---------------------------------------------------------- AeroElectric-List Digest Archive --- Total Messages Posted Wed 03/12/08: 13 ---------------------------------------------------------- Today's Message Index: ---------------------- 1. 02:41 AM - Test (Bill Settle) 2. 07:29 AM - Re: Re: "Coat Hanger" antenna (Robert L. Nuckolls, III) 3. 08:09 AM - Re: Z-19 Series Architecture (Mark Sletten) 4. 08:53 AM - Re: Z-19 Series Architecture (mikef) 5. 09:50 AM - Re: Re: Z-19 Series Architecture () 6. 10:33 AM - Re: Re: Z-19 Series Architecture (Robert L. Nuckolls, III) 7. 10:34 AM - Re: Re: Z-19 Series Architecture (Robert L. Nuckolls, III) 8. 02:33 PM - Re: Diodes versus switches 101 ? (Charlie England) 9. 05:31 PM - Re: switch wiring (Ken) 10. 06:57 PM - Re: switch wiring (Robert L. Nuckolls, III) 11. 07:47 PM - Re: Diodes versus switches 101 ? (Robert L. Nuckolls, III) 12. 09:02 PM - Re: Diodes versus switches 101 ? (Bill Schlatterer) 13. 09:56 PM - Re: Diodes versus switches 101 ? (ZuluZephyr) ________________________________ Message 1 _____________________________________ Time: 02:41:01 AM PST US From: "Bill Settle" Subject: AeroElectric-List: Test ________________________________ Message 2 _____________________________________ Time: 07:29:08 AM PST US From: "Robert L. Nuckolls, III" Subject: Re: AeroElectric-List: Re: "Coat Hanger" antenna At 05:30 PM 3/11/2008 -0700, you wrote: > >From: "Robert L. Nuckolls, III" > <nuckolls.bob@cox.net> > >Subject: Re: "Coat Hanger" antenna > > >>At 07:59 AM 3/8/2008 -0800, you wrote: > > >>No offense to Ernest or Bob, the old coat hanger > >>antennas that you terminated the coax with crimp on > >>lugs has not been used in production airplanes since > >>the 50's or early 60's. > >> > >>Coat hanger antenna > > > >><http://www.aircraftspruce.com/catalog/avpages/av534.php>http://www.aircraftspruce.com/catalog/avpages/av534.php>http://www.aircraftspruce.com/catalog/avpages/av534.php > >> > >>What Eric says about "EM field energy", which I recall > >>bits from physics and armature radio, the coax, > >>BNC connector is way more efficient. That inch of > >>exposed shield/core & important lost insulation does > >>count. The "coat hanger antenna connections are > >>subject to corrosion and fatigue way more than a BNC > >>connector. Just my opinion. > > >But without quantification. The advantages of the > >modern antennas are mainly mechanical. They don't > >twist in the insulator. In high-dollar models there > >are p-static immunities due to the DC grounded > >fabrication . . . and they look sexier and have lower > >maintenance costs. > >Bob its more than sexy it is indeed that last inch of exposed >dielectric, often close to some conductive material, even >in a composite plane (don't they need a ground plane for a 1/4 >wave dipole?) that is the issue, at least to me. Sure. But let's consider the "loaded" antenna. An antenna that's physically too short for efficient operation at the desired frequency. The ideal, full sized quarter-wave mobile antennas operating at say 14 Mhz would LIKE to have about 17.5 feet of antenna on the bumper. See: http://tinyurl.com/2dyfss This is not practical on a moving vehicle so some means of electrically lengthening a "too short" antenna is called for. This is done in a variety of ways but all involve adding some reactive component (inductive or capacitive) to the antenna. My last HF mobile installation featured a 4' stub mast, an adjustable center loading coil and an 8' whip antenna above all this. Here's a modern example of this style of artificial antenna lengthening: http://www.hiqantennas.com/images/MVC-462.jpg Let's examine a practical example where there's value in making an antenna shorter while minimizing effects on performance. A few years ago I got a call from the production line on some new ELT transmitters that were shutting down with a high SWR fault. Here's the antennas in question: http://tinyurl.com/2u5c8u The original system integrator installed dual antennas (125.5 and 406 Mhz) under the fiberglas 'instep cap' of the vertical fin. There were some composite air ducts with metallic components also located in the fin cap. These not only made it necessary to fold the longer antenna over, proximity of the metal further de-tuned it. See: http://tinyurl.com/38symp Older versions of the ELT transmitter would tolerate this degradation of antenna performance, the new ones would not. I wrote a white paper suggesting a variety of options. The elegant solution was a top-loaded antenna physically short enough to mount under the fin cap and avoid close proximity to the duct-metal. http://tinyurl.com/2vv5e4 I crafted an exemplar antenna that was only 9" tall and took it to the lab for comparison with a full sized 24" antenna. It's radiation angle was higher and maximum radiation levels were down by less than 2 dB from the full sized quarter wave. Entirely satisfactory for this installation. We could have considered some other scenarios. See: http://tinyurl.com/3aaocs There are an infinite number of combinations of top, center and base loading that would minimize SWR at the frequency of interest. However, all versions other than top-loading increased fabrication labor and produced a lower performance. Let's take the ideas illustrated above and consider when one chooses to use the stone-simple, rod and feed-thru antenna. Open conductors have inductance. As a rule of thumb, we ball-park an open wire (and in this case, an extension of the antenna base below the ground plane) at 