Adding a Humidity Sensor

[evilhonda]

2014-09-12_07.26.51evilhonda.msq

(the values in this tune are skewed for scaling and resolution) I ordered a 5v humidity sensor yesterday, on E-Bay. My tune is set up to compensate for barometric pressure inherently by means of properly scaling the tables to react correctly to the manifold pressure as barometric pressure changes. To give a visual understanding, when the scaling is correct, the change in barometric pressure is revealed as a border change of the curser path on a map that includes pressure. When baro is low the curser can not reach the highest pressure ranges accordingly. As the baro rises the curser moves further into the high pressure areas and fails to reach into the lowest pressure ranges. Effectively expanding and contracting the possible extremes of value(a smaller or larger square of accessibility to the pressure ranges on the map(s) as baro changes. All of that leaves the baro metric pressure input available to compensate for humidity. Right now I use a barometric pressure sensor and a corresponding logarithmic curve on the baro table to compensate for humidity, but around here, most of the time the moisture content is consistent with the usual weather patterns, and occasionally weather arrives that has more or less moisture content than the usual weather systems. This causes slight variations to the tune's performance.
I plan to remove the barometric sensor and replace it with a humidity sensor. Then, for compensation, The barometric pressure table will be changed from a logarithmic curve to a negatively exponential curve starting with high values in low humidity and downward as moisture content increases. The only problem I see is where to place it. Like every other sensor, placement is critical to it's performance. I'm thinking that the best place is under the hood, near the intake, any ideas?

[JamesBroChill42]

What kind of sensor is it exactly? I know that many humidity sensors can be sensitive to lots of moving air. So it may work fine around town but on the freeway it might become inaccurate.

[evilhonda]

Your right, I got the same problem from every sensor. The solution for the other sensors was to match the tuning strategy with the placement of the sensors. For example, the intake air temp would be cool while driving down the freeway and then after sitting at a red light, the sensor would heat up. I had a cold air intake so that didn't reflect what the real incoming air temperature was. To match strategy with placement, I realized that when the car is sitting still at idle, the cool air also cooled the cylinder walls and consumed excess fuel because I had to add fuel just to burn the cooler air completely, according to the o2 reading. When I eliminated the cold air intake(to draw air from high under the hood) the hot air added volatility to the combustion charge, the result was, less fuel per air mass(to get stoich), less air mass, a more consistent idle speed and better idle fuel economy, and when I take off the air flowing into the manifold was cool anyway because the moving car quickly cools the engine bay and the air sensor reflects the real air temperature entering the engine. I had to run an egt sensor in place of the coolant sensor because of the same problem with the sensory input not reflecting the real conditions, the egt sensor(placed inches from an exhaust valve) follows the true engine heat exactly and it turned out that the coolant never indicates how much heat is already present in a combustion event. So far I think the humidity sensor will work best under the hood, reflecting the real time changes to the air supply as influenced by the engine and environment. The humidity effects more than the just (displaced)air mass of a combustion charge, it also effects volatility during a combustion event, more humidity means a longer cooler burn per mass/pressure of air(requiring more fuel at higher rpm) and less humidity means a hotter faster burn per air mass/pressure. When it is humid out, the longer burn time shows up in the o2 value as richer because the latency of burning fuel/air makes the o2 sensor hotter, not because the oxygen in the charge has been more thoroughly burned. I will probably need to adjust other parameters to compensate for changes to the air conditions around the humidity sensor, not sure what yet, but I usually have to be creative to resolve problems like that, for example, faster moving air around the sensor will indicate a higher humidity unless the sensor is well designed(a comparison of a control and a variable), if it does skew the values they will likely be variances that are consistent with other driving conditions and I will just take advantage of the differences to further tuning compliance elswhere. The humidity sensor that I'm trying is for Arduino off of Ebay, they're a few dollars each, but I'm still waiting for the shipment.