---------------------------------------------------------- AeroElectric-List Digest Archive --- Total Messages Posted Wed 02/03/21: 9 ---------------------------------------------------------- Today's Message Index: ---------------------- 1. 08:34 AM - Re: EFIS Erratic Temp Readings (Dick Tasker) 2. 08:39 AM - Re: EFIS Erratic Temp Readings (Robert L. Nuckolls, III) 3. 09:24 AM - Re: Some Thoughts on Z101 (Eric Page) 4. 11:19 AM - Re: EFIS Erratic Temp Readings (Robert L. Nuckolls, III) 5. 12:34 PM - Re: EFIS Erratic Temp Readings (Dick Tasker) 6. 01:43 PM - Re: Poor Man's Battery Tester (user9253) 7. 04:28 PM - Re: EFIS Erratic Temp Readings (Robert L. Nuckolls, III) 8. 05:19 PM - Re: EFIS Erratic Temp Readings (user9253) 9. 06:26 PM - Re: EFIS Erratic Temp Readings (Dick Tasker) ________________________________ Message 1 _____________________________________ Time: 08:34:28 AM PST US Subject: Re: AeroElectric-List: EFIS Erratic Temp Readings From: Dick Tasker I know there are two junctions and am quite well aware how thermocouples work and how to use them - one of the many jobs I had as an Electrical Engineer was designing thermocouple amplifiers to measure temperature in industrial settings. This was before the nifty ICs that have internal reference junction compensation so we used a temperature sensitive resistor network to compensate for the reference junction temperature. My point was more of a technical clarification. The voltage that we measure is NOT generated at a junction (or two junctions). It is generated along the whole length of the wire based on the temperature gradient along the wire in question. The only reason we get anything to measure is that the two wires used in the thermocouple have different coefficients. Not that all this is terribly pertinent in actual applications, but one should know how electrical phenomena actually work. Dick Tasker Robert L. Nuckolls, III wrote: > At 02:12 PM 2/2/2021, you wrote: >> >> Actually, the voltage is not generated at the junction. It is effectively generated along the whole length of a conductor due to the temperature difference between the two ends. In fact, the only >> reason we get any output to measure is that the two wires of a thermocouple are chosen to have different Seebeck coefficients. > > Actually no. There are TWO junctions. One is > attached to the temperature of interest, the > second resides at some REFERENCE temperature. > WAaayyy back when, multiple thermocouples scattered > about a test airplane at Cessna were paired with > 'reference' thermocouples immersed in an distilled > water ice bath (0 degrees C). > > The real temperature of interest is deduced by > comparing the test thermocouple with the reference > thermocouple. Views -A- and -B- in this excerpt > from the 'Connection. > > http://www.aeroelectric.com/articles/Thermocouple%20Junctions.pdf > > The reference junction doesn't have to be an ice bath, > it can be part of the indicating instrument wherein > 'reference' is cabin environment temperature . . . there > will be compensating thermistors in the instrument to > improve accuracy over a range of cabin environments. > > Non-electronic thermocouple instruments of old actually > measured that difference voltage with meter movements that read > full scale with tens of millivolts applied. In these > cases, there was considerable current flowing in the > thermocouple leads compared to electronic displays. > Hence, thermocouple wire lengths and wire gages were > critical to accurate measurement. > > Electronic displays don't need to draw current from > the thermocouples hence, wire length and gage > don't matter . . . there are no voltage drops in the > lead wires. JOINTS in those runs are critical > in that they need to occur in pairs so that potentially > parasitic junctions cancel each other out. See chapter > 10 in the 'Connection. > > > Bob . . . > > Un impeachable logic: George Carlin asked, "If black boxes > survive crashes, why don't they make the whole airplane > out of that stuff?" > ________________________________ Message 2 _____________________________________ Time: 08:39:46 AM PST US From: "Robert L. Nuckolls, III" Subject: RE: AeroElectric-List: EFIS Erratic Temp Readings > Is this a new condition or did it crop up after a period > of normal ops? > >Aircraft has never flown, only noticed this while testing engine. Good data point. > Are the readings believable when the engine is not > running? Does turning the alternator on/off make > a difference? > >Readings seem stable and accurate when engine not running, will test >this again. > >Will run to-morrow and then disconnect alternator Also, while exploring effects of alternator operation, you might dis-mount temperature sensors from engine and let them hang free not touching engine. > Are your thermocouples on the engine electrically > isolated . . . i.e. not grounded to the engine? > > Not sure what this means > > I have K Type bayonet EGT sensors that clamp onto > Exhaust and 1 yellow and 1 red wire that connects directly to MGL RDAC unit > > CHT sensors are under plug with same wire setup The reason for this question is founded in characteristics of legacy thermocouples designed to drive stand-alone instruments. Given that there were no wires at the instrument ends attached to ground or other ship's wiring, the engine thermocouples could be quite simply constructed with the junctions literally 'grounded' to the sensor shell. Such sensors might not play well in the modern EFIS sandbox. I rather suspect that everyone builds isolated (insulated) thermocouples such that no electrical continuity exists between the thermocouple conductors and the sensor enclosure. These would be functionally friendly in any engine management system. > The gross excursions of reading sound more like > a ground loop issue as opposed to connector quality. > > I have taken a ground from the EFIS directly