Vance Harral

This is an interesting lesson in expectations. If the altitude hold on our Brittain maintained the selected altitude within 20' using climb rates of 50-200 fpm as shown in the OP's video, I'd consider it to be doing a fantastic job - much better than expected. For us, it's a pretty good day when it holds altitude within 100' or so, which is all it promises.

Sure, but we're talking about certification standards, not practical strategy. The original conversational offshoot was about when a certified rate-of-turn indicator is legally required, and whether or not various electronic displays are certified as primary rate-of-turn indicators for this purpose. A bit of searching around suggests to me that a pair of GI-275s are legally certified to replace the entire steam gauge 6-pack. But I couldn't actually find the installation manual, so don't quote me on it. Those same searches turn up posts that say actually getting rid of your steam gauge ASI, ALT, etc. may not necessarily be a good idea. Setting aside questions of redundancy, some people feel the size of the instrument isn't kind enough to aging eyes to be a practical replacement for every individual steam gauge. Probably not an issue if you can quickly and easily read 14pt type at 2-3 ft, but not everyone can.

Yeah... sorta. A student-oriented check list does have items that aren't necessary for an experienced pilot. For example, "TAIL - INSPECT" is adequate for an owner's preflight checklist, but a student checklist typically breaks out the individual surfaces and/or linkages. That said, the preflight checklists I write just say "INSPECT" for these items, regardless of how they're broken out, so a student is going to have to ask me what exactly they're supposed to look for, until they absorb how the system they're inspecting actually works. Similarly, I write "MIXTURE - SET FOR TAKEOFF" for students (and myself), rather than spelling out how to set the mixture. To me, that's one important difference between a "Check" list and a "Do" list. A pilot using a checklist should know how to check/do things based on their systems knowledge. The checklist is just a safeguard against tactical mental lapses. Fortunately, this philosophy is synergistic with multiple goals. Avoiding prescriptive details keeps checklists short and tidy, which in turn reduces the risk of skipping items. And as a CFI, the way a student works through such a checklist gives me a pretty good idea of their level of understanding of the systems they're checking/operating.

Not only that, but both GI 275s are powered by the same source (the alternator), so one could argue the terms of AC 91-75 aren't met. It's unclear to me if/how the backup batteries in these devices play into the AC.

I had forgotten about that AC, thanks for the reminder. Note, however, that the absence of a TC still requires two certified attitude sources. Not one primary AI and a Stratus, or GTX-345 with ADHARS, or Dynon pocket panel, or whatever.

