---------------------------------------------------------- Avionics-List Digest Archive --- Total Messages Posted Sun 01/19/03: 6 ---------------------------------------------------------- Today's Message Index: ---------------------- 1. 12:01 AM - Re: Air Data Compter (Werner Schneider) 2. 04:17 AM - Re: Air Data Compter (Tony Cann) 3. 07:07 AM - Re: Air Data Compter (Tom Brusehaver) 4. 03:04 PM - Re: Coolent flow (Rowland & Wilma Carson) 5. 11:20 PM - Re: Re: Air Data Computer (Bruce Boyes) 6. 11:20 PM - Re: Air Data Computer (Bruce Boyes) ________________________________ Message 1 _____________________________________ Time: 12:01:49 AM PST US From: "Werner Schneider" Subject: Re: Avionics-List: Air Data Compter --> Avionics-List message posted by: "Werner Schneider" Tom, you might have more luck in the aerolectrics list werner (using RKM and a memory module for the task) ----- Original Message ----- From: "Tom Brusehaver" Subject: Avionics-List: Air Data Compter > --> Avionics-List message posted by: Tom Brusehaver > > > I am building something like the airdata computer that was in the > December Circuit Cellar magazine. I was kind disappointed in the > article, there was no code, only snippets of schematics, and > fomulas with no units, but it was enough to get me started. > > I started using a HandyBoard, but it only has 8 bit A/D > convertors in it, so I couldn't get the resolution I needed. > Looking around for other A/D convertors, I found the Dallas > DS2438 one wire A/D (battery monitor), and I have had a > TINI gathering dust for about two years, so I thought it > would be perfect! > > I found David Bray's One Wire Barometer page, and thought > I'd steal parts of his amplifyer to make this more useful. > > So I was looking for stuff about connecting LCDs to the > TINI board, and google finds this Avionics-Archive from a > June 2000-March 2001 where folks were talking about TINI > and airplanes, and I felt I needed to subscribe. When I > looked further I realized I WAS subscribed to the list. > > Are folks still building stuff? I am kinda jazzed about > one-wire in airplanes, since it seems like a really easy > way to build modular things, and using the Tini as the > central controller, send information to various PDAs and > other displays. > > I have 2 modules in mind at the moment, the airdata, and > maybe if I can get it working an AOA meter. I could see > engine monitors, and general purpose sensor packages, > including fuel, landing gear, etc. > > If I am wacko, just delete this message, and I'll go away. > > Tom > > > > > > ________________________________ Message 2 _____________________________________ Time: 04:17:46 AM PST US From: Tony Cann Subject: Avionics-List: Re: Air Data Compter --> Avionics-List message posted by: Tony Cann Thanks for the discussion on 8 bit vs 12 bit. A couple of questions: - are there any places on an aircraft where we do need 12 bits? -- fuel quantity? -- fuel flow? - what do you think of multiplexing things like 6 EGT probes into one A/D? Are there accuracy or response time problems when you are leaning? Tony > Time: 08:58:07 PM PST US > From: "Jim V. Wickert" > Subject: Avionics-List: Air Data Compter > > --> Avionics-List message posted by: "Jim V. Wickert" > > Tom, > > I do a great bit of Industrial Automation and Process control which > requires precision A to D conversion for critical silicon wafer processing > control, but on an airplane what is wrong with 8bit? Many process control > situations with critical process interjection of change are using 12 bit > algorithms, very seldom do you see 16 bit,..... why, process conversion > time. For an oil, head, water or any other converstion, 8 bit is quite > adequate. A typical 8 bit conversion for a single channel will take about > 4 to6 Mil sec. for a 11 to 12 bit you are looking at 10 to 14 Mil Sec. for > a 16 bit converstion you maybe looking at 18 to 21 Mil sec this all depends > on the processor I know but what precision are you showing on the display > whole numbers we do not need 16 bit precision and the lag in most of the > Engine Information Systems is much slower that the update of the 16 bit > precesor as well?? > > We do not need to kill ourselves with A to D conversion algorithm software > development in something like C++ or Vicual C when we have at best 10 > channels of 8 bit conversions. We are better off developing intuitive > diagnostics that tracks history and control levels and will monitor other > critical functions of the operating system that will indicate a problem > before it hapens if we want to labor with software development. > > Just some simple thought process as this same ideal model has reduced cost > and improved production and quality for Intel Semiconductor production. > > Jim Wickert > Vision #159 ________________________________ Message 3 _____________________________________ Time: 07:07:33 AM PST US From: Tom Brusehaver Subject: Re: Avionics-List: Air Data Compter --> Avionics-List message posted by: Tom Brusehaver Cool, you folks don't think I am a nut :-). When I plug the MPX4115 directly into the 8 bit A/D converter, the resolution was only about .17 inches of mercury. (ATIS says altimeter 30.02, I