20 nH/inch. Assuming 1.5" of bare conductors total, we can expect something on the order of 30nH of series inductance at the base of the antenna. This works out to about 25 ohms of reactance at 130 Mhz. A full length, 24" antenna's center frequency would shift downward due to this loading inductance. A purest would get out the antenna analyzer and trim the antenna's physical length (shift toward capacitive reactance) to balance reactive effects of the base loading inductance. The next time I have occasion to get the test ground plane out, I'll take a look at the effects of exposed conductors at the base of the stone-simple antenna and report the amount of shortening required to shift the antenna's center frequency back to the design optimum. These exposed conductors ARE at the base of the antenna where the highest currents are encountered which go hand-in-hand with the highest radiation. So indeed, we are squirting some RF into the aircraft's interior spaces that would not be present with other antennas . . . but in practice this is seldom a problem either for EMC issues or overall performance. > > >>No one is seriously using the coat hanger wire antenna > >>on new OEM aircraft any more. > > >for the reasons stated plus some others but > >none related to observable performance. > >Bob I have to take your word for it. However in my experience >I have seen these antennas SRW go up and up and up with time >because of dissimilar metal corrosion and bad crimps. Absolutely! But connectors corrode too . . . and suffer from bad craftsmanship at assembly. I tossed out the large-area, low-pressure ground terminal supplied with the stone-simple antenna and substituted a PIDG terminal properly bonded to the skin with a 10-32 screw torqued up tight. Similarly, I made sure that the PIDG terminal at the base of the antenna had the super-mash on it between two nuts. If you don't have gas-tight joints, they will degrade with time. > Like I >said most pilots only demand short range line-O-sight >communication, not DX comm. The term observable >performance should include maintenance and durability in >my humble opinion. To be sure. The DIY antennas do have cost of ownership issues that are strongly affected by craftsmanship of the original installation. But if one understands the potential 'gotchas' with ANY antenna and exercises due diligence for craftsmanship, there's good value to be realized from either of these antennas. 't know what this "ducting" stuff is. > >Well Wikipedia to the rescue: >http://en.wikipedia.org/wiki/Coaxial_cable > >I use the word "ducting" in an informal way to describe how >a coaxial cable propagates elect mag wave energy. > > "Coaxial lines solve this problem by confining the > electromagnetic wave to the area inside the cable, > between the center conductor and the shield. The > transmission of energy in the line occurs totally through > the dielectric inside the cable between the conductors." > >"ducting" - was a casual my way of describing how coaxial >propagates EMW. Where open wire transmission lines, like >you have that inch before the "coat hanger" antenna, two >parallel wires, have the property where the electromagnetic >wave propagating down the line, extends into the space >surrounding it, an undesirable characteristic. You say the >loss is small? OK I guess, sure but every little bit helps >or hurts. The fact is radios are so good with filters and >excellent sensitivity and stability, a less than optimal >antenna can be some what tollerated. Understand. > > > >"Weak" is un-quantified. However, just like concerns about > >"iron poor blood" it may help sell alternative antennas. > >True it is unqualified, but we could get into dB or signal strength, >but all I am saying is in Aviation we are talking about strong >signals at short distances, most of the time, period. However if >you want real long range communications at distance than it >does matter. Yes "matter" is an unqualified term as well. Wouldn't argue that the stone-simple antenna does not represent the best-we-know-how-to-do. But neither is a Toyota starter bolted to a Lycoming on an adapter plate carved out on a bandsaw. I would recommend that any builder take advantage of the convenience of pre-fabricated, optimized antennas. But I think we would be remiss as teachers if we did not define the simple-ideas that control performance and service life of the DIY approaches. I would not discourage anyone from doing the best they can with what they have to work with . . . whether by choice or circumstance. > > >Bottom line is that we're fabricating OBAM aircraft. > >Yes, there are sexier, more convenient and probably > >more 'efficient' antennas to be considered over the > >stone-simple, rod and feed-thru insulator antennas > >that were quite popular 40 years ago. It all comes > >down to the builder's preferences for trading his/her > >personal expenditure of $time$ versus purchasing > >the product of someone else's $time$ for a more > >sophisticated design. > >Well sexier I guess, but cost is only times 2 and when you >have a $1000 or $3000 radio, $60 more is a drop in da bucket. Agreed. But one can do the stone-simple antenna with self procured components for much less than the $60 sale price . . . and there are very few builders who have $3000 tied up in a radio. > > >Performance risks for going the low-dollar, DIY > >route are low and easily managed. > > > >Bob > >Well I guess we disagree with the last part, but I agree >making antennas and testing them is FUN! VHF