to the > engine block. The engine mounting bolt is grounded to battery and > to the starter motor > > I have also tried it with the RDAC ground attached > next to the EGT sensor but this made no difference When you installed this experimental ground, were there any OTHER grounds in operation? I.e. DC power ground? To absolutely eliminate the possibility of this being a ground loop problem. ALL grounds would need to go ONLY to crankcase. However, if your thermocouples are of the insulated variety, ground loop(s) are almost assuredly not the problem. Have you conversed with the manufacturer about this? > > >None of the leads are joined, they go directly from the sensor to >the MGL RDAC unit Okay. Bob . . . Un impeachable logic: George Carlin asked, "If black boxes survive crashes, why don't they make the whole airplane out of that stuff?" ________________________________ Message 3 _____________________________________ Time: 09:24:24 AM PST US Subject: AeroElectric-List: Re: Some Thoughts on Z101 From: "Eric Page" nuckolls.bob(at)aeroelect wrote: > Can you elaborate on your design goals? - Redundant electrical supply for the engine (auto conversion with ECU and electronic ignition) - Failure of one alternator or regulator cannot cause AOG - Redundant electrical supply for minimum/emergency VFR equipment (EFIS, GPS, comm, landing light, Hobbs) - Brown-out-proof EFIS installation > Re brownout booster: I'm pondering a plan-d . . . or are we up to 'e'? There are dozens of suitable step up devices on the market but I have no knowledge of their 'spool up' times nor can I personally vouch for their specifications without putting the device under test on the bench and doing some measurements. > > I think I have a way to craft a brown-out booster of KNOWN performance thus avoiding the risks for incorporating the Chinese unknowns. Watch this space. > > A full schematic of the proposed boost system would make it easier to 'grok'. A schematic for the booster I've been fiddling with is attached. It uses TI's LM3481: https://tinyurl.com/yy9twn9t+ It will grunt 7A all day. It takes about 12ms for the output to come up into a 2A load when it's powered up from cold, and if it's already powered, output never falls more than a few mV below 12V when input is cut from 14V to 10V. It seems to work well but it's never going to be a drop-in, jellybean solution; it's all SMD, including fine-pitch and leadless components. Thanks, Eric Read this topic online here: http://forums.matronics.com/viewtopic.php?p=500549#500549 Attachments: http://forums.matronics.com//files/brown_out_booster_rev_b_520.pdf ________________________________ Message 4 _____________________________________ Time: 11:19:48 AM PST US From: "Robert L. Nuckolls, III" Subject: Re: AeroElectric-List: EFIS Erratic Temp Readings > It is generated along the whole length of the wire based on the > temperature gradient along the wire in question. The only reason > we get anything to measure is that the two wires used in the > thermocouple have different coefficients. > >Not that all this is terribly pertinent in actual applications, but >one should know how electrical phenomena actually work. That's the assertion I was having trouble with. There are thermocouple installations in petroleum refineries that are hundreds of feet long and traverse a huge range of temperatures . . . if 'gradient along the length' is in play, how can such installations offer meaningful readings from 'the far end'? I would draw your attention to View -B- of the illustration I posted. It's entirely valid to EXTEND a thermocouple any necessary distance with two copper wires . . . as long as you KNOW the temperature of the environment where the transition from t/c wire to copper takes place. That joint becomes the reference junction. I've not encountered a situation where length or surrounding environment of extension conductors have an effect on calibration of a 'zero current' thermocouple installation. Bob . . . Un impeachable logic: George Carlin asked, "If black boxes survive crashes, why don't they make the whole airplane out of that stuff?" ________________________________ Message 5 _____________________________________ Time: 12:34:07 PM PST US Subject: Re: AeroElectric-List: EFIS Erratic Temp Readings From: Dick Tasker Robert L. Nuckolls, III wrote: > That's the assertion I was having trouble with. > There are thermocouple installations in petroleum > refineries that are hundreds of feet long and > traverse a huge range of temperatures . . . if > 'gradient along the length' is in play, how > can such installations offer meaningful readings > from 'the far end'? Think about it. The thermocouple that is hundreds of feet long that goes through lots of temperature transitions from one end to the other generates voltage all along its length - sometimes positive and sometimes negative, depending on the temperature gradient along that particular section. The only thing that matters is the total temperature difference between the two ends. All other intermediate transitions cancel out. It's like putting lots of batteries in series - some pointing one way and others pointing the other. The net result is just the sum of all the batteries - positive and negative adding and subtracting respectively. > I would draw your attention to View -B- of the > illustration I posted. It's entirely valid to > EXTEND a thermocouple any necessary distance > with two copper wires . . . as long as you KNOW > the temperature of the environment where the > transition from t/c wire to copper takes place. > That joint becomes the reference junction. You are correct and this works because you have extended both wires with the same material. You don't have to use copper, although that is normal because it is relatively cheap and available. You could extend them with gold wire if you wanted and you would get the same result. The key is that both