can enter 29.97 or 30.14, I thought I could do better). So that is why I am looking at the amplifyer, to move the usable range of the sensor into the ball park of the A/D converter. Using a serial A/D converter pretty much throws out the speed question anyway. It takes longer to clock the data in than convert it :-). The DS2438 has two A/D channels, so I think it is a fine choice for the air data comuter. I am only trying to measure altitude, and airspeed (and temprature). Rate of change of these items is pretty slow, sampling 10 times a second would probably be good enough, but I'll go as fast as possible. Jim V. Wickert wrote: > --> Avionics-List message posted by: "Jim V. Wickert" > > Tom, > > I do a great bit of Industrial Automation and Process control which > requires precision A to D conversion for critical silicon wafer processing > control, but on an airplane what is wrong with 8bit? ________________________________ Message 4 _____________________________________ Time: 03:04:24 PM PST US From: Rowland & Wilma Carson Subject: Re: Avionics-List: Coolent flow --> Avionics-List message posted by: Rowland & Wilma Carson At 21:53 -0500 on 18-01-03, Tim & Diane Shankland wrote: >Avionics-List message >Does anyone know the coolant flow rate of the >Status suburu Tim - isn't there a Subaru list where you'd have a better chance of an informed answer? I'd be worthwhile checking the selection hosted by Matronics or maybe those available on Yahoo groups. Come to that, can't the Subaru aero-conversion engine suppliers give you this sort of information? It's certainly important; I know many early adopters of the Subaru in the Europa had cooling problems - although those may have been caused by airflow rather than liquid coolant flow restrictions. regards Rowland | Rowland Carson Europa Club Membership Secretary | Europa 435 G-ROWI (450 hours building) PFA #16532 EAA #168386 | e-mail website ________________________________ Message 5 _____________________________________ Time: 11:20:25 PM PST US From: Bruce Boyes Subject: Re: Avionics-List: Re: Air Data Computer --> Avionics-List message posted by: Bruce Boyes At 04:13 AM 1/19/2003 -0800, you wrote: >--> Avionics-List message posted by: Tony Cann > >Thanks for the discussion on 8 bit vs 12 bit. A couple of questions: >- are there any places on an aircraft where we do need 12 bits? > -- fuel quantity? > -- fuel flow? >- what do you think of multiplexing things like 6 EGT probes > into one A/D? Are there accuracy or response time problems when > you are leaning? > >Tony Given that 8 bits resolves better than 1/2 of 1%, I don't think my eyeball can detect that much on an analog EGT or any other gauge. The trick is applying that 8 bits to the region of interest, say 900-1400 F for EGT, that would be a resolution of 2 degrees. 0-1400F is about 6 degrees, still probably plenty. We often use 12-bit converters because they are cheap and available, and discard the low two bits or average the data so that we get 10 bits or so effective resolution. If you get an EGT update more than say twice per second while leaning that's probably fast enough. The thermocouple is not instantaneous, anyway. That would be the limiting factor - the probe's rate of change. Similar issues on other parameters - let the resolution of the sensor drive the ADC design. I don't have a clue what fuel flow gauges can resolve but I bet someone on the list does. Bruce Boyes ________________________________ Message 6 _____________________________________ Time: 11:20:26 PM PST US From: Bruce Boyes Subject: Re: Avionics-List: Air Data Computer --> Avionics-List message posted by: Bruce Boyes At 04:07 PM 1/18/2003 +0000, you wrote: >--> Avionics-List message posted by: Tom Brusehaver > > >I am building something like the airdata computer that was in the >December Circuit Cellar magazine. I was kind disappointed in the >article, there was no code, only snippets of schematics, and >fomulas with no units, but it was enough to get me started. > >I started using a HandyBoard, but it only has 8 bit A/D >convertors in it, so I couldn't get the resolution I needed. >Looking around for other A/D convertors, I found the Dallas >DS2438 one wire A/D (battery monitor), and I have had a >TINI gathering dust for about two years, so I thought it >would be perfect! I'm a big fan of TINI, in fact we will soon be manufacturing them ourselves: www.tstik.com However, the tool should always fit the job. 1Wire does not do well in electrically noisy environments, so I'm not sure how well it would do in an aircraft with the ignition system and multiple radio transmitters. I've wondered about CAN net - which TINI also supports - CAN is designed for noisy environments, and is in fact being used in air- and water- target drones by the US military. It's also used in mission critical car, truck, and bus applications. Cygnal systems now has some low cost 8051 nodes with 12-bit ADC, PGA, and -- CAN. So use those as the sensor nodes and tie them back to something like TINI for monitoring, display and logging. An iButton device would make a dandy ignition lock / security system, too. Might be a great thing for clubs, etc. An electronic log of who flew when. Bruce Boyes