Setting aside my own mistakes (which are many), let me give a CFI perspective on this. I've watched dozens of students miss dozens of items on checklists. In every single case, I'm able to identify something about the checklist itself which contributes to the error. So one possible snarky reply to "How many times have you missed a step on the check list?" is, "Never, with a well-designed checklist". That's not really a true statement, of course. But if you're going to pooh-pooh checklists, pointing out the foibles of a poorly-designed one is kind of a strawman argument. I actually wrote a short treatise on checklist design for a CFI candidate recently, pasted below for your consideration. First, good checklists are audience-specific: e.g. items that are helpful to a renter student pilot are distracting to an experienced owner, and vice-versa. One-size-fits-all would be simpler, and I used to think I could teach new student pilots to just grow into the kind of lightweight, streamlined checklist owners prefer. But experience has taught me differently, and given me an appreciation for the compromises made by people who write factory POH checklists. So... the checklist I give to a pre-solo student for a Cessna will work well for them, but be totally bloated for a Cessna owner; but that's fine - there isn't going to be a single checklist that works best for everyone. As an example, my normal procedures checklist for students has an item for "Taxi to Departure Runway". This is a stupidly obvious action for anyone with even a few flights under their belt, and I didn't have it there initially. But I've had no less than three different greenhorn students try to start performing pre-takeoff and runup checks while idling on the ramp, because they lack the context and experience to understand/remember that those checks are done from the run-up area. Adding an item for taxi provides an obvious, helpful separator. Second, while my student checklist is geared toward students, it still avoids detailed, prescriptive actions. e.g. it says, "Prime as needed", rather than, "Prime 3 strokes in cold weather, 1-2 strokes in warm weather, not at all if engine was recently shut down". This makes the checklist more compact, which helps avoid skipping things (see below). But it also forces students to learn the nuances of and purposes behind various actions, rather than performing them mindlessly. To put it in FAA terms, I am trying to ensure they quickly move from the rote level of learning, through understanding, to application. The end goal is for them to use checklists as actual *check* lists, rather than *do* lists. Failing to understand this is a mistake I made when I first started creating custom checklists, and that I see repeated by others - especially mid-time pilots transitioning to new aircraft. It's tempting to fill your custom checklists full of prescriptive details, but doing so winds up creating more problems than it solves. Finally, the most serious and most common checklist mistake I see students (and experienced pilots) make, is simply skipping an item outright, without realizing they've done so. Sometimes it's just an inexplicable brain fart (myself included). But I've observed a couple of specific issues caused by the design of the checklist itself, and my checklists try to avoid those: 1) Adjacent items that look almost the same are prone to getting skipped. Consider a checklist that says: Lighting - ON Headsets - ON Avionics - ON ... vs. one that says: Lights - AS DESIRED Headsets - ON Avionics - ALL ON At first glance it seems like there's no meaningful difference between these, but each line of the lower one is distinctly, visually different. Whether you're using your finger or your eyeballs or both to track to the next item, you're more likely to skip an item in the first case. This is analogous to how some old airplanes have switches that all look alike, while newer ones have distinctive shapes for landing gear, flaps, etc. Some of the wording in my checklists is designed to provide this sort of visual distinction. 2) You can teach students to use a finger or thumb as a checklist place holder to avoid skipping items, but only if the checklist mechanically supports doing so. A few checklist operations require both hands, so there is no perfect solution. But you can't thumb-track the checklist at all, if it's a big, multi-page and/or multi-column monstrosity. So within bounds of reason regarding eyesight, smaller is better, and that's why my checklists use a "long and skinny" format. 3) The more pages a checklist has, the longer it takes to find the section you care about (especially in an emergency); and the cardinal sin is a checklist that requires you to turn pages in the middle of a single sequence. For example, a lot of flight school 172s have laminated copies of the POH procedures pages in the map pocket, bound with a ring. The POH checklists themselves are fine, but the ring-bound "flip book" is unwieldy, and I've lost track of how many times I've seen a student skip items in the BEFORE TAKEOFF list because it spans two pages. For whatever reason, turning the page causes them to blow past items at the end of the preceding page and/or the beginning of the next page; and they'll often keep making that mistake even after multiple corrections. So again within bounds of reason regarding eyesight, smaller/fewer pages is better, and this is another reason for my "long and skinny" format. In summary, the whole point of having a checklist is to ensure every item on it gets checked, so anything you can do to minimize the chance of skipping items is valuable. The ergonomics of the checklist are at least as important - and maybe moreso - than what items you actually put on the list.

If you mean "traditional", standalone turn coordinators, sure. But to be clear, IFR flight requires a rate-of-turn indicator per 91.205 (except for a few exotic corner cases with three attitude indicators, one of which works through 360 degrees of pitch and roll). So you can't legally fly IFR unless something in your panel is certified as a primary rate-of-turn indicator. The G5 is certified as a primary rate-of-turn indicator, I'd bet at least a dollar that the GI-275 is too. VSI instrumentation is not required for any kind of flight, VFR or IFR.

In older Mooneys with the spring bungee system on the elevator, the position of the trim wheel affects the neutral position of the yoke on the ground. Trimming full down moves the elevator down and the yoke close to the panel, making entry/egress a little easier. Newer Mooneys with bobweights naturally "fall" to full down elevator and yoke close to the panel regardless of trim setting.

One possibility on the vacuum regulator is that the tab or safety wire which is supposed to hold the adjusting nut in place has come loose, and the nut has spun all the way in/out/off/whatever to the "unregulated" position.

Got curious about this thing: Turns out it's presumably connected to an old-school air data computer for winds aloft info: http://www.insightavionics.com/tas1000.htm. Seems like the Aspen would make this obsolete, curious why it's still there. Maybe just a "it ain't broke and I like it" item?

4 hours to completely change the pan gasket? Heck, it took me and my ham-fisted meat hooks about that long just to remove the lower cowl bowl last time we changed an alternator! OK, not really quite that long just for the lower cowl. But let's think about this, and I'm just going off the top of my head: all the intake tubes have to come off because they connect to the manifold that passes through the pan. The entire exhaust has to come off because it runs underneath the area where the pan would come off. That means disconnecting all the EGT probes in addition to the exhaust itself. You absolutely have to remove the lower cowl to do this job, which requires disconnecting the oil cooler, the air intake boot, the ram air door cable, the cowl flap linkages, and the landing light connection. I think you're probably correct that the fuel injection servo has to come off, my recollection is that's one of the things that gets in the way of tightening some of the bolts. Did I forget to mention that the throttle and mixture control cables pass through a bracket that's hung off the pan? Gotta deal with that too. All that is just the stuff I can remember, I'm sure there are other complications. If I were doing all the work myself, I'd bank on at least a week, if I didn't screw anything up. A pro with experience could do it much faster, of course, but I think it's a big job no matter who's holding the wrenches.

