DI accuracy

[drauk]

Two questions, the first purely theoretical and just as an aside really, the second the practical one:

1) Is "precessing" the only practical way in which a gyro driven DI goes "off" from the compass? That is to say, if the DI stops matching the compass, assuming nothing catastrophic has happened to any of the instruments, is the cause always "precessing"?

2) What constitutes a DI with a problem? Is there some tolerance that a decent DI should be able to maintain and if so how is it defined? Qualitatively you might say "Oh, the DI seems a bit off - I have to reset it every few minutes", but is there a quantative definition? For example, something along the lines of "no more than 1 degree after 10 minutes of level flight" or "no more than 5 degrees after 5 360deg turns at 45deg bank angle". Or is it measured in some more technical way, rather than by practical flying?

[FlyingForFun]

There are four things which cause a DI to wander:

- Imperfections in the DI. These will vary from one DI to another, so there's no way you can take them into account. To answer your second question, I would guess that the answer is that a DI is bad if it has enough imperfections to cause it to wander noticeably. I've never seen any actual limits.

- Earth rate. If you put a DI at the north pole, indicating that the Grenwich Meridian is "north" (for the sake of argument), and leave it stationary, 12 hours later it will be pointing in the opposite direction, indicating that the Grenwich Meridian is now "south". This is because the earth is rotating, but the compass remains in the same place. Conversly, if you place a DI on the equator, it will not wander at all - north is still in the same direction relative to space, even though the earth is spinning. Anywhere in between, the wander will be at a rate that depends on the latitude.

- Latitude nut. Because the earth-rate wander is so predictable, it can be corrected. A DI is fitted with a "latitude nut". This nut is adjusted according to the latitude at which the DI will be used, and it causes the compass to "wander" in the opposite direction to the earth rate wander, thus cancelling it out. At least, that's the theory. If it's set incorrectly, it won't work. If you have a DI which constantly wanders in one particular direction, then this may be the cause of the problem.

- Transport wander. Similar to earth rate wander, except that it's caused by the DI moving across the surface of the earth (presumably due to being in a movign aircraft) as opposed to the earth moving under the DI. The amount by which it will wander over a journey depends on how far east or west you move, as well as the average latitude throughout the journey. The rate at which it wanders depends on your east/west groundspeed as well as the average latitude. There is clearly no way of correcting this, since the manufacturer can't tell in advance which direction you're going to be moving in or how fast.

This is all required knowledge for ATPL exams (as well as the formulae for the last 3 types of wander), but at PPL level it is all simplified down into "precessing" - and that's probably all you ever need to know for any practical purposes.

Hope that helps!

FFF
-------------

[rustle]

Good answer FFF. You just done your exams or something? ;)

[iainpoll]

OK, this is actually quite complicated and covered in detail during ATPL exams.

Firstly DI 'pressesion' as we know it is actually a gash term for gyroscopic wander. This is broken down into two catagories, 'real' wander, and 'apparent' wander. Real wander is caused by imperfections in the instrument, i.e. the gyroscope. Your second point would relate to real wander, I dont have any figures but real wander increases with instrument age, wear and tear, stuff like that.

Apparent wander is caused by a mixture of 'Earth rotation', 'Transport Wander', and errors/corrections induced by somthing called a 'Latitude nut'!

There is a lot of maths involved here, and as my ATPL manuals are now in the loft, and I'm sure you don't want boring to tears, I'll leave it to someone far more qualified than me to explain in more detail!

Happy landings :ok:


FFF got his answer in first, and, as usual it is far better than mine.
:cool:

BTW, Baz at Bristol, with Nanuk of the north, and Bongo of the Congo, really helped me nail this.......

:ok:

[drauk]

Thanks FFF. It certainly helped answer my first question, which was really about the fact that whenever I hear people talk about the DI going "off", if they're trying to sound smart, they say it's precessing, and I wondered if that was the only explanation.

Unfortunately the important bit was the second part of my question, which hopefully an engineer may be able to answer. If an engineer refits a serviced DI to a plane, how does s/he know it's working? I suspect the answer is "fly around for a bit and see if it still agrees with the compass", but what I'd like to know is, what's "a bit"?

[Evo]

:D Rustle

I've seen DIs go U/S twice now - once was with an instructor on a PPL navex and the other time was with Genghis en-route to Duxford. The first failure nearly took me into Solent's airspace and the second nearly took us to Luton. :eek:

Both times we were seeing 30 degree errors develop within a few minutes (or less - I think we developed something like a 20 degree error in the turn from Base to Final in the Popham circuit). When that happens it's bl**dy easy to follow it to somewhere you shouldn't go.

[drauk]

Evo, you're doing your IMC rating at the moment, right? Try flying with a DI that has to be reset all the time when doing a procedural approach. Of course, you give up and use the compass, which is not much fun.

FFF, another thing: Doesn't doing your ATPL exams rather shoot yourself in the foot as far as your moniker is concerned?!

