X-Virus-Scanned: clean according to Sophos on Logan.com Return-Path: Sender: To: lml@lancaironline.net Date: Tue, 05 Jan 2010 13:13:49 -0500 Message-ID: X-Original-Return-Path: Received: from imr-da03.mx.aol.com ([205.188.105.145] verified) by logan.com (CommuniGate Pro SMTP 5.3.0) with ESMTP id 4065284 for lml@lancaironline.net; Tue, 05 Jan 2010 12:19:52 -0500 Received-SPF: pass receiver=logan.com; client-ip=205.188.105.145; envelope-from=Sky2high@aol.com Received: from imo-da01.mx.aol.com (imo-da01.mx.aol.com [205.188.169.199]) by imr-da03.mx.aol.com (8.14.1/8.14.1) with ESMTP id o05HJA6g025014 for ; Tue, 5 Jan 2010 12:19:10 -0500 Received: from Sky2high@aol.com by imo-da01.mx.aol.com (mail_out_v42.5.) id q.bf2.688a2faa (14501) for ; Tue, 5 Jan 2010 12:19:08 -0500 (EST) From: Sky2high@aol.com X-Original-Message-ID: X-Original-Date: Tue, 5 Jan 2010 12:19:08 EST Subject: Re: [LML] Re: Fuel Planning - Capacitance probes X-Original-To: lml@lancaironline.net MIME-Version: 1.0 Content-Type: multipart/alternative; boundary="-----------------------------1262711948" X-Mailer: AOL 9.5 sub 155 X-Spam-Flag:NO X-AOL-SENDER: Sky2high@aol.com -------------------------------1262711948 Content-Type: text/plain; charset="US-ASCII" Content-Transfer-Encoding: 7bit Gary, Interesting. In the 320, the wire exiting the wing probe travels downward as it is penetrating the BL50 rib. The outboard wing to stub wing joint is sealed with white bathtub caulk - appropriate for this location. The trailing edge at the rear spar is not sealed because of deduced wheel well airflow. It is possible that splash from the wheel could possibly make its way around the mid-stub wing rib and possibly to the almost hidden probe end. I will look into this and make such a result even less likely. You are right - the header tank and its probe are, uh, less exposed to the problems you mentioned. I will happily report (or the NTSB will) any case where the systems I am using in my experimental airplane misinforms me so badly that I ignore the Gestalt of all data and human senses and run out of fuel. See my response to Bob. Scott Krueger PS "Science" has also led us to the Copenhagen Conference and its output ;<) In a message dated 1/5/2010 10:52:12 A.M. Central Standard Time, casey.gary@yahoo.com writes: Scott, No, we're talking about the same thing. The device uses a simple oscillator with the variable element being the capacitance in the RC network. Therefore, the frequency varies inversely with level. The total capacitance is fairly small, but not too small. As long as the DE of the fuel is consistent, the electronic components don't drift or change with temperature or supply voltage, there are no stray capacitance issues, and no water is present the sensor will work very well. The header tank is probably a good application since it is in a relatively constant temperature dry environment. In the IV/ES wing tank application what would happen if the wing/fuselage joint were to leak just above the sensor and wires? Water could find its way into the electronics, which aren't sealed, and that would create a large error or even total failure. The comments are based on science and the design of the sensor, not experience. As Bob suggested, if one were using experience alone the total field population isn't enough. Perhaps if there were 100 million out there, then maybe. In summary, our comments(if I can speak for Bob), are that the design and construction of the sensors don't justify total faith in their accuracy or reliability. They're adequate, but I would like to see better. Gary ____________________________________ Bob and Gary, I have no idea how the capacitance probes you talk about work. Here is how the VM Fuel system probes work: There are three wires emanating from the probe. One is ground. One is a regulated 5 VDC to the probe. One carries a square wave frequency output from the probe that varies with the interaction of the fluid level on the probe. For Example, the 17 gallon wing tanks go from about 7500 counts at empty to 4500 counts at full. The header setup got to be quite different - VM said the min probe length was about 14" - I wanted my header indicator to be very accurate and the probe vertical - that meant it was only 8" long. The fuel computer could not handle frequencies above 20,000. So, after studying some simple circuits from some books at Radio Shack, I built a CMOS frequency divider with Schmidt triggers to clean up the square wave - it is just a bulge in the connecting cable. Piece of cake - the header frequencies run between 12000 and 18000 and are very accurate (calibrated by 1/2 gallons). The probes do not seem to be affected by radio frequencies - of course, I am not looking at fuel level gauges during radio transmissions. Oh well, you use your experience and I'll use mine. Scott Krueger Lancair 320 PS The wing tank calibrations are so close that if I had to switch to a more modern display that only shows two tanks instead of the three I can see now, I could switch the wing output to one tank display and leave the header on the other. -------------------------------1262711948 Content-Type: text/html; charset="US-ASCII" Content-Transfer-Encoding: quoted-printable
Gary,
 
Interesting.
 