Especially when the gut is a little rounder than it probably should be. We call it, "trim for egress!"

The answer depends on the exact devices you choose. For the G5, seems like you already answered your own question: the terms of the STC allow it to serve as a primary attitude indicator, and/or primary DG/HSI, and/or primary turn coordinator. The terms of the STC do not allow it to serve as a primary altimeter or airspeed indicator, even though the instrument is connected to the pitot static system and can display altitude and airspeed. If you actually saw a Mooney panel with two G5s and no ASI or altimeter, it is either illegal in terms of required instrumentation, or it has some sort of very special one-off field approval.

Over the course of 17 years, five complex endorsements, three instrument ratings, one commercial certificate, one CFI, and one CFII, I'm sure our airplane has accumulated "several" collective hours in the red RPM arc. Neither the prop nor the engine has fallen off the airplane yet, though admittedly that's just one data point.

We can't afford an engine overhaul! Too busy wasting money on foolish panel toys...

Sounds like we're kindred spirits. We have also R&R'd the prop governor (twice), and replaced oil lines, in addition to the other stuff I mentioned, over the last 17 years, all primarily to deal with oil leaks. I've come to the conclusion that chasing oil leaks is just regular maintenance on airplanes that only fly 100-ish hours a year. I know it's de-rigueur to pooh-pooh the calender-time-based component of TBO recommendations, and I concur with Mike Busch on that sort of thing. But it's not as if the concept is simply made up out of thin air. Gaskets and seals and their ilk really do wear just with the passage of time, to some degree. So those of us that take decades to reach 2000, or 2500, or 3000 hours, are going to have more of these sorts of problems than the flight school 172 that gets its 3K in 6 years. When I call our oil leaks a "mess", that may reflect my personality as much as any objective measure. Let me try to be a little more scientific about describing it. If I completely wash the airplane, then we run it for three, one-ish hour flights, without cleaning, here's what we'll see: 2-3 definitive black streaks of dirty oil running back from places like the aft edge of the port side cowl cheek, a stop-drilled crack in the front nose bowl, etc. 2-3 more streaks running back down the nose gear doors, which I think is being blown out the cowl flaps enough oil pooled at the bottom of the cowl flaps to be noticeable, though not enough to actually drip a thin film of oil around the cowl mouth, and a smattering of oil on the windshield (we have the old "guppy mouth" cowl, which actually has a net outward airflow at cruise speed) an irritatingly dirty belly for having only flown 3 flights, though some of that is just ordinary combustion residue out the exhaust rather than un-burned oil Looking around the airport, I've seen worse, so I don't think any of the above is really that big a deal. But it's objectively worse now than it was a decade ago. Used to be the airplane would stay looking pretty decent for a dozen flights between cleanups.

When we first definitively identified the pan gasket as a source of leaks, we did try re-torquing the bolts near the point of the leak. Didn't help, and the mechanic was not surprised. For one thing, you can't get to all the bolts without removing the exhaust and other components, so we only re-torqued "some" bolts and not others. Doing that around a flat surface is always a little sketchy. For another, once a gasket loses enough suppleness to start seeping, squeezing it tighter is a Hail Mary idea anyway. We have not tried the gooping on exterior sealant. I'm not opposed to that trick necessarily, just seems like an Aggie-engineered thing (I can say that cuz I iz one), that's about as likely to cause a new problem as it is to fix an old one. At this point we simply live with the leaks, and have instituted a voluntary program in the partnership of trying to wipe down the cowl after every flight, so we have a better chance of noticing if they suddenly get worse. It is amazing how little oil it takes to create a giant mess. All these leaks have no observable effect on the number of hours we go between adding make-up oil. It's right at one quart every 10 hours today, just like it was in 2004.