on a plane >is pretty easy, and I have no problem with a DIY, however >as I said in my opinion, if you are going to buy an antenna, >like most builders, get a $120 antenna with a BNC connector. >My focus is all towards metal planes, composites have more >flexibility to experiment with internal antennas. > >A BNC will be more secure & robust. You can have BNC >problems and other issues, but keeping the coaxial cable >intact to the last inch, is not totally trivial. True. We could build a bracket to mount the connector but it needs to be attached to the airframe with rivets, not by assembly forces of the ceramic feed-thru insulator. The problems of establishing gas-tight joints at both the ground plane and at the antenna feed-point are the same for both metal and plastic airplanes. See: http://tinyurl.com/yu38rz I would hope that the efforts I made toward that goal over 35 years ago paid off in trouble-free performance over the lifetime of the installation. The last time I had occasion to install a rod-n-cone antenna was on a wrecked straight-tail, C182 during the rebuild. It was a belly mounted antenna for a #2 nav/com. The antenna survived the accident but the belly skin was wrinkled and got replaced. > >With a composite plane the connection will be on the "back >side of the ground plane? Right? There will be signal loss and >reflected RF into the airplane, IMHO. True. The effects are non-zero but for most users are not operationally significant. > >Good discussion. The proof is in the eating of the pudding. >Fly a plane with both antennas and test them against a >ground station for both TX and RX. I think you will see a >difference in one antenna getting a weak incoming signal >and getting out a stronger signal. Of course bent whips >are a compromise. The radiating element should be as >perpendicular to the ground plane as possible but most >pilots like the looks and lower drag of the bent whip for a >little more SRW. Also the "Blade" high speed antennas >have "wider" band width. There are good enough antennas >and better, best antennas. It is definitely esoteric but for >just secure connection the BNC has it over the crimp >and screw exposed twin lead approach, at least in the >VHF band of freq's, in my opinion. Old ways are still >good but there is a reason new methods & connections >where developed. It may take more sensitive lab equip >to notice but I can't believe it makes NO difference, . . . no perceivable difference from the pilot's seat. We often get wrapped around a performance axle trying to tweak things based on test measurements that return no operational benefits. I had one instructor tell me there was benefit after a balked landing to turn off the alternator so that we could get more horsepower to the prop! Intuitively we know there's a measurable difference but the practical benefits are someplace between zero and not much. > . . . but than you say you tested it, and I have not. All of the testing has been in the lab were we do indeed see quantifiable differences. But in practice, tens of thousands of these antennas have flown and offered the user good return on investment too. > All I can go >by is when trouble shooting poor radio performance on >old planes, the first place I look is at those old corroded >connections on the "coat hanger". All very much on point . . . and I would suggest that any readers of this List make note of the fact that some installations are more deserving of preventative maintenance than others. Use of good assembly practice when making up the joints will forestall if not eliminate future maintenance events. Bob . . . ________________________________ Message 3 _____________________________________ Time: 08:09:22 AM PST US Subject: AeroElectric-List: Re: Z-19 Series Architecture From: "Mark Sletten" Listers, I've posted this question a couple of times via the Matronics Forum interface, but I've not gotten a response. My posts are coming back to me via the daily list digest, but I'm not sure Bob is getting them. Should I email directly to the Matronics list instead of posting via the Forum? Regards, Mark Mark Sletten wrote: > Bob, > > On the Z-19 rev M and Z-19/RB rev A drawings the E-BUS normal feed is from > the Main PWR Dist block thru a 7A breaker and diode. > > On all other drawings it appears the E-BUS normal feed is directly off the > Main PWR Dist Block binding post (not a breaker) thru a diode. > > Is there a reason for the different setup on the Z-19 series? > > > Regards, > > Mark Sletten -------- Mark Sletten Legacy FG N828LM http://www.legacyfgbuilder.com Read this topic online here: http://forums.matronics.com/viewtopic.php?p=169244#169244 ________________________________ Message 4 _____________________________________ Time: 08:53:29 AM PST US Subject: AeroElectric-List: Re: Z-19 Series Architecture From: "mikef" Mark, Your posts are getting through, I've not seen any replies to your question either. There was a related Z19 question a few weeks before xmas about updating the Z19 to reflect separate power feeds for coils (and possibly injectors). No further replies seen on that either. Mike Read this topic online here: http://forums.matronics.com/viewtopic.php?p=169263#169263 ________________________________ Message 5 _____________________________________ Time: 09:50:23 AM PST US Subject: RE: AeroElectric-List: Re: Z-19 Series Architecture From: Mark, I questioned that too. Not sure I ever got a strait answer. As far as I am concerned that constitutes the weakest link for what is supposed