What the table does
Your Target Lambda table sets the mixture you want under normal conditions. This correction table modifies that target when coolant temperature says conditions aren't normal. The ECU then fuels to the corrected target.
The key detail: the values are added to Target Lambda, not multiplied. A negative number makes the target richer. A positive number makes it leaner.
Examples
| Target Lambda | Correction | Corrected target | Gasoline AFR |
|---|---|---|---|
| 0.85 | -0.03 | 0.82 | about 12.1:1 |
| 0.85 | 0.00 | 0.85 | about 12.5:1 |
| 0.85 | +0.03 | 0.88 | about 12.9:1 |
If NSP is set to display AFR rather than lambda, the correction shows in AFR units too, so the same small step appears as a bigger number.
Axes
By default the table uses coolant temperature across the horizontal (x) axis and manifold pressure up the vertical (y) axis. Both can be switched to any ECU channel. To change the breakpoints, select the table and press F3, then edit an existing value or add a new one. When you edit an existing axis value, NSP adjusts the table so its overall shape stays the same.
Having manifold pressure on the second axis is useful. You might want extra enrichment at high temperature only under boost, while leaving cruise alone.
How we use it
- Cold engine: a little richer helps a cold engine run smoothly, since fuel doesn't vaporize well on cold surfaces.
- Hot engine: extra fuel under load when coolant is high adds a margin against knock.
- Normal operating range: leave it at zero, so the Target Lambda table is exactly what you get.
Start with small corrections and confirm what the wideband reads in logs. Big numbers here are usually covering a problem elsewhere.
This whole table relies on an accurate coolant reading. If the sensor or its calibration is off, the correction lands at the wrong temperature. Haltech's coolant sensors come in M12 x 1.5, 3/8 NPT and 1/8 NPT threads.
The other lambda corrections work the same way: air temp, fuel composition and generic. The overview is in lambda corrections.