[Evo]

Evo, you're doing your IMC rating at the moment, right? Try flying with a DI that has to be reset all the time when doing a procedural approach. Of course, you give up and use the compass, which is not much fun.


Yeah I am, and yeah it isn't - but I'm rubbish at most things partial/limited panel at the moment :(

What worries me is how i'd even spot a DI failure in IMC. Both times before I spotted something was wrong because the view out of the window wasn't quite 'right' ("What's that fuggin' great airport?" :ooh: ). I guess in IMC i'd notice something was wrong when the navaids started to behave oddly (maybe?) but whether I'd figure it all out I dont know. A gradual AI failure is even more worrying.

[drauk]

Evo, how you'd spot a gradual DI failure is a whole other discussion, but actually I reckon it is one of the easier failures to spot for three reasons: (i) as you said, you'll find your ADF and VOR instruments seemingly not behaving as you'd hope (ii) more obviously, it's easy to check against the compass and (iii) most obviously of all (if under their control) ATC will keep asking you to "Say heading", because you're not flying what they last gave you!

[Andy_R]

Sorry for answering for FFF, but can learning not be fun when it is related to something you enjoy doing???

[FlyingForFun]

Absolutely right, cloud. Although I doubt anyone would claim that ATPL Air Law is "fun" :eek:

Although many commercial pilots view "little aeroplane" as a necessary hurdle to getting to fly a jet, I do know quite a few who own shares in aircraft which they fly when they're not working. If you can call it "working" of course! And that's definitely the category which I'll be in!

Anyway, too many people know me as FFF to be able to change my name. :)

FFF
-------------

[IO540-C4D5D]

Most DI drift I have encountered was at least an order of magnitude above anything induced by the earth's rotation at the UK latitude, so perhaps for the average renter the practical answer is "lack of maintenance". I don't think an aircraft even needs a working DI to be VFR-legal.

Having been flying with a slaved HSI for a year I would dread to go back...

[iainpoll]

Slaved HSI - now your talking!:cool: :cool: :cool:

FFF, I must admit I did think of the little aeroplane as a hurdle - then I suddenly realised how much fun it was. I actually took a guy flying yesterday CPL/IR all the rest, he loved it, in fact his comment was 'wow - forgotten how much fun this can be'!

ihp

[drauk]

In respect of the joke about FFF, my point was that some people fly for reasons other than fun, e.g. commercially. I didn't suggest (nor imagine) that some commercial fliers don't find flying fun!

Since nobody has been able to answer question 2 then I'm beginning to assume there is no definition. Any engineers around to refute that theory?

[Flying Tooth Driller]

The FAA says that a precession error of no more than 3 degrees in 15 minutes is acceptable for normal operations.

Sounds sensible. I wish all the planes I've flown had DIs that good...............
Well, a few have :-)

[drauk]

Thanks FTD. That is exactly the kind of info I was after.

[FlyingForFun]

I've only ever flown two aircraft with unservicable DIs.

My Europa's vacuum pump broke, hence no DI at all. Fixed now, although I haven't had a chance to fly it since, but I was just starting to get the hang of compass turns.

The other aircraft was more interesting - a Super Cub which I was renting. I tend not to look at the instruments too much when I fly, especially when the viz is good. And this particular Cub was in Arizona, where the viz was always good. So I'd set my heading, pick a point on the horizon and fly towards it. Every 5 minutes or so I'd check my DI, and most of the time I'd be right on course. But occassionally the DI would have "wandered" well off course. Not by a few degrees, but by anything from 10 degrees to over 150 degrees! Re-set it against the compass, and found that I was still on track. Didn't really cause me any difficulty, but it was very strage nonetheless.

Then, one day, I caught it. I made one of my regular 5-minute-interval glances at the instruments. And there was the DI, spinning round and round! Then it just stopped, on whatever random heading it happened to be on! I would guess that over the course of the 2 months I was flying the aircraft, I saw it do this maybe 4 or 5 times. Each time, it would be maintaining heading perfectly. Then it would start spinning, spin right around several times in the space of a half a second or so, then stop. And it happened dozens more times when I wasn't watching it. Certainly didn't take any guidelines like "3 degrees in 15 minutes" to tell that it was sh@gged!

Anyway, as IO540-C4D5D says, an aircraft doesn't require a working DI to fly VFR. Many aircraft don't have a DI at all, let alone a working one... I would guess that any guidelines or rules could only apply to IFR operations.

FFF
-------------

[foghorn]

How much does a slaved HSI cost these days? I would love to put on ein our aircraft, but I doubt that the non-IR / IMC members of the group would go along with me...

[Keef]

Slaved HSI is fabulous but the last time I looked, it was in the region of £5k. With a few bells and whistles, yet more than that.

We had one on the late lamented G-BDKV. The big snag was that after the "initialisation" period (ie working out which way is North and pointing there), it turned down the speed of change to something like rate 4 maximum.