In the 320, the wire exiting the wing probe travels=20 downward as it is penetrating the BL50 rib.  The outboard= wing to=20 stub wing joint is sealed with white bathtub caulk - appropriate for this= =20 location.  The trailing edge at the rear spar is not sealed because= =20 of deduced wheel well airflow.  It is possible that splash from= the=20 wheel could possibly make its way around the mid-stub wing rib and possibl= y to=20 the almost hidden probe end.  I will look into this and make suc= h a=20 result even less likely.  You are right - the header tank and its pro= be=20 are, uh, less exposed to the problems you mentioned.
 
I will happily report (or the NTSB will) any case where the systems= I am=20 using in my experimental airplane misinforms me so badly that I ignore the= =20 Gestalt of all data and human senses and run out of fuel.
 
See my response to Bob.
 
Scott Krueger
 
PS "Science" has also led us to the Copenhagen Conference= and its=20 output ;<)
 
In a message dated 1/5/2010 10:52:12 A.M. Central Standard Time,=20 casey.gary@yahoo.com writes:
Scott,
No, we're talking about the same thing.  The device uses a sim= ple=20 oscillator with the variable element being the capacitance in the RC net= work.=20  Therefore, the frequency varies inversely with level.  The to= tal=20 capacitance is fairly small, but not too small.  As long as the DE= of the=20 fuel is consistent, the electronic components don't drift or change with= =20 temperature or supply voltage, there are no stray capacitance issues, an= d no=20 water is present the sensor will work very well.  The header tank= is=20 probably a good application since it is in a relatively constant tempera= ture=20 dry environment.  In the IV/ES wing tank application what would hap= pen if=20 the wing/fuselage joint were to leak just above the sensor and wires?=20  Water could find its way into the electronics, which aren't sealed= , and=20 that would create a large error or even total failure.

The comments are based on science and the design of the sensor, not= =20 experience.  As Bob suggested, if one were using experience alone= the=20 total field population isn't enough.  Perhaps if there were 100 mil= lion=20 out there, then maybe.  In summary, our comments(if I can speak for= Bob),=20 are that the design and construction of the sensors don't justify total= faith=20 in their accuracy or reliability.  They're adequate, but I would li= ke to=20 see better.
Gary

Bob and Gary,
 
I have no idea how the capacitance probes you= talk=20 about work.  Here is how the VM Fuel system probes work:
 
There are three wires emanating from the prob= e. =20 One is ground.  One is a regulated 5 VDC to the probe. On= e=20 carries a square wave frequency output from the probe that varies with= the=20 interaction of the fluid level on the probe.  For=20 Example,  the 17 gallon wing tanks go from about 7500 cou= nts at=20 empty to 4500 counts at full.  The header setup got to be quite dif= ferent=20 - VM said the min probe length was about 14" - I wanted my header= =20 indicator to be very accurate and the probe vertical - that meant= it was=20 only 8" long.  The fuel computer could not handle frequencies above= =20 20,000.  So, after studying some simple circuits from some books at= Radio=20 Shack, I built a CMOS frequency divider with Schmidt triggers to clean= up the=20 square wave - it is just a bulge in the connecting cable.  Piece of= cake=20 - the header frequencies run between 12000 and 18000 and are very accura= te=20 (calibrated by 1/2 gallons).  The probes do not seem to be affected= by=20 radio frequencies - of course, I am not looking at fuel level gauges dur= ing=20 radio transmissions.
 
Oh well, you use your experience and I'll use= =20 mine.
 
Scott Krueger
Lancair 320
 
PS The wing tank calibrations are so close th= at if I=20 had to switch to a more modern display that only shows two tanks instead= of=20 the three I can see now, I could switch the wing output to one tank disp= lay=20 and leave the header on the other.

 -------------------------------1262711948--