Link Tuning Examples

Link Tuning Examples

ChadDouglas

Started 8/31/02

Chad Douglas' car is the first one that I will use as an example of what is done with basic Link tuning starting from scratch. Here are the paticulars on Chad's car.

Car Specs:

95 M-Edition 1.8L Late '95 model with higher CR pistons
FMII unmodified
NGK 1283 plugs (7 heat range - Unsure of gap - didn't have a gapper at the time, but they looked close to .030 - Will regap tonite before datalogging)
Car has PS and AC - I DO use the AC when ambient gets above 95F
FM Turbo single exhaust (if that matters)
Stock Automatic tranny (car is a 5-spd) radiator with FM SPAL Stage 2 Fans
Hard Dog Hard Core Hard Top bar (in case you want to figure it's weight in) ~45 lbs, I think...
Driver weighs ~280
Engine: 107,000 Miles - I did a compression check just before the FMII install and it tested at 210psi DRY in all 4 cyls with no more than 3psi variance. I did not do a wet test after getting a number that high. I ran Top Engine Cleaner through and based on the gunk that came out of the exhaust, I did get some of the carbon out, but I have not run a compression check since I cleaned it out.

Chad loaded the Turbo 1 defaults and then auto tuned the car in the non-boost rows. Then he data logged the car at 180 kpa. Here is the datalog from that first run. Also below are the 3d graphs of the fuel and timing. I also have shown a plot of the boost run from that datalog.

Here is a graphic of the fuel map. As you can see, there are some pretty significant dips in the 5 row fuel. I will smooth the 4 row, and add fuel to the holes in the 5 row to smooth it as well. The default maps for fuel would have worked here, but in most cases, they are too rich. You can see the smoothed fuel map below.

Here are the initial changes to the fuel map. You can see that the fuel curves have been smoothed and the overall fueling reduced by about 3-4 points.

Here is the timing map that we started with. You can see by the holes in the plot that there were some knocks during the initial boost control tuning. This is normal. You will see some knocks when you are first setting up the boost control. When you do the store to save the boost settings, if the car has pulled timing to keep from knocking, it will also store the timing pulls. The Link pulls 6 points, or 1.5 degrees of timing for every knock that it kills. You can see them really well in this plot.

If you look in the 5 row, you will see how the timing has some real holes in it. This is where the car knocked and the Link pulled the timing. Below is the tuned timing map with the knock pulls taken out and the timing map resmoothed.

Below is a plot of the car from the first datalog. This log shows the boost, the InjW time, the ignition time, air temp, water temp and other things that I need to see when tuning the car.

Here you can see that the InjW time is not following the map plot at all. This means that the car will have some lean spots going into boost as well as some too rich spots when in boost. The fueling really does not start to settle in and flatten out until almost 5000 RPM. There is also an increase in output from the knock sensor. This is probably due to an increase in the torque since the fuel and timing really are closer to what they should be after 5000 RPM. If you look at the fuel plot above, you can see in the fuel map where the fuel values after the 550 zone start to increase and provide more towards what should be a proper amount of fuel. For a car running 12 psi, I expect to see the InjW times that are plotted to be about 120 to 130 with the timing between 12-15 degrees to start with in the knock prone zone. This is at max boost between 4000 and 5500 RPM. The MLL estimated power in this map is about 160 hp and 155 lb/ft of torque.

Here is the first plot with the car running the tuned map on the mechanical wastegate boost level. There is some engine noise at the upper end, but no real knock yet. The fueling looks better, but still a little lean at lower RPMs. You can see that from where the boost peaks until about

This also shows the MLL estimated horsepower and torque using the defaults. As you can see, it is showing about 150 hp and about 135 lb/ft of torque. I am not really looking at what the values are, just what they look like. The torque plot compared to the fuel plot is what I am looking at. How fast does the torque come on and how long does it hang in there. We want maximum value with no knock and maximum area under the curve. This is not looking too bad for 7 psi to start. I will add more fuel from 3000-4000 in the 4 row to get the fuel and torque up. The InjW is showing that it is too lean. You can see here with the first tuned map that the hp and torque are approaching what he had at 12 psi before.

Here is the log that he did at 12 psi with the fist tuned map. You can see that the InjW times are much better and flatter. This also shows the power and torque plots. Torque is much flatter and comed on a lot faster. It is rolling off too much at the upper end. I would also suspect that the knock he is getting as the boost is coming up is from the fuel map being too lean in the 4 row. This was shown in the InjW plot from the 7 psi log plot above. I also made some changes in this boost control. All that I did was lower the +V and the -V from 10 to 8. You can see that it helped settle out the oscillations in the boost from the first plot that was done without any tuning. I would say that this boost control is very very close to being dialed in. Remember that these are 3rd gear pulls and you will have a bit of overshoot in the lower gears. After we are done with the fuel and timing changes, then we will work on the boost control some more. Here we are at about 185 hp and 160 lb/ft. Lots more room to go.