to be an essential bus system. I did see something Bob mentioned about the feed coming from both ends. One side is fed from the 7A fuse, and the other from the switch which is powered via the backup battery via its contactor switch. I believe Bob's logic is that one will backup the other. If you turn the switch off, the e-bus still has 7A and if you lose the 7A job, the switch will power the e-bus through via the backup battery contactor. The diode keeps the juice from going back other way (through the fuse) when powered by the switch. Earlier I questioned the 7A limit, but if you start breaking down the load, that can support a lot of today's panel. A Dynon and an sl30 are still within values. That's all I need to get down. The fuel pumps, ignition and engine stuff would all feed off of the secondary power switch not the e-bus. To hell with idealism - for my Sub, the e-bus will work both ways based on the intended design or an improvement in it. 7A is not going to make anything glow. In an emergency we'll be glad for the 7A. What I think is missing or confusing is the process (a standard issue for technos). Be sure you have a sound practice for throwing switches and knowing what is working etc. I may even include a warning light which closes if that ole 7A takes a pill. How'z that project going? Glenn http://n661gl.blogspot.com -----Original Message----- From: owner-aeroelectric-list-server@matronics.com [mailto:owner-aeroelectric-list-server@matronics.com] On Behalf Of Mark Sletten Sent: Wednesday, March 12, 2008 11:06 AM Subject: AeroElectric-List: Re: Z-19 Series Architecture --> Listers, I've posted this question a couple of times via the Matronics Forum interface, but I've not gotten a response. My posts are coming back to me via the daily list digest, but I'm not sure Bob is getting them. Should I email directly to the Matronics list instead of posting via the Forum? Regards, Mark Mark Sletten wrote: > Bob, > > On the Z-19 rev M and Z-19/RB rev A drawings the E-BUS normal feed is > from the Main PWR Dist block thru a 7A breaker and diode. > > On all other drawings it appears the E-BUS normal feed is directly off > the Main PWR Dist Block binding post (not a breaker) thru a diode. > > Is there a reason for the different setup on the Z-19 series? > > > Regards, > > Mark Sletten -------- Mark Sletten Legacy FG N828LM http://www.legacyfgbuilder.com Read this topic online here: http://forums.matronics.com/viewtopic.php?p=169244#169244 ________________________________ Message 6 _____________________________________ Time: 10:33:49 AM PST US From: "Robert L. Nuckolls, III" Subject: Re: AeroElectric-List: Re: Z-19 Series Architecture >Mark Sletten wrote: > > Bob, > > > > On the Z-19 rev M and Z-19/RB rev A drawings the E-BUS normal feed is from > > the Main PWR Dist block thru a 7A breaker and diode. > > > > On all other drawings it appears the E-BUS normal feed is directly off the > > Main PWR Dist Block binding post (not a breaker) thru a diode. > > > > Is there a reason for the different setup on the Z-19 series? No reason other than observance practice for protecting any small wire feeder of 6" or more in length. If the e-bus is located right next to the main bus fuse block, then you can make direct connections through the diode without fuses or breakers to protect the wires. Bob . . . ________________________________ Message 7 _____________________________________ Time: 10:34:33 AM PST US From: "Robert L. Nuckolls, III" Subject: Re: AeroElectric-List: Re: Z-19 Series Architecture At 08:50 AM 3/12/2008 -0700, you wrote: > >Mark, > >Your posts are getting through, I've not seen any replies to your question >either. > >There was a related Z19 question a few weeks before xmas about updating >the Z19 to reflect separate power feeds for coils (and possibly >injectors). No further replies seen on that either. I'm not in a position at present to modify the Z-figures for the purpose of customizing. Remember, these are intended to illustrate architectures that offer attractive failure modes effects analysis, I.e, failure tolerance. Exactly what loads are tied to what busses by what sizes of wire/fuse is up to the builder to work out as appropriate to their particular engine and suite of appliances. Bob . . . ________________________________ Message 8 _____________________________________ Time: 02:33:20 PM PST US From: Charlie England Subject: Re: AeroElectric-List: Diodes versus switches 101 ? > > Bill Schlatterer wrote: >> >> There was some discussion about Z19 using diodes in a critical ECU >> circuit and the question arose as to what kind of failure modes apply >> to diodes and their reliability as compared to switches. I think the >> trust of that thread (sub list) and post below is that diodes might >> be "automatic" but switches are safer and less likely to fail? ( I >> have no opinion but would like to understand the rationale ) This >> was posted and I was just wondering what the thought might be on the >> AE list since most of Bobs diagrams "suggest" a diode in the E-Bus >> circuit as opposed to a switch. That might not be AS critical as the >> ECU/EFI circuit on an all electric engine but the same concerns still >> apply? >> >> Posted to the Sub list: >> >> "After 30 years in the industrial electrical construction and >> maintenance field I can say without equivocation that a robust switch >> is much less likely to fail than a robust diode. The science of the >> switch is much simpler - use a lot of good conducting material that >> is