I'm Sorry, Dave. I'm Afraid I Can't Do That (Me and Autopilots)

Somewhere over the Central Valley...

I get an unexpected extra day off work over the 4th of July weekend, so I do a sobering but weirdly-pleasant IFR-in-VMC flight on my own to Stockton (KSCK), where I do the ILS RWY 29R with the missed as published, all hand-flown (for fun!) under the guidance of the G1000's flight director, but not under the Cone Of Stupidity or in actual, of course. Following that, a leisurely VFR-only flight back to Hayward (KHWD) including a diversion to Livermore (KLVK) for some touch-and-goes on runway 25L there (and what a pleasure that always is!).

I’m rusty as hell on real-world IFR procedures, and it shows — and it’s the reason I do this flight on a real IFR flight plan but in clear VMC, and to a destination and along a route and including an approach all of which I'm familiar with. I can barely remember how to get my clearance on the ramp at Hayward (which takes me all of ten seconds to get right, but still…), I bungle several radio calls, I input the wrong heading on a vector, I can’t initially get the autopilot to do the right thing with altitudes (below), and I ask for the wrong approach at Stockton. I make it sound terrible, but really, nothing was so bad that I busted any limits or anything, as far as I know (the heading was 10 degrees off and I caught it quickly, for example), but it still left me seriously glad I did this today rather than in IMC. After this flight, I actually feel like I could do it for real again now; as always, it's all about complexity management.

The autopilot? A real brain failure on my part here, as usual. Initially on the departure out of Hayward I can’t seem to get it to do altitude arming using what I think is the “obvious” way, but it’s another of those weird problems I have with altitude arming on both this (the Garmin GFC 700) and the older non-integrated one in the other DA-40’s: I confuse arming and capture, and somewhere over the Central Valley on my way to Stockton I feel like kicking myself when I realise (with a few seconds of metaphorically standing back and thinking) what I should be doing (that “ALT” button in both cases means “capture / hold”, not arm, unfortunately). The good news here is I didn't obsess about it, or spend a lot of time debugging it on-the-fly (as it were) — I just shrugged and kept hand flying. The autopilot in the DA40 is a technological wonder — and increases single-pilot IFR safety an order of magnitude over hand-flying, and it’s more obvious in use than the other models — but it’s still a bit intimidating when it sits there (silently) doing the autopilot equivalent of “I'm sorry, Dave. I'm afraid I can't do that”. I'll get over it….

Traffic's A Coyote At 8,000 Feet

I'm under the Cone of Stupidity in Cessna 051 somewhere a mile or two out from the runway threshold on the Stockton (KSCK) RNAV 29R approach when the tower controller tells us to exercise caution for (as I hear it) a coyote at 8,000'. Several things immediately go through my mind as we bump along the approach: a) Umm, what?! It's bad enough that the approach is seriously bumpy, but we have to contend with a coyote at 8,000'?! b) Is "Coyote" actually some sort of aircraft type, and he's calling traffic? c) We're below 1,000' — why would he call out anything at 8,000' unless it was dropping like a brick (or, for that matter, a coyote). It's a mystery to me, so I quickly ask Evan H., my safety pilot sitting in the right seat, what the hell the controller just said.

Evan patiently explains that it sounds like there's been a report of a coyote on the runway at around the 8,000' marker. Well, that makes much more sense, but I still ask the controller to say again. He does, a bit more clearly this time, and I reply with an enthusiastic "Yeah! we'll definitely keep a lookout for that!". In the end there's no sign of any coyotes anywhere on this or the next few landings, but it's hardly the first time I've either seen or been warned about a fox or coyote or dog on the runway here at Stockton or Livermore or even at my home base, Oakland (KOAK).

And it's already been that sort of flight, not the flight I expected, and a flight full of little errors or miscommunications, none of which amount to much, but all of which when put together make me feel glad I was under the hood in clear (bumpy) VMC rather than in actual IMC.


* * *

It started a few days ago, when Evan and I were supposed to do this flight on a weekday evening (rather than a Saturday morning), but the flight was scrubbed when I noticed that the port side (red) nav light was inoperative (no matter how much I banged on the wing); we rebooked for this morning.