Here is the first tuned map that I sent him with the changes.
Here is the datalog from the7 psi datalog
Here is the datalog from the12 psi datalog

Stay tuned, more tuning coming!

08/31/02

Here is the results from the second tuning map.

You can see from this plot that the InjW has leveld out a bit more. Also you can see that the torque is not dropping off as fast. I also see from this that the car is making a bit more consistent power by the linear rise in the noise from the knock sensor. There is not as much of a stepped appearance to it. I want to get the InjW leveled out a bit more than it is. It is falling just a bit too fast at the upper range from about 4500 and up.

Here are the notes from the 3rd map changes in response to this datalog.

Raised droop suppresion to 8 from 10 and spike suppresion to 6 from 8 to try to get the oscillations down after boost target it reached.

Fuel. Increased the fuel by about 2 points from 4500 and up in the 4, 5 and 6 row. Looking at the datalog, the InjW was dropping off a bit after 4500 as was the torque. I added a bit more fuel to try to bring the timing up in these areas for more torque. Will use the slight increase in fuel to see what we can do for power by increasing the timing.

Increased the timing in the boost rows by about 2 points or 1/2 degree. Resloped the timing in the 4,5 and 6 rows so that the initial timing going into the boost zones will be lower, hopefully the car will be less prone to knock by starting lower.

The slight increase in the fuel and leveling of the fuel plus the 1/2 degree of timing increase should bring the torque up and level it out with the additional fuel providing a bit more protection from knock. Need to datalog this map and see how it affects knock and power.

Here is the fuel map from the 3rd tuning map

You can see from this that there is more fuel at 4500 and up in the 4, 5 and 6 rows.

Here is the timing map.

You can see that I resloped the timing going into the boost zones to cut down on knock hysterisi.
This also has 2 points more timing in the boost areas. There is also a slight increase in the timing after
5500. I have made the timing slopes steeper in an attempt to try to start brining the torque up after the
knock prone areas. These cars knock the most between 4000 and 5500.

Here is thedatalog after the second tuning map.

Here is thethird tuning map.

The best way to look at all of these changes is to load the second datalog. Then add the first map as a benchmark, the second one as a benchmark, and the third one as a benchmark. Then you can see all of the changes represented by just changing the benchmarks in the zone data screen.

This one is getting close for boost control. Now all we need to do is see what we can do for power. Bear in mind that the road that he is using, the slope of the road, the weight of the car and other things can affect the MLL power calculations. I am just using the default weigh and gearing along with tire size. If Chad can get his car weighed and we get to the dyno after the tuning is done, then we can compare the MLL power with the dyno power.

9/3/02

Here is the log plot from the 3rd map

There is not much evidence of knock yet. There does seem to be the more normal rise in engine noise near redline. Not much change in power from the last log. The torque curve is flatter and holding longer. The boost control is not too bad. Based on this log, I added 1 degree of timing to the ignition map to up the torque and power. Added more timing to the 3500 -5500 area and just a bit more to the top end. I don't want to induce high rev detonation. I also added 2 points of fuel to the fuel map to help any prevent any detonation or pre-ignition that may be brought on by the increase in timing.

Here is the datalogfrom the 3rd tuning map.

This is thefourth tuning map.

09/06/02

Chad has loaded the 4th map, datalogged it and sent the results. Here is where the tuning is just slow and
methodical. You will see from this plot that the power and torque is up. There is more area under the curve for the torque. Now it is just a matter of tweaking the timing up and addind a bit of fuel until the power gains stop, go down, or the car starts to knock and we can't kill it with fuel. Generally, I will add 1/2 to 1 degree of timing and 2-3 points of fuel to get the power squeezed up. At this point, I would say that it would probably take about 2-3 more datalogs to get the car dialed in, then I would recommend a trip to the dyno for confirmation of the tuning and a wideband O2 plot of the fuel. I also decreased the +V and -V boost control settings to help with the oscillations in the boost that he is getting coming out of boost assist. You can see where the WG duty cycle drops from 98% and the boost regulation starts, but the map pressure continues to rise to near boost target, which is 180 kpa. This works fine. What we are trying to stop is the droop after it hits target. Taking some out of -V will slow the amount that the boost falls. Taking some out of +V will slow how fast the boost rises. This should dampen the ringing.

Here is the plot.

Here is thedatalog from this run.

Here is themap file that I sent in response to this log.

I flattened the timing in the 4 row to keep any knock from starting down low. This log shows an increase in knock sensor output at 3000 RPM. I also took a bit of timing out of the 4 row and 5 row at 3000-3500 RPM. I added just a bit of fuel, 2 points, to the 4 and 5 row between 3000 and 3500. I also added 1/2 degree of timing to the 4 and 5 row to bring up the torque and power a bit. As I said, at this point it is just a matter of adding a bit of timing and a bit of fuel until it either stops gaining power, knocks, or the power falls off. I would also start tweaking individual zones to flatten and broaden the torque.

Off to Miatas in Yellowstone!