not likely to corrode and support it with stout mechanical parts >> that hold tightly. >> >> Now, by contrast, the science of the diode starts with getting good >> silicon material and contaminating it in just the right proportion >> with just the right material at just the right temperature for just >> the right time. THEN you can start on building the junctions onto >> some sort of heat dissipating holder. I have replaced dozens of solid >> state devices that failed (always failed open) and I never found the >> root cause of failure. I put in an exact replacement and the circuit >> worked just fine for years. The trade off you get for the "automatic" >> switching of a diode vs. the manual switch is in the area of >> RELIABILITY. I'll be using manual switches for all critical loads." >> >> Not wanting to stir anything up, just want to understand the thinking >> and factual matter in the post above? >> Thanks Bill S >> 7a Z13/8 Z32(HD E-Bus) Z35(7ah) >> > Not a direct answer to the question of whether a diode is more likely to fail than a switch, but a response to the method used to reach his answer. Old pre-1970's era point-type ignitions are based around switches (the points). Just about every production car since then has used variations on diodes (doped semiconductor junctions) in all the various components of modern solid state ignitions (transistors, FETs, photocells, etc). Using his analysis technique, it's obvious that point-type ignitions are the way to go, since they are in fact cheaper & *much* simpler. Yet we don't see them in new cars. The real question is, which is more reliable in real world applications? Charlie ________________________________ Message 9 _____________________________________ Time: 05:31:01 PM PST US From: Ken Subject: Re: AeroElectric-List: switch wiring Kevin You may need a starter contactor or relay but an ON-OFF-momentary ON switch can probably be wired to do the mags like that. I don't know if you will find such a ready made diagram though (or anyone to recommend it), as I think such a scheme would greatly increase the risk of injury from a live mag. I can understand an OFF-Batt-Alt switch but generally I try to avoid those 3 position switches as I find them awkward, especially at night. Also it may be more common to use a separate switch for each alternator. That way one failed switch can't disable both alternators. Ken Kevin Klinefelter wrote: > > > Bob and all, > > I have a couple questions. > > I am wiring a Rotax 914 in my Europa per Z-16(sort of). I want to use a > switch for each "mag" that will be down-off, middle-on, momentary-up of > both to start. Is there a drawing somewhere on the site that shows how > to wire this? > > I also want to wire a switch to control two alternators, the rotax built > indynamo and a Denso IR alternator mounted on the vac pad. I would like > to use one switch; down-dynamo middle-off and up-main > alternator(denso).Is there a drawing showing how to wire that? > > Thanks, Kevin > ________________________________ Message 10 ____________________________________ Time: 06:57:44 PM PST US From: "Robert L. Nuckolls, III" Subject: Re: AeroElectric-List: switch wiring Kevin Klinefelter wrote: Bob and all, I have a couple questions. I am wiring a Rotax 914 in my Europa per Z-16(sort of). I want to use a switch for each "mag" that will be down-off, middle-on, momentary-up of both to start. Is there a drawing somewhere on the site that shows how to wire this? I also want to wire a switch to control two alternators, the rotax built indynamo and a Denso IR alternator mounted on the vac pad. I would like to use one switch; down-dynamo middle-off and up-main alternator(denso).Is there a drawing showing how to wire that? Thanks, Kevin For the ignition switches you need two, two-pole, three position, center off, one side momentary devices. Another way to describe this switch is 2P3T on-off-(on). This would be a B&C S700-2-5 switch. See: http://www.bandc.biz Wire one side of each switch to their respective ignition modules such that the switch is closed in the down position and shorts the ignition module to an OFF condition. Wire the other sides of the two switches in series such that both have to be lifted to the START position to engage the starter contactor. Recommend you keep the two alternator switches separate. You don't want failure of one switch to cause both alternators to become unavailable to you. Bob . . . ----------------------------------------) ( . . . a long habit of not thinking ) ( a thing wrong, gives it a superficial ) ( appearance of being right . . . ) ( ) ( -Thomas Paine 1776- ) ---------------------------------------- ________________________________ Message 11 ____________________________________ Time: 07:47:32 PM PST US From: "Robert L. Nuckolls, III" Subject: Re: AeroElectric-List: Diodes versus switches 101 ? There was some discussion about Z19 using diodes in a critical ECU circuit and the question arose as to what kind of failure modes apply to diodes and their reliability as compared to switches. I think the trust of that thread (sub list) and post below is that diodes might be "automatic" but switches are safer and less likely to fail? ( I have no opinion but would like to understand the rationale ) This was posted and I was just wondering what the thought might be on the AE list since most of Bobs diagrams "suggest" a diode in the E-Bus circuit as opposed to a switch. That might not be AS critical as the ECU/EFI circuit on an all electric engine but