Then came the forecast: I'd filed a real IFR flight plan to and from Stockton because the forecast the previous night had implied there'd be a deep coastal stratus layer over much of the Bay Area today, and the weather had been rather weird for the previous few days (it very nearly rained in Oakland last night, which in California terms is surely one of the four horsemen of some sort of apocalypse). But the day dawns bright and clear, if a little cold and blustery. So none of the benign actual IMC I'd really hoped for, unfortunately.

I check NOTAMS and such, and notice that Oakland's radar — a large part of the Bay Area's approach system — is NOTAMED OTS, and with it out, Oakland's Class C surface area and associated services are temporarily gone. No big deal — this has happened before when I've been flying and it's not going to make any difference to the flight.

Then doing the rental paperwork at Oakland Flyers, we hear that 051's G1000 has some sort of problem that supposedly makes the database out of date, and that cross-filling from the PFD to the MFD also isn't working. OK, I guess we can do this flight VFR with practice approaches then… (I prefer to do things on a real IFR flight plan just to keep current on things like copying clearances, etc.).

Out at the hangar I notice the lock's missing from the hangar door and point it out to Evan, thinking "that's careless…". Then I open the door and burst out with a surprised "Where's the bloody plane?!" The entire hangar is empty. OK, time to actually look at the book — sure enough, in tiny characters on the front where it used to say "Hangar XYZ" it now says "Hangar ABC" — RTFM, I guess, but no one at the club mentioned it'd been re-hangared a few hundred metres away — and it'd been in the original hangar only a couple of days ago when the original flight was scrubbed. No big deal, I guess.

We pre-flight 051, then get in and start up. Three things are immediately obvious: 1) the database is actually up to date — maybe we can do this on the IFR flight plan, after all; 2) Yes, there's some sort of weird PFD / MFD cross-fill problem, which is going to make things irritating but not too difficult with the two of us paying attention (I wouldn't want to do this in single pilot IMC); and, 3) The standby (mechanical) AI seems to have lost its tiny mind and is revolving and swirling like something from The Exorcist (a movie I've never actually seen, but never mind). The standby AI doesn't settle down after a couple of minutes, and since it's quite distracting, Evan covers it with one of his inop covers. We decide we'll depart VFR and just do the approaches as practice approaches — after all, this is just an IFR currency flight, and the weather is clear VMC as far as Stockton, at least.

By the time we've done the runup off 27R the standby AI's settled down, but we still depart VFR. We head for the Central Valley — and that lovely weather turns out to be bumpy. Not quite continuously, and not severely, but it's a factor in almost every approach and cruise segment for the next two hours, and it's the sort of bumpiness that anyone not a pilot would probably find tedious at best, sick-making at worst.

We get the practice approaches set up with a bit of extra programming (Evan just replicates on the MFD what I'm doing on the PFD), and after plodding on bumpily across the Central Valley we do the RNAV 29R into Stockton with full pilot nav, including the hold-in-lieu at OXJEF. The bumpiness is close to too much for the autopilot, but I sadistically keep it engaged for most of the way to see how well it could cope. Not very well, is the answer, and after the coyote thing and a smooth touch-and-go, we go missed back to NorCal approach. I re-engage the autopilot at about 2,000' in what I think is the correct altitude mode, but the bloody thing just keeps the plane climbing; at 2,200' I cut the autopilot out of the equation and start diving back to the assigned 2,000'. I can sort the issue out later… and, after setting up the next approach that's what I do.

This time around on the same approach it's just too bumpy for the autopilot, and I hand-fly the entire approach, a little roughly, but quite a lot better than the autopilot last time. However, this time my landing is really pretty bad — I blame the sudden windshear just as I'm flaring, but whatever it was, we bounced, and I had to use the throttle to ensure we didn't land hard on the second bounce. Not too bad, but a lesson in wariness. And, once again, on the missed I engage the autopilot in what I think is straight and level mode at 2,000' — and it starts climbing. This time I'm more on top of it and we don't bust altitude too badly, but I'm always a bit aghast at how much trouble I have with this unit in altitude mode (it's all me, the unit itself is just fine). Just to add to the overall feeling that not everything is going as well as I'd like, I keep making minor radio errors (missing calls, not understanding the controller, etc.) for the approaches (it seems to clear up as I get closer to Oakland for some reason).

We request, set up for, and get the ILS 29R with vectors, and things go better with the autopilot this time around, but I watch it like a hawk. And this time around my landing is OK, considering the windshear, but there are still no coyotes to be seen, which kinda disappoints me. We go missed as published and I let the autopilot drive us around the hold at ORANG, just because I really enjoy seeing this particular chore automated (and I am, after all, an engineer who delights in technology like this). After once around the hold we depart with VFR flight following back to Oakland.

