---------------------------------------------------------- AeroElectric-List Digest Archive --- Total Messages Posted Wed 01/11/12: 2 ---------------------------------------------------------- Today's Message Index: ---------------------- 1. 01:15 AM - Re: Re: Capacitive Fuel Level Sensor - How do they work? (Robert L. Nuckolls, III) 2. 07:37 PM - Re: (Paul Millner) ________________________________ Message 1 _____________________________________ Time: 01:15:24 AM PST US From: "Robert L. Nuckolls, III" Subject: Re: AeroElectric-List: Re: Capacitive Fuel Level Sensor - How do they work? >I believe the capacitive probe will be linear over its length as >well; it won't "know" about the cross sectional area of the tank at >any given sensing point, just whether or not fuel is present there. True for off-the-shelf probes . . . wherein the designer has no first-hand knowledge of tank geometry for the proposed installation. It is possible to build a capacitive probe that presents an output that is not linear with relation to liquid level. This patent speaks to some examples: http://tinyurl.com/6ogp4ve The advantage may be that a long capacitive probe can extend through a tank designed such that a float type sensor, on one end or the other, will 'range out', because when one end of the tank is full, the other end is still not full, or vice versa. It's also possible, though difficult, to custom bend a capacitive probe so that there's more range (flatter slope) in greater cross-sectional areas of the tank, and less range (steeper slope) in smaller areas of the tank; that tends to linearize the output. But it's a tough thing to accomplish with high resolution. Until the micro-controller came along. The last trade study I did for HBC spoke to a system with a self calibrating feature. You put the signal conditioner for a pure linear probe into a calibrate mode, put unusable fuel in tank and then tell the signal conditioner "this is zero fuel". Then fill the tank with 10% steps for capacity stopping each time to tall the signal conditioner to "remember this value as x%" ending up at 100% or full. When taken out of the calibrate mode, the signal conditioner does a linear interpolation of the fuel level across the constellation of calibrated data points to offer a very close representation of available fuel irrespective of tank and probe geometry. The approach to sensing and signal conditioning would produce a device that's adaptable to virtually any airplane and combination of probes. This produces a gage that is "too accurate" in that substantial uncertainty in calibration points is introduced by dielectric constant of various loads of fuel and temperatures. The "ultimate" system includes a calibration probe at the bottom of the tank that is always submerged and offers the signal conditioner a real-time sample of variables for that particular load of fuel. The electronics becomes so simple that it costs more to build an install a sturdy set of probes in something like a Hawker 4000 than the electronics which processes data from those probes! Bill of materials for the electronics came in at about $10 in production lots. >In the 80's in process instrumentation industry, there were a number >of well-designed (non-interactive adjustment) signal conditioners, >for under $100, that had three or four linearization adjustments >across an input range. I took a whack at such a device for Cessna about that time. They were proposing a spring cartridge in the elevator controls that could be 'wound up' to offer a constant stick-force-per-G under the range of flight conditions of airspeed and CG. Made my head get real tight! The last iteration drove a non linear cam to 'get the curve'. The system we proposed never made it into the airplane. Don't recall how they solved the problem. A couple years later, we (Electro-Mech) did what I belive is the first micro-processor driven trim system that flew on the aborted Mooney M30 program. This did use the lookup table I described for fuel level signals to drive the pivot location actuator for a servo/anti-servo tab. That showed more promise but the project got canceled (prototype missed target weights by about 500 pounds!). In any case, it proved the future utility of software for slaying such dragons. Bob . . . Bob . . . ________________________________ Message 2 _____________________________________ Time: 07:37:41 PM PST US From: Paul Millner Subject: Re: AeroElectric-List: On 1/4/2012 7:55 AM, Ken Lehman wrote: > Some rnav units that have airspeed inputs calculate drift and simply use > a lookup table to convert track to magnetic dg info. How would that work, Ken? Seems like knowing only airspeed and groundspeed, you don't know if the headwind/tailwind component is directly ahead/behind you or off a wingtip... so how does airspeed and groundspeed and track allow you to calculate heading? Paul -- Please note my new email address! millner@me.com ------------------------------------------------------------------------------------- Other Matronics Email List Services ------------------------------------------------------------------------------------- Post A New Message aeroelectric-list@matronics.com UN/SUBSCRIBE http://www.matronics.com/subscription List FAQ http://www.matronics.com/FAQ/AeroElectric-List.htm Web Forum Interface To Lists http://forums.matronics.com Matronics List Wiki http://wiki.matronics.com Full Archive Search Engine http://www.matronics.com/search 7-Day List Browse http://www.matronics.com/browse/aeroelectric-list Browse Digests http://www.matronics.com/digest/aeroelectric-list Browse Other Lists http://www.matronics.com/browse Live Online Chat! http://www.matronics.com/chat Archive Downloading http://www.matronics.com/archives Photo Share http://www.matronics.com/photoshare Other Email Lists http://www.matronics.com/emaillists Contributions http://www.matronics.com/contribution ------------------------------------------------------------------------------------- These Email List Services are sponsored solely by Matronics and through the generous Contributions of its members.