Compass Turns - IMC

[PoppyPilot]

Hi, can anyone give me some useful tips for doing accurate compass turns on a partial panel. I have just started my IMC so in fact any uselful 'rules of thumb' or tips related to helping me through this rating would be really appreciated. Thanks

[Contacttower]

I was told by my instructor not to do compass turns...just do timed ones instead. 3 degrees/second = rate one. They can be done though and this is what I was taught:

If you undershoot when turning onto northerly headings by 30 degrees and overshoot the same for southerly ones then they usually work out OK. Just remember to always give the compass time to settle and also that when you look at a W compass you have to turn away from the digits to get to them.

[tmmorris]

Agree timed turns, followed by let the compass settle and check, are the way forwards, and I was only taught this method - precisely to avoid the dip problems of compass turns. Though in the event of electrical failure (hence failure of the turn indicator/coordinator) you'd have to know your bank angle for a rate 1 turn (calculated from speed, of course, though I forget the formula because I know what it looks like on the ac I usually fly!)

Tim

[rustle]

Though in the event of electrical failure (hence failure of the turn indicator/coordinator) you'd have to know your bank angle for a rate 1 turn (calculated from speed, of course, though I forget the formula because I know what it looks like on the ac I usually fly!)

Tim

TAS/10 + 7 works at these low speeds :8

[Contacttower]

17.5 degrees in the PA28 at 100knots I believe. :8

[tmmorris]

That's the one!

Tim

[waldopepper42]

For those, like me, who need acronyms in those high workload moments.....

"UNOS"

Undershoot North, Overshoot South!

[maxdrypower]

Bit odd for me to say this as I have just done an IMC but why dont you ask the guy sat in the right hand seat ? I assume you trust him and value his input or you wouldnt be paying out a shxt load of money . Thats what he is there for surely

[DFC]

1. Put your heading on the ADF Card. It helps you to remember what you should see on the compass. It also helps orientate you so that you know if xxx heading is a left or right turn.

2. Using timed turns is easier than reference to the compass but a combinaion of both is best.

3. Look at the ADF. Heading Digits are placed every 30deg.

Rate 1 is 3 deg/ second.

Thus when asked to turn onto xxx heading complete the following;

1. Confirm that the compass heading is still what it should be. If not, reset the ADF card.

2. Count the number of 30deg segments and multiply by 10 seconds.
3. Count the remaining 10 deg segments and multiply by 3 seconds.
Add both together to get timing.

eg heading 290 and asked to turn onto 180

look at card you have 3 30degree segments plus 2 10degree segments. = (3*10) + (2*3) = 33 seconds.

It is a timed turn so Start time then start turn. When time is up, stop turn.

Remember that the turn "co-ordinator" is calibrated for a specific speed and only shows rate 1 if you have 1g, ball in the middle and this speed. However, for the IMC rating simply peg rate 1 and fly in balance.

Having leveled the wings and re-established straight and level flight, check compass. For adjustments of up to 20 degree simply count 1 banana, 2 banana etc For 20 deg plus do another timed turn.

For the compass turn remember to undershoot the nearest pole. Thus in the northern hemisphere undershoot north, overshoot south, no error east or west.

Keep bank angles to 15deg or less to avoid locking the compass card.

The error is about 30degrees at north or south and 0 at East or West. That means if turning onto a heading of north, you stop the turn 30 degrees before you get there. If turning onto south you stop the turn 30 degrees after you pass south.

If turning onto east or west then you stop the turn on the heading.

For points in between simply judge how much to use i.e. a turn onto 045 degrees is under shoot by roughly 15 degrees. i.e. stop tuen 15 degrees before you get there.

As I said previously, the best thing to do is a combination of the two. i.e. Doing a timed turn but if passing through east or west during the turn check that this happens at about the right time.

The most important thing is to be able to fly a constant heading when not turning. It is wandering all over the sky during straight and level that fails most instrument pilots. The turns are expected to firstly be in the correct direction, secondly get you into the ball park heading and thirdly to be finally corrected to the chosen heading which is then maintained.

Hope that helps.

Regards,

DFC

[Contacttower]

Remember that the turn "co-ordinator" is calibrated for a specific speed and only shows rate 1 if you have 1g, ball in the middle and this speed. However, for the IMC rating simply peg rate 1 and fly in balance.



Are you sure about that DFC? Reason I question it is when I'm doing climbing rate one turns at 80knots and the co-ordinator is aligned for rate one the AH shows about 12 degrees of bank. Repeat at 100knots and it shows about