This part — the VFR flight back to Oakland — is not without its hazards either. For much of the past 45 minutes we've been hearing parachute drop zone announcements (there were several drop zones NOTAMED for today in the area), and we have to be careful and listen to the controller to ensure we don't get caught in anything bad. There's a drop in progress over Lodi, but that's quite some distance north of us, and another supposedly closer in to Stockton, but nothing was ever announced on air as we flew by. And just as we're getting towards Byron (C83), Evan suddenly points out a flight of four or five military jets slightly below us, maybe two miles away, going rather quickly towards Sacramento. A nice sight (they're maneuvering in not-quite-aerobatic ways as they shoot past), but I kinda wish NorCal had pointed them out to us as traffic. Oh well. We watch them disappear rapidly northwards, hoping they don't reappear back in our direction.

All through the flight from just after takeoff one of the G1000's MFD fuel gauges intermittently shows a red 'X' for a few seconds and then reverts to normal; it seemed to stop doing this around Stockton, but it's back again. Oh well, another thing to squawk when we get on the ground.

We set up for and request the RNAV Y 27L back in to Oakland, and potter on towards JUPAP as requested. Nothing much to report about this segment except it's still bumpy, but (predictably) the air's clear and cool. We get vectored for the approach, and at first I let the autopilot handle things, which it seems to be doing OK. But just past the final approach fix, even though it's captured the (LPV) glideslope, it just can't seem to cope with the near continual windshear, and we end up two dots above the glideslope. Time to take over, I think, and I disengage the autopilot and hand fly the rest of the approach successfully and fairly smoothly down to about 400'. It's a weird sort of day when my hand flying is smoother and more accurate than the autopilot, but there you are. But however smooth the approach, I land badly on 27L. Once again, due, I think, to low-level windshear in the flare, I billow down the runway and have to salvage things with some throttle work. Nothing terrible, but a bit of a blow to the ego, I guess.

Back in Oakland Flyers I do the paperwork and notice that the tach time we took from the MFD is way out. I suspect that maybe the G1000 cross-fill problem has affected the tach time (or not — I just don't know) on the MFD without affecting the rest of the engine instruments. Whatever the cause, it means I have to trudge back out through the security gate to the hangar and back again just to see what the tach time is on the PFD; it's much more believable this time (and yes, I check what the MFD displays — we didn't misread it, it was maybe ten hours out). Back in the clubhouse again, my pen stops working as I'm trying to finish the paperwork. It's been that sort of day, really….

What Could Possiblie Go Wrong? (Rust Never Sleeps, Part 2)

Things are going quite well for me under the cone of stupidity on Yet Another Quest For Currency (the aviation sort, not the hard stuff I can use to get boutique bagels and exotic coffees down at the neighbourhood local). John's in the right seat, watching carefully. We've left Stockton (KSCK) behind after having successfully completed several approaches and landings in the still darkness of a Central Valley autumn night.

John's given me his iPad to play with on the next approach, the ILS into Livermore (KLVK), and it's working nicely with my new Dual XGPS150 GPS unit sitting up there on the dash. I've also got my little iPhone in my hand as I watch how well it works without coupling to the Dual (it'll only couple to one Bluetooth device at a time). Well, it does quite well, actually, and I feel pretty impressed with all the technology surrounding me (but it's really me doing all the real work, telling the G1000 what to do and such — well, that's what I like to tell myself). Suddenly I can't help it: "Siri! Set up a practice ILS 25R approach into Livermore from our current position, please!" Siri does nothing — either she's not listening, or she's as confused by my accent as everyone else is, and I revert to doing it all myself again. Oh well; I'm guessing we'll have to wait a few years before that'll work (and if you don't know who Siri is, this blog's probably not for you :-)).