the same concerns still apply? Posted to the Sub list: "After 30 years in the industrial electrical construction and maintenance field I can say without equivocation that a robust switch is much less likely to fail than a robust diode. The science of the switch is much simpler - use a lot of good conducting material that is not likely to corrode and support it with stout mechanical parts that hold tightly. Now, by contrast, the science of the diode starts with getting good silicon material and contaminating it in just the right proportion with just the right material at just the right temperature for just the right time. THEN you can start on building the junctions onto some sort of heat dissipating holder. I have replaced dozens of solid state devices that failed (always failed open) Interesting! Most solid state failures I've observed were shorted devices (overheated and or voltage spiked). The devices that did go OPEN were transistors that shorted first and then opened their emitter bond-wires due to over current. . . . and I never found the root cause of failure. I put in an exact replacement and the circuit worked just fine for years. The trade off you get for the "automatic" switching of a diode vs. the manual switch is in the area of RELIABILITY. I'll be using manual switches for all critical loads." Not wanting to stir anything up, just want to understand the thinking and factual matter in the post above? About 1980 I was designing a new solid state speed controller and runaway monitor system for the Lear 55 and ultimately the entire fleet of 30 series aircraft. It was the first time I was tasked with doing a formal failure mode effects analysis and mean-time-between-failures (MTBF) study on a new product. I dug out Mil-HDK-217 and began slogging through the part-by-part service life prediction algorithms. By the time some 200+ items of solder joints, transistors, resistors, capacitors, etc were all accounted for, I was pleased to turn the crank and a really nice MTBF number on the order of 10,000 hours fell out. But wait, there was this really expensive, mil-spec, hermetically sealed 25A power relay used as a last-ditch backup disconnect. After factoring this device into the grand scheme of things, my shinning MTBF number fell to something on the order of 900 hours! Several times over the last 25 years I've asked the guys at Electromech's field service shop how the Lear trim controllers are holding up. The failure items they have to report were surprising. MOST of the field failures were in mass-terminated ribbon cable connectors followed by loss of ground through the mounting hardware that held the two boards together. Seems this was before I learned about having solid, on-purpose wiring grounds and NOT to depend on mountings that loosen and/or corrode. Electronically, most failures were the usual gang of jelly bean components, solder joints, etc. Most failures were in the monitor boards that had 3x the parts count of a controller board. There were NO failures of the controller that caused a runaway and a tiny fraction of the failures incapacitated the system. There had been no failures for that piece-o-#@@$, high-dollar relay that crashed my MTBF study. Bottom line: Was this relay badly misjudged? No, I had a pretty good understanding of contact physics and designed the electronics such that relay contacts were closed before electronics actually caused the motor to run. Similarly, electronics shut the motor down and waited some tens of milliseconds before opening the relay. Hence, the relay was never required to actually switch any power. No current was flowing when the relay contacts opened or closed and current was not allowed to flow until contacts stopped bouncing on closure. This is why Lear bought my design with the somewhat distorted MTBF numbers because in thousands of hours of flight, the relay never really saw any switching service. Nowadays, there are more accurate considerations of how a part is used that will provide a more realistic prediction of failure rates. Does this mean that the very robust, mil-spec relay is the golden child of the design and the electronics were left holding the bag for all the failures? No, most failures were due to bonding/connection issues. This controller resides in the vertical fin, just under the leading edge of the stabilizer. This has to be the worst environment in the airplane for environmental extremes. For the most part, the electronics seem to be living up to predictions for a long and quite satisfactory service life in thousands of hours. My personal experience since supports an assertion that the silicon rectifier diode has an expected service life that will far outpace any electro-mechanical device (relay or switch) that is tasked with controlling current flow. Some of my most vexing field failures problems to solve involve relays or switches. I've never had to chase down root cause for failures of energy steering diodes in an airplane. Now, let us consider the gentleman's contrary assertion. No doubt he has replaced blown diodes and perceives them to be less robust than a switch. I cannot help but wonder if his problem children were not subject to overheating and or over-voltage due to industrial line transients or local lightning. Whatever the root cause of his observed failures, there is nothing in my experience or that of my colleagues in the aircraft industry that parallels his experience. I'm not suggesting that he is being untruthful; only that we're probably observing