Well, yeah, but not literally forever up to infinite time. There is a point on the far right of the "bathtub" curve where the theoretical risk creeps back up to about the same level as a newly overhauled engine. That's the sweet spot we'd all love to get to under benign circumstances. No disagreement from me that the overall fleet statistics show the right side of the bathtub way beyond published TBO, for engines that are run regularly, and ours is. But anyone with a basic understanding of statistics knows that fleet numbers only apply to an entire fleet of machinery. We don't own a fleet, we own one unit. Deciding how to manage a single unit is a lot more like gambling. We'll place our bet at some point, and it likely won't be any time soon, for all the reasons elucidated in this thread. But it's still totally reasonable, as the engine goes past 2000 hours of service, to start thinking more about what triggers feel right to you/me/us. We use all the right tools to guide us: compression, cutting filters, pulling the pickup screen, UOA, periodically checking performance against book, etc. But it's simply undeniable that feel/guess/hope is a component too, unless your philosophy is to keep running until gross failure. The end game I'm hoping for is to see a definitive knee of the curve in some combination of declining compressions, uptick in oil consumption, slightly reduced performance, more than the very occasional sliver/flake of metal in the filter, etc. If several of those things came together over the course of a year, I'd gladly overhaul the engine that winter. I'd also fully understand that doing so exposes us to a whole new set of risks for the next couple hundred hours. But like everyone else, I'll tell myself the lie that I guessed perfectly, and that the risks over those first couple hundred hours are no worse than the risk of having continued to run the old engine. In reality, no one knows the answer to that question for a single unit, so you just tell yourself whatever makes you feel good.

Nah. The manufacturers define TBO numbers because the FAA's engine certification standards require them to do so: CFR 14 Part 33, Apendix A, 33.3(a)(6). Even the lawyers know a single number for all circumstances is bogus, but the companies have to pick something to put in the ICA, otherwise no type certificate. If it were as liability-driven as your tounge-in-cheek comment suggests, there would be all kinds of efforts to get the 2000 hour TBO significantly reduced, or to push everyone much harder to value the calendar time over engine time. But there aren't, even though plenty of engines don't get anywhere near TBO before some sort of catastrophic failure. The number of aircraft operators running piston engines under anything other than Part 91 where it's not meaningful, is so small that there is no value in debating changing the existing recommendations. Whether that's a cause or effect is an interesting debate, though. One wonders if there might be a lot more piston Part 135 operations if TBO numbers were different. The sad tale told in this Mooneyspace post seems relevant.

I should have included emojis or elaborated. I am not seriously suggesting that fixing an oil pan gasket leak is not much different from an overhaul. But it is very invasive, expensive, and won't make the engine "dry" anyway - as I said, we have evidence of other minor leaks too. The more of those that occur and the greater the rate at which they leak, the greater the risk of having a more serious problem masked. So we're unlikely to replace the oil pan gasket, much more likely to keep living with it until "someday", when a collection of those leaks and/or other symptoms push us over the threshold. I'm all for running these engines long, and I don't have any serious concerns about our 2200 hour/25 year engine today. At the same time, my ideal endgame is accumulating enough small, compound, symptoms of wear, to the point we decide to pull the overhaul trigger under benign circumstances. The alternative is to run it until gross malfunction. Maybe that's cool if the gross malfunction is just a surprise bunch of unexpected metal in the filter at the next oil change, from parts we're going to replace during the overhaul anyway. But if it's from a ruined crankshaft, or worse yet the engine throws a rod in flight, I won't be too proud of the extra hours of runtime we got in exchange. The good news in our case is, we've already gotten everything we could ask from the engine, and everyone in the partnership is mentally and financially prepared for the overhaul. The complex question in our case is simply estimating the point at which the relative risk of continuing to run the old engine crosses the risk of infant mortality in a new one. There is no definitive answer for that, only speculation and gut feel.

This may be what eventually triggers us, and I'm watching this thread with interest since our engine is about the same age as the OP. We're currently at about 2175 tach hours and 25 calendar years () since overhaul on our airplane. Oil analysis, borescope, and compressions all reasonable, takeoff and cruise performance right at book numbers, no metal in oil filter or pickup screen, oil consumption has been steady at about 1 quart every 10 hours since we bought the airplane in 2004. Basically no indication whatsoever that it'll need an overhaul any time soon. But... it leaks enough oil to make a giant mess inside/around the cowl and down the belly, even just a couple of hours after cleaning everything up. We've addressed the easy stuff: valve cover gaskets, drainback tubes, etc. But the main leak source is the gasket between the oil pan/intake manifold and the crankcase, which seems difficult to address (would have to disassemble so much to drop the pan that it feels like you may as well do an overhaul). There's also some evidence it's seeping a little at the cylinder bases and/or case split, though it's just a drop or two if at all. None of those leaks are individually concerning or reason to overhaul by themselves. But collectively they indicate all the "soft seals" that are designed to last the life of the engine are timing out; and what bothers me about it is they could mask something more serious that you'd otherwise know about right away in a dry engine: crack in the case, loose accessory seal, etc.

My airplane partner tried reaching out to Cici via e-mail just last week, but no response. Hopefully she is well, and I continue to root for Brittain. But the longer they stay in this state of "deep hibernation" - where they can't even supply parts, much less service - the less optimistic I am about Brittain ever again being a viable concern. It's possible to keep the Brittain system going as long as the only thing that fails are the simple mechanical components. But if our control head unit with its printed circuit board and big multi-plane control switch goes south, I don't know of any viable path - either legal or gray market - to repair it.