Anyway, that's not the "what" that could possiblie go wrong in the title (diehard Simpsons fans will probably get the reference). What went wrong was the string of mostly weather-related cancellations and postponements that lead up to this flight — all sparked off (I'm absolutely certain) by my ending an earlier email to John with the words "what could possibly go wrong?". I'm just not the sharpest tool in the tool shed: one day I'll learn not to tempt fate like that. So what did go wrong? An example: we had to cancel our previous attempt when a howling gale caused by an unusually-strong high pressure system sitting somewhere over Utah was driving high-speed rivers of wind down over the Sierras and the coastal ranges towards the Pacific. As locals will know, this makes for unpleasant turbulence pretty much anywhere within sight of a mountain range or even hills (and there are ranges everywhere in California…). I looked up the winds on DUATS: 65 knots at 6,000 — a nice tailwind if you can get it, but we'd be going straight into it for at least half the flight, and that's a fierce source of rotors and bumps even over the Valley or in the lee of the Berkeley Hills (especially in the lee of the Berkeley Hills).

I might have been up for it even given all that, but my knee's still a bit iffy and I thought the better of it. So here we are, several non-existent flights later (I'll spare you the details of the other delays and cancellations).

The ILS into Livermore goes reasonably well — I'm getting better with practice — and we do the low approach past the tower over 25R when I come out from under the hood. Woohoo! (I always feel like doing a victory roll on low approaches — a hangover from my aerobatics days — but I know better, honestly). And now it's off to Oakland for the ILS 27R approach back home….

This goes fairly well as well, but we get the slam onto the localiser from somewhere between GROVE and UPACI intersections, and I can barely bring myself to program in the required 1,100 fpm descent to capture the glideslope from below. We just don't get low enough, and in the end I take over and hand fly the last few minutes (which kinda negated the point of this approach — it was supposed to be about managing the G1000. Oh well). The landing on 27R goes well (they've all gone well tonight, which was pleasing after all the time off from flying) and we taxi back to the Port-A-Ports.

* * *

Yes, I'm current again — but still a bit rusty. There's a touch of the old death grip creeping back to haunt my control of the plane, and my radio work definitely needs some polishing, but overall, this flight went basically as planned, despite all the preceding hiccups. We started it by taking off from Oakland's runway 9L, a bit of a treat for me: the winds nearly always favour one of the 27's, and I hadn't taken off from 9L for literally years, and we ended it an hour or two later with me feeling fairly OK about my flying, especially the landings and the big-picture procedural issues on the approaches. My hand flying on one of the approaches was definitely a little agricultural, but hey, I can't do everything.

* * *

Earlier, while prepping the flight I do due diligence with DUATS and notice the following NOTAM:

!FDC 1/1954 SCK FI/T IAP STOCKTON METROPOLITAN, STOCKTON, CA.
   ILS OR LOC RWY 29R, AMDT 19...
     S-LOC 29R MINIMUMS NA.
     CIRCLING MINIMUMS NA.
     JOTLY LOM HOLDING NA.
     TERMINAL ROUTES ORANG TO JOTLY LOM, LIN VORTAC TO JOTLY LOM,
     WRAPS TO JOTLY LOM, ECA VOR/DME TO JOTLY LOM NA.
     MSA WITHIN 25 NM OF ECA VOR/DME, 070-160 3000, 160-250 5300,
     250-340 3400, 340-070 4200.
     JOTLY LOM DECOMMISSIONED.

Hmmm, I think — what the hell does that mean? It basically seems to mean there's simply no way to do my fave ILS for currency at the moment. But why make the approach unavailable that way? Why not just NOTAM the entire approach N/A? I ask John — he's not entirely sure what the FAA's thinking with this either, so I decree that we'll just do the RNAV 29R approach out there instead. It's all flying to me….

Tree Of Life

We're pottering along in Cessna 051, 2,000' above the darkness of a quiet Central Valley night, on a wide downwind vector for the ILS back into Stockton (KSCK). A few minutes earlier we'd gone missed as planned on the KSCK RNAV 29R approach after a nice stop-and-go, the first of a projected four or five approaches to get me IFR current (and proficient) again. We're discussing John's new bluetooth GPS for his iPad (which he's brought along and is showing me how it works in real-world use), waiting for another vector back towards the localiser. We're also (of course) idly listening with one ear to NorCal Approach's frequency, when — out of nowhere — a voice we connect with a certain semi-local bizjet we'd heard earlier on-frequency pipes up and asks in a rather faux innocent tone something like "so when did you guys [ATC] start saying 'tree' and 'fife' for 'three' and 'five'? It sounds really funny to me" (yes, our controller's been saying 'tree' and 'fife' quite by the book all the while).