an apples/oranges situation. The environments in which our diodes live and the manner in which they are applied in airplanes are sufficiently different from his industrial environment to prevent a useful comparison. So please folks, don't rip out your diodes and put in switches. There are two reasons to support this suggestion: First, we design for failure tolerance. What are the risks if a diode DOES fail? How will the pilot know it? Is it pre-flight detectable? Will the failure cripple the system such that it becomes a hazard? If the answers to any of these questions give you cause for concern about comfortable termination of flight, then redesign the system to eliminate the potential for stress. Consider his statement: "I'll be using manual switches for all critical loads." Our common usage of steering diodes for power is one of TWO power paths to an e-bus. So even if it does fail open, it has backup. If it fails shorted, we can detect this during pre-flight. Second, there is nothing in the service history of diodes (or any other solid state device) to suggest they are recognized problem children waiting for a chance to ruin your day or drive up your cost of ownership. Bob . . . ________________________________ Message 12 ____________________________________ Time: 09:02:23 PM PST US From: "Bill Schlatterer" Subject: RE: AeroElectric-List: Diodes versus switches 101 ? Thanks Bob, that tells me what I need to know. Wasn't planning on taking the diode out anyway but it's good to understand the why and why not? My only regret about my building process is that once I'm done, I won't have much use for the things I have come to understand from this list. Education is a good thing but practical use makes it excellent. In my case, I have learned just enough from this list to recognize what I don't know yet? That may be the reason I become a repeat builder ;-) Thanks Bill S -----Original Message----- From: owner-aeroelectric-list-server@matronics.com [mailto:owner-aeroelectric-list-server@matronics.com] On Behalf Of Robert L. Nuckolls, III Sent: Wednesday, March 12, 2008 9:41 PM Subject: Re: AeroElectric-List: Diodes versus switches 101 ? --> There was some discussion about Z19 using diodes in a critical ECU circuit and the question arose as to what kind of failure modes apply to diodes and their reliability as compared to switches. I think the trust of that thread (sub list) and post below is that diodes might be "automatic" but switches are safer and less likely to fail? ( I have no opinion but would like to understand the rationale ) This was posted and I was just wondering what the thought might be on the AE list since most of Bobs diagrams "suggest" a diode in the E-Bus circuit as opposed to a switch. That might not be AS critical as the ECU/EFI circuit on an all electric engine but the same concerns still apply? Posted to the Sub list: "After 30 years in the industrial electrical construction and maintenance field I can say without equivocation that a robust switch is much less likely to fail than a robust diode. The science of the switch is much simpler - use a lot of good conducting material that is not likely to corrode and support it with stout mechanical parts that hold tightly. Now, by contrast, the science of the diode starts with getting good silicon material and contaminating it in just the right proportion with just the right material at just the right temperature for just the right time. THEN you can start on building the junctions onto some sort of heat dissipating holder. I have replaced dozens of solid state devices that failed (always failed open) Interesting! Most solid state failures I've observed were shorted devices (overheated and or voltage spiked). The devices that did go OPEN were transistors that shorted first and then opened their emitter bond-wires due to over current. . . . and I never found the root cause of failure. I put in an exact replacement and the circuit worked just fine for years. The trade off you get for the "automatic" switching of a diode vs. the manual switch is in the area of RELIABILITY. I'll be using manual switches for all critical loads." Not wanting to stir anything up, just want to understand the thinking and factual matter in the post above? About 1980 I was designing a new solid state speed controller and runaway monitor system for the Lear 55 and ultimately the entire fleet of 30 series aircraft. It was the first time I was tasked with doing a formal failure mode effects analysis and mean-time-between-failures (MTBF) study on a new product. I dug out Mil-HDK-217 and began slogging through the part-by-part service life prediction algorithms. By the time some 200+ items of solder joints, transistors, resistors, capacitors, etc were all accounted for, I was pleased to turn the crank and a really nice MTBF number on the order of 10,000 hours fell out. But wait, there was this really expensive, mil-spec, hermetically sealed 25A power relay used as a last-ditch backup disconnect. After factoring this device into the grand scheme of things, my shinning MTBF number fell to something on the order of 900 hours! Several times over the last 25 years I've asked the guys at Electromech's field service shop how the Lear trim controllers are holding up. The failure items they have to report were surprising. MOST of the field failures were in mass-terminated ribbon cable connectors followed by loss of ground through the mounting hardware that held the two boards together. Seems this was before I learned about