extension wires are the same material and generate the same voltage from end to end so any temperature difference between the extension point and the meter (electronic display) in question is irrelevant (the voltagesgeneratedin each copper wire cancel out). > I've not encountered a situation where length > or surrounding environment of extension conductors > have an effect on calibration of a 'zero current' > thermocouple installation. Correct. As I suggested earlier, see this and related articles for more info tinyurl.com/3njog7qv Dick ________________________________ Message 6 _____________________________________ Time: 01:43:02 PM PST US Subject: AeroElectric-List: Re: Poor Man's Battery Tester From: "user9253" XH-M609 Battery Low Voltage Disconnect Module is available on eBay for as little as $5. Let me know if there is an error in my circuit. -------- Joe Gores Read this topic online here: http://forums.matronics.com/viewtopic.php?p=500555#500555 Attachments: http://forums.matronics.com//files/battery_load_tester_110.png ________________________________ Message 7 _____________________________________ Time: 04:28:28 PM PST US From: "Robert L. Nuckolls, III" Subject: Re: AeroElectric-List: EFIS Erratic Temp Readings >Think about it. The thermocouple that is hundreds of feet long that >goes through lots of temperature transitions from one end to the >other generates voltage all along its length - sometimes positive >and sometimes negative, depending on the temperature gradient along >that particular section. The only thing that matters is the total >temperature difference between the two ends. All other intermediate >transitions cancel out. It's like putting lots of batteries in >series - some pointing one way and others pointing the other. The >net result is just the sum of all the batteries - positive and >negative adding and subtracting respectively. Suppose I brought 10' of chromel wire together at a sense junction with 100' of alumel wire. Assume further that the far ends of both wires terminate at a temperature compensated thermocouple meter. Suppose further than the 100' strand was coiled up and immersed in 100C water while the 10' strand remained at room temperature. Is it your understanding that serious errors in temperature reading would be generated by gross differences in temperature dependent voltage sources within those two wires? I.e. thermocouple class voltages (microvolts) are generated by warming two identical molecules of a conductor? What is this phenomenon called? Can you point me to literature that explains it? Bob . . . Un impeachable logic: George Carlin asked, "If black boxes survive crashes, why don't they make the whole airplane out of that stuff?" ________________________________ Message 8 _____________________________________ Time: 05:19:07 PM PST US Subject: AeroElectric-List: Re: EFIS Erratic Temp Readings From: "user9253" One time at work I measured millivolts at the end of a two wire fire alarm cable that was buried underground. Water had gotten into the cable and turned it into a battery. -------- Joe Gores Read this topic online here: http://forums.matronics.com/viewtopic.php?p=500557#500557 ________________________________ Message 9 _____________________________________ Time: 06:26:03 PM PST US Subject: Re: AeroElectric-List: EFIS Erratic Temp Readings From: Dick Tasker Robert L. Nuckolls, III wrote: >> Think about it. The thermocouple that is hundreds of feet long that goes through lots of temperature transitions from one end to the other generates voltage all along its length - sometimes >> positive and sometimes negative, depending on the temperature gradient along that particular section. The only thing that matters is the total temperature difference between the two ends. All >> other intermediate transitions cancel out. It's like putting lots of batteries in series - some pointing one way and others pointing the other. The net result is just the sum of all the batteries >> - positive and negative adding and subtracting respectively. > > Suppose I brought 10' of chromel wire together at a > sense junction with 100' of alumel wire. Assume further > that the far ends of both wires terminate at a > temperature compensated thermocouple meter. > > Suppose further than the 100' strand was coiled up and > immersed in 100C water while the 10' strand remained > at room temperature. > > Is it your understanding that serious errors in > temperature reading would be generated by gross > differences in temperature dependent voltage > sources within those two wires? I.e. thermocouple > class voltages (microvolts) are generated by > warming two identical molecules of a conductor? No, assuming the ends of each wire are at room temperature on the meter and the temperature to be measured at the other end. The only thing that matters to your measurement is the temperature difference between the two ends of the thermocouple. What happens in between cancels out. The Alumel will generate an internal voltage where it goes into the 100C water, nothing along the 100' immersed in water and generate the opposite voltage where it comes out. This does of course assume that the alumel and chromel wires are homogeneous throughout their length. Manufacturers go to great lengths to make sure this is so. If not, all bets are off. > > What is this phenomenon called? Can you point > me to literature that explains it? It is called the Seebeck effect. I previously included a link that explains the effect: tinyurl.com/3njog7qv Also refer to the links in that article for more information. For practical purposes you can certainly continue to say that the voltage is generated at each junction and that nothing happens in between. All I am doing is explaining what actually happens. Dick ------------------------------------------------------------------------------------- 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.