall below 1606 and become unstable or "jittery". If you look at the exact time the MAS starts to lose count of airflow, you will see the injector pulse signal do the same thing, become unstable. That is because the ECU is getting a corrupt signal from the MAS and it doesn't know how much fuel to give. It trys to mimic the MAS signal and the result is a corrupt injector open time signal. As you can imagine, that results in detonation starting and then timing retards to try and stop the knock. It's a vicious cycle, one thing leading to another.

What you would think to do to cure this problem might be add more fuel, that's not right though. All you need to do is get the MAS to not lose count of the air and then the MAS signal will smooth back out and so will the injector signal and the ECU will be able to give more fuel and the knock will go down and your timing will stay advanced. :-) How you do this is by BYPASSING air around the portion of the MAS that counts the air, cutting up the lower passage. Remember what we talked about earlier though about the fuel trims. If you bypass too much air it will run fine at WOT but the idle/part throttle will be too lean as the fuel trims will be maxed out at 140% and still not enough fuel. This is when you need an AFC to add even more fuel at idle/part throttle. Another way around this is converting to the 2G MAS setup but you will definitely need an AFC or equivalent then.

KNOCK SUM

Knock can be caused by many different things. Bad gas for one, too low of a/f ratio, too high of boost on pump gas (even with say an O2 reading of .96v, very rich, you can get knock from too hot of charge air temps), this could be from intercooler heat soak, too high of boost etc. For the most part you can turn your boost up until you show knock then back it down a little bit. This is where you will get the best timing advance and most power.

NOTE, It has been found that if you give rapid throttle input (even just "goosing" the throttle) you may see a rise in knock sum then have it bleed down slowely. This also includes the TMO Stage III stutterbox mod. If you use the stutterbox to launch, you can cause a false knock signal to retard timing. The key to still using it seems to be slow throttle input. If you simply push the clutch in and step on it the engine revvs and you get a knock sum of ~20 or so. It slowly falls away if you stay there long enough. If you SLOWLY bring up the revvs to the stutter rpm then floor it there is no audible knock and no knock sum.

NOTE, a knock sum of 43 is bad, as in not good. But IMO that alone does not mean you are melting your pistons as we speak. Every car I've logged, mine included, with the exception of one has been knocking up to 43 at high rpms in high gears. Almost everytime it's been a case of too high of boost setting. Turning down the boost resulted in less knock and the car actually running better. I'm not saying go out and turn it up until you get knock but if you see a real high sum or even 43, don't have a heart attack and panic, just turn the boost down until you figure out why the knock is there.

This absolutely kills power but saves your engine. If timing has been retarded to maximum value, but knock is still present, the ECU will use the boost control solenoid to try and lower boost pressure. This may be indicated by the common "LED mod.".

OXYGEN SENSOR

There is no absolute O2 voltage that is "safe." It will vary from car to car. Lack of detonation is a much more important than O2 voltage. With that in mind, on 92+ octane pump gas, most cars can run around 0.90v to 0.92v with a minimum of detonation. With race gas it's a whole different story. :-) You can run significantly lower O2 voltages, in the 0.8v range or even lower without knock. A nice smooth and steady O2 voltage reading is desired throughout the rpm range while at WOT.

NOTE, oxygen sensors and WOT O2 readings seem to vary from car to car and even in the same car under different circumstances. Some cars won't read higher than say .94v no matter how much fuel you add, others may read .98v or even 1.0v. It also seems that your O2 voltage may change from pass to pass at the track without any adjustments made to the fuel curve, i.e. .85v one pass, .82v the next, .88v the next, etc.

If you get a sporatic "dip" in your O2 sensor reading, this can be a sign of a misfire.  You didn't actually go lean during that "dip", but you had a excessive amount of unburnt fuel AND O2 go into your exhaust. The O2 sensor just reads how much O2 in the exhaust, therefore if you have a misfire, the excess O2 that is flowing by the O2 sensor shows up as a lean condition.

THROTTLE POSITION

One key note here. Make sure it reads 100% when you go WOT. If it doesn't you need to adjust the throttle cable to make the TPS read 100% when you have it floored. I had to adjust mine because the TB wasn't opening all the way.

AIR TEMPERATURE

On VPC cars, the VPC will tell the ECU that you have a constant temp of 67 degrees at the sensor. This is not adjustable on the VPC.

BAROMETER

On VPC cars, the VPC will tell the ECU that the barometer is 0.99 bar. This is also not adjustable on the VPC. If you have a dime store fuel cut defencer, this adjusts the barometer reading to make the ECU think it's at a higher or lower altitude. I've seen this work and cause readings from .84 bar up to .99 bar.

TIMING ADVANCE

This is my personal favorite gauge on the TMO gauge package. You can watch you timing retard as you go through the gears. The more it retards the more knock you have. Ideally you want around (no less than) 15degrees advance (maybe slightly more) at WOT. As stated on the main page, any more advance and you can probably turn up the boost a little more, any lower advance and you are getting too much knock and it's killing power. This is the same for race gas and pump gas. Putting in race gas and not turning up the boost or not leaning it out will get you more advance but assuming you want to go as fast as possible, you should make more boost and/or lean it out more and then the advance may go back down depending if you have knock or not. Personally I wouldn't let the timing advance go to less than 15 degrees total advance at WOT. With race gas you may not get it to go below 18 or 20degrees even at 25psi boost.

This value is directly related to knock sum however it's not so with actual boost. The amount of boost you can run depends on a lot. Most people think about the amount of fuel, and they are right. But you can't forget about charge temps. Let's say you don't have an intercooler at all. You won't be able to run as high of boost even with plenty of fuel, you'll get knock because of the higher intake charge temperatures. Getting a good intercooler (front mount is best) really takes the charge temps down and will allow you to run a little higher boost than without it. The efficiency of the turbo you have also comes into play. A 14b turbo is not as efficient at 20psi as a 16g is, and the 16g is not as efficient at 20psi as a 20g is. If you push a small turbo too hard, all it does it heat up the air too much.

Keep in mind that all cars are different and therefore will take a slightly different state of "tune". Use these tips to get started in the right direction, soon you'll be tearing up the tracks and street!