having solid, on-purpose wiring grounds and NOT to depend on mountings that loosen and/or corrode. Electronically, most failures were the usual gang of jelly bean components, solder joints, etc. Most failures were in the monitor boards that had 3x the parts count of a controller board. There were NO failures of the controller that caused a runaway and a tiny fraction of the failures incapacitated the system. There had been no failures for that piece-o-#@@$, high-dollar relay that crashed my MTBF study. Bottom line: Was this relay badly misjudged? No, I had a pretty good understanding of contact physics and designed the electronics such that relay contacts were closed before electronics actually caused the motor to run. Similarly, electronics shut the motor down and waited some tens of milliseconds before opening the relay. Hence, the relay was never required to actually switch any power. No current was flowing when the relay contacts opened or closed and current was not allowed to flow until contacts stopped bouncing on closure. This is why Lear bought my design with the somewhat distorted MTBF numbers because in thousands of hours of flight, the relay never really saw any switching service. Nowadays, there are more accurate considerations of how a part is used that will provide a more realistic prediction of failure rates. Does this mean that the very robust, mil-spec relay is the golden child of the design and the electronics were left holding the bag for all the failures? No, most failures were due to bonding/connection issues. This controller resides in the vertical fin, just under the leading edge of the stabilizer. This has to be the worst environment in the airplane for environmental extremes. For the most part, the electronics seem to be living up to predictions for a long and quite satisfactory service life in thousands of hours. My personal experience since supports an assertion that the silicon rectifier diode has an expected service life that will far outpace any electro-mechanical device (relay or switch) that is tasked with controlling current flow. Some of my most vexing field failures problems to solve involve relays or switches. I've never had to chase down root cause for failures of energy steering diodes in an airplane. Now, let us consider the gentleman's contrary assertion. No doubt he has replaced blown diodes and perceives them to be less robust than a switch. I cannot help but wonder if his problem children were not subject to overheating and or over-voltage due to industrial line transients or local lightning. Whatever the root cause of his observed failures, there is nothing in my experience or that of my colleagues in the aircraft industry that parallels his experience. I'm not suggesting that he is being untruthful; only that we're probably observing an apples/oranges situation. The environments in which our diodes live and the manner in which they are applied in airplanes are sufficiently different from his industrial environment to prevent a useful comparison. So please folks, don't rip out your diodes and put in switches. There are two reasons to support this suggestion: First, we design for failure tolerance. What are the risks if a diode DOES fail? How will the pilot know it? Is it pre-flight detectable? Will the failure cripple the system such that it becomes a hazard? If the answers to any of these questions give you cause for concern about comfortable termination of flight, then redesign the system to eliminate the potential for stress. Consider his statement: "I'll be using manual switches for all critical loads." Our common usage of steering diodes for power is one of TWO power paths to an e-bus. So even if it does fail open, it has backup. If it fails shorted, we can detect this during pre-flight. Second, there is nothing in the service history of diodes (or any other solid state device) to suggest they are recognized problem children waiting for a chance to ruin your day or drive up your cost of ownership. Bob . . . ________________________________ Message 13 ____________________________________ Time: 09:56:09 PM PST US Subject: AeroElectric-List: Re: Diodes versus switches 101 ? From: "ZuluZephyr" I have followed the switch vs diode controversy on a number of forums with interest since I am building a Sportsman with a Subaru H6 Engine. I am aware of crash involving a failed diode. I believe the diode was not rated for the actual amp load which was the most likely reason for the diode failure and the ensuing controversy. There is another design recommending the elimination of the diodes and using one switch to supply the fuel pumps, ECU and EFI. This provides a single point of failure which I think violates the design goals of the Z19 drawing of providing redundant paths, components and circuits. Being a committed party ( I have the engine) and wanting the simplest most reliable electrical system possible, I have designed a variation of the Z19 drawing using the same parts count with diodes (rated for the load) than includes a bypass circuit in case of diode failure (unlikely). Please see attached drawing. I welcome all comments and suggestions. Rocky Morrison Read this topic online here: http://forums.matronics.com/viewtopic.php?p=169493#169493 Attachments: http://forums.matronics.com//files/proposed_fuel_pump_eci_efi_circuit_for_eggenfellner_subaru_h6_233.pdf ------------------------------------------------------------------------------------- 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.