Home  |  News  |  Forums  |  FAQ  |  Classifieds  |  Events  |  914 Info  |  Blog  |  Members Map
IPB
 
 
 
 
 

Welcome Guest ( Log In | Register )

 

entry May 20 2005, 03:56 PM
Narrow-band O2 sensors v. wide-band O2 sensors:

Engines run on air/fuel mixtures ranging from 10:1 (very rich, lots of fuel) to 17:1 (very lean, little fuel). Best power is achieved with the mixture around 12-13:1. Best economy (gas mileage) is achieved with the mixture around 16-17:1. Best emissions with a modern catalytic convertor are achieved with 14.7:1, which also happens to be the "chemically correct" mixture for getting 100% combustion when burning gasoline in air. This ratio is called "stoichochoimetric", which many people shorten to "stoich".

Narrow-band sensors were invented to detect in a very unambiguous way when the air/fuel mixture was EXACTLY 14.7:1, since emissions were the primary concern. They act as a three-way switch: rich, stoich, lean. It's the nature of these "switches" that you can really read them as: way rich, a bit rich, a tiny bit rich, stoich, a smidge lean, somewhat lean, way lean. However, you can't reliably attach a number like 14:1 to "a smidge rich", nor can you rely on all sensors behaving properly beyond the three way rich/right/lean scale. So, the 10-20 LED meters out there aren't really telling you the truth. The main reason for this is the sensors weren't designed to be accurate beyond 14.7:1, and they're very temperature sensitive. Exhaust gas temps vary pretty wildly as load, mixture, and ignition timing vary, and these sensors have no way to compensate for the temperature fluctuations. The sensors work by generating a voltage between 0.1 and 0.9v, where 0.5v is just right. I think > 0.5v is rich, but I really don't remember. Any voltmeter can be used to read these sensors.

Wide-band sensors use a different technology, and are accurate from at least 10:1 to 20:1, down to about 0.1 point. They were originally invented for the cars that deliberately run quite lean for excellent economy. However, where the narrow band sensors are very simple devices that require very little effort to read, the wide-band sensors require much more complicated electronics to make them work. 6-7 years ago, wide-band setups cost $800-1000, and the sensors alone were over $400. Today, the sensors are down around $25 each, and the control electronics are as cheap as $150 retail. Now, you can have a meter with 10-20 LEDs on it that really mean something. You can reliably tune an engine to run exactly at 13:1 or 15:1 or 18:1.

Places to get wide-band units are Tech Edge (Australia), Innovate Motorsports (California), AEM (US), and SDSEFI (Canada).

Tuning a Type 4 using a narrowband sensor is asking for serious trouble. Cylinder head temps peak a bit lean of stoich, and if a Type 4 is tuned to run under load at stoich, CHT will soar, and you'll very likely destroy your engine. Since you can't accurately tune a narrowband sensor to run a bit rich (there's no perceptible difference, as far as the sensor is concerned, between rich and really rich), this setup is not all that useful for setting the mixture on a 914. Wideband units are now cheap enough that there's little reason not to get one in preference to a narrowband setup.

 
« Next Oldest · My Blog · Next Newest »