I once landed in nil wind, did a 180 on the runway to take off in the other direction, and it gave up. The examiner (this was a renewal checkride) was delighted since it saved him having to cover it up. He spent the next ten minutes trying to get it to synch again (and failed).

[drauk]

FFF, my questions stem from flying IFR. I bet "catching" the DI spinning round was quite amusing, given that you were flying without the need for it. I've not been doing it long enough to experience anything like that. Anyway, the DI in question is not nearly so obvious, it just wanders off and I was looking for a way to quantify it.

[FlyingForFun]

Ah, but drauk, when did I ever claim that any of my posts would actually be useful? :D

[knobbygb]

Just out of interest, how does a slaved HSI work? Does it have its own internal mag. compass?

I ask because some of the Warriors I fly have these and some don't. I always turn them to manual because I 'don't trust' the automatic ones. I'd rather align it with the compass myself, as per my training. Perhaps understanding how it works would help me overcome this. I asked an instructor once but he didn't know.

[Evo]

I think an HSI uses a magnetometer as a reference? (edit: a quick search (http://www.google.co.uk/search?q=cache:tL2qgk3IUjUC:www.aviationshop.com.au/avfacts/editorial/chsi/index.asp+%22horizontal+situation+indicator%22+&hl=en&ie=UTF-8) refers to a "flux valve" - not sure if this is the same thing or not :confused: )

[FlyingForFun]

Knobby,

It basically uses some very sensitive coils, usually placed in the wing-tip because that's where they get the least interferance, to detect the earth's magnetic field. The detector is held in a "hook joint" which enables it to remain vertical as the aircraft banks and pitches.

The DI is connected to a similar set of coils, which are connected to the detector in such a way that a current is produced if the two are not aligned. This current is used to turn the DI. When it aligned with the detector unit, no current is generated and so the DI remains stationary. As it drifts (which all DIs will do, as we've discussed) a current is produced and it synchronises itself again.

The DI is normally aligned at a rate of 2 degrees per second. However, there is a "fast" mode for initial alignment when it is first turned on (or after it topples). I also recall something about the detecting unit being ignored at high angles of bank, and it basically becomes a standard DI, which then re-syncronises itself with the detecting unit when the aircraft is rolled level again - but I can't find anything about that in my notes.

Hope that helps!

FFF
-----------

[IO540-C4D5D]

Keef

There was something wrong with that slaved HSI/DO you had if it could not follow heading changes quickly.

Normally these are gyro instruments so they respond like any gyro DI to heading changes, i.e. immediately. The slaving part of it merely updates the heading very slowly (a few degrees per minute or so) from a fluxgate magnetometer (a crude form of solid state compass) mounted usually in a wingtip.

I think the total cost of a slaved HSI system is nearer to 10k than 5k. But once you have a slaved HSI you can add a slaved RMI for a mere 4k :O The same heading signal can also be used to rotate a gps and a stormscope display, etc.

This stuff is ludicrously expensive for what is actually in there.

[Tinstaafl]

Easiest way to think of a HSI is as a 'standard' DI that gets given an automatic slight 'nudge' to keep it aligned. The nudge coming from an electronic compass mounted somewhere remote from interference.

Add an automatic fast nudge to get the thing aligned in the first place & a means to disconnect the nudge.

Once disconnected it's very like a standard DI, only you use an electric switch to set alignment instead of a mechanical knob.

Other bells & whistles include an integral VOR display.

[Keef]

IO540 - it must've been worn out. It would certainly align itself pretty quickly after "power up", and was fine in normal use. But a quick 180 on the ground after "spin up" would "topple" it every time.

We didn't worry too much, because we didn't do a lot of that. We didn't have that particular Arrow for long anyway - the engine blew up over the Solent and the insurance wrote it off.

It's certainly a nice bit of kit, but I think I'd go for a GNS430 before I'd go for one of those.

Don't reckon I'll be in that league again, being a poor retired fella with a part-time job that doesn't pay anyway ;-)

[tmmorris]

Which is the expensive bit? The plane I usually fly already has a 'remote indicating compass' (rather an old presentation, vertical card like a DI but the needle revolves instead of the card, so you 'fly the needle' to take up a new heading. It suffers from dip but does seem more accurate than the one on the coaming. Is this using a flux-gate? If so, is the HSI or RMI then quite cheap to add?

Tim

[IO540-C4D5D]

tmmorris

I am responding without reference to current price lists, but to install a slaved HSI you need

- the HSI itself, e.g. KI525A
- the fluxgate magnetometer assembly, e.g. KM112
- remote directional gyro, e.g. KG102A
- various bits e.g. auto/manual switch/meter (KA51B)
- installation of the whole thing
- possible approvals from the CAA

You can contact your local avionics show for a quote but you will need a coffee beforehand, and another one afterwards :O

This stuff is mostly early 1980s but is still current fit in higher end brand new planes.

I don't believe you will get any change from £10k. There are far cheaper alternatives available for U.S. experimental category planes...