Argh, I think. Not this again — they've been saying it that way for years now, especially the newer controllers (and it's even in the AIM) — where's this guy been all that time? Or is he just trying to be funny? The controller responds with a fairly non-committal answer along the lines of "that's the way we're supposed to say it now", and the other guy persists with several lines about how silly it sounds; and, several times, we can hear the controller's increasingly "I-don't-really-want-to-be-having-this-conversation" responses coming back between the calls to other planes for real stuff.

Just as I'm snidely remarking to John that hell, that's the way we in large parts of UnAmerica have been saying it for a while now, and what's with all the Good Ole Boys in America coming out of the woodwork and complaining about it, a new voice pipes up on frequency — a (much) older no-nonsense voice (I'd almost swear it's Lou except the voice isn't quite his) that simply points out that 'tree' has been the way it's been for a long time, and what's to complain about? (and, implicitly, "get with the program!"). So much for stereotypes. I'm tempted to add my own accented voice to the conversation (something along the lines of "bloody whingeing yanks!" :-)), but, as I've complained about in the past, my accent really marks me out on air, and I'd never get away with it ("Cessna 051! That was you, wasn't it?! Expect to hold at JOTLY for 35 minutes!"). And, in any case, as John reminds me, "American" is actually now one of my (several) nationalities, so who am I to complain?! Humph. The controller tries to end it all by handing the original complainer off to the next sector, but the pilot on the other end can't help getting in a parting shot that (I guess) was intended to be humorous, but just comes off as lamely missing the point (again). We plod on in the darkness over the Central Valley, marveling at the state of radiophonic discourse and aviation standards-keeping, waiting for that vector….

A few seconds later it comes, and we head off towards the localiser, my brain back in approach mode. I hand fly the ILS (under the hood) steadily down to 150' with John watching like a hawk in the right seat — not bad for a flight that started out with me feeling very rusty indeed. It surely helps that the wind's fairly constant (a quartering tailwind right down to the ground) and the traffic's nonexistent.

We go missed again, and back with NorCal we ask for and get full pilot nav for the Tracy (KTCY) RNAV 26 approach, starting at our current position with OMWAP as the IAF, and including the course reversal hold. This goes well (with the G1000 steering us around the hold), and John gets to do the landing after I look up at the MDA. About the only interesting thing about this approach is the use of waypoint names like "OMWAP" or "IVABE"; John and I argue about how the latter should be pronounced, and the controller avoids using it at all. Back with NorCal on the missed we head back to Oakland for the ILS 27R.

Somewhere close to GROVE for the localiser we start hearing a non-stop string of approach clearances, vectors for traffic, and ATC instructions on air, and just as I'm remarking to John about how it's suddenly a zoo out there, we get a square 360 for traffic on the ILS — traffic we can see on the G1000's TIS as several white diamonds converging on the localiser in an interesting-looking cluster. Par for the course for Oakland, I guess, but I enjoy this sort of thing, and I'm usually happy to go a long way out of my way for faster traffic and to make the controller's life easier (it's all flying, dammit, and it's always good IFR flying exercise). By the time we're back on vectors for the localiser and cleared for the ILS, there's only one plane competing with us for the approach, an Amflight Navajo bearing down on us from a dozen miles behind us Out There somewhere beyond our TIS display.

The ILS is an approach I know like the back of my hand — an approach I nearly botch in the last few moments as I'm hand-flying under the cone of stupidity and John casually suggests setting the flaps in a specific way that combines a momentary throttling back with flap deployment (nothing tricky, just done in a way I don't normally do). I make the mistake of thinking about it and then doing it, which puts me dot or two to the left and a dot low on the ILS before I recover — I should have just ignored John or done it by instinct, I guess. Still, no one dies as a result of my over-thinking, and the landing is a success.

We taxi to the Business Jet Center for fuel, and discover it's a bit of a zoo as well — a CHP helicopter, a King Air, and an anonymous and expensive Gulfstream, or Challenger or whatever are all departing from in front of Business Jets simultaneously, and we make damn sure Cessna 051's out of the way and has its control locks in. I stand outside on the ramp for a minute in the thick of things watching the CHP helicopter depart sideways to avoid hitting 051 and the business jet, and the noise of all three aircraft in the night is astonishing — the sort of thing you usually either love or hate (you can probably guess which applies to me).

After hangaring 051 John signs my logbook, and I'm IFR current again. What matters at least as much, of course, is whether I'm proficient as well, but after a bunch of hand-flown approaches and similar exercises, I don't feel too bad.