Couple of ways this can be done. First, the terms of the STC for various Brittain autopilots allow them to be driven from various CDI indicators, several of which are compatible with Garmin navigators (either by name or by interface). In this setup, the CDI outputs an analog DC voltage to the autopilot giving it left/right deviation info, and the autopilot converts that to steer left/right signals. So the autopilot is not really being "driven" by the navigator in this case, at least not directly. Rather, the navigator drives the CDI, and the CDI drives the autopilot. This works OK for basic course tracking, but it won't correctly do anything" fancy" like flying around a hold. It's also unclear to me whether it is actually legal - strictly speaking - to engage the autopilot when the CDI is in GPS mode. So I have never done this, of course! Second, Brittain did have drawings to interface their autopilots to at least one GPSS roll steering unit (I think it was the DAC GDC31). So there are a few people flying around with a 430W or other Garmin navigator driving a roll steering unit, and the roll steering unit in turn feeding heading data to the Brittain system. You make this work by putting the autopilot in heading mode, and flipping a switch that muxes the autopilot heading input between the roll steering unit and the conventional heading interface for the Brittain. Again, Brittain made drawing for this back when Jerry was still alive. This is the slicker setup, as roll steering will follow more complicated GPS courses, and it is completely on the up-and-up to fly GPS courses this way. The problem we're facing in the time of the Garmin G5/GI-275 is, we'd like to connect the DG/HSI to a Garmin GAD29B adapter, which can produce the analog signals the Brittain needs; then interface the GAD29B to the Brittain autopilot. That's different enough that no existing approved STC or drawing from Brittain addresses it. Jerry was working on these drawings around the time he passed away, and there are copies floating around showing the connections. But from a legal perspective it doesn't matter, because they were never approved by the FAA.

I don't know what constitutes "limited authority", and as you note, our opinion doesn't matter - only the opinion of the decision makers at the FSDO. To be fair to those guys, it's not difficult to construct a sound argument either way. You've constructed the pro- argument, which I agree with. The con- argument goes like this: "The autopilot has three pairs of servos that connect to aileron/rudder/elevator, and it can hold a heading, intercept and fly a course, and not only hold altitude, but also pitch attitude. That's neither 'single axis' nor 'limited control authority'". Maybe I shouldn't say that out loud...

I'm in the middle of trying to get this question answered for a Brittain B5 (very similar to the B6), through the Denver FSDO. Here's what I know so far: The DER I'm working with says feasibility hinges on whether the FSDO personnel who would grant the field approval consider the B5 to be a, "simple, single-axis autopilot system with limited control authority", as defined in table row D.13h(1) of the "Major Repair and Alteration Job Aid" document which accompanies FAA Order 8300.16. That's the governing document for autopilot system modifications, and whether they can be field approved. This is a tough sell for the B5, because (1) it includes a pitch control and altitude hold system, albeit one that operates completely independent of the lateral control system involving the DG/HSI; and (2) even ignoring the pitch control system, there are servo boots connected to the rudders as well as the ailerons, which arguably makes it at least a 2-axis system. Frankly, I'd be in better shape if we had an older AccuFlite rig, but even that might not be viewed as "simple". If my DER can convince the FSDO to consider the B5 a "simple, single-axis autopilot system with limited control authority", he is willing to help me pursue a field approval for a G5/Brittain connection. His estimate is that it would cost between $5000 and $10,000 to do so, i.e about half the cost of throwing away the B5 and installing a Garmin GFC500 in its place. It's unclear to me if all that money would go in his pocket for consulting, or if the FAA actually bills applicants for their time (you'd think it would be "free" being a federal government organization, but who knows). But it's extremely unlikely I could do this without going through a paid DER consultant anyway: I lack detailed knowledge of the rules and process, I lack any direct connections to the FAA, and I lack the patience to develop these things. If the FSDO decides the Brittain B5 is not a "simple, single-axis autopilot system with limited control authority", the only path to certification is a one-time STC. My DER says certification and flight test for such a thing runs about $50K, so obviously a non-starter. Net result of all this is that I'm unlikely to pursue a field approval, but I did agree to pay the DER for a couple of hours consulting time to get an answer to the "simple, single-axis autopilot system with limited control authority" question. I'll report back here when he gets back to me. The frustration, of course, is that none of this has anything to do with whether a B5/G5 interface would work. Everyone agrees it would. But if you want your certified airplane to continue being legally certified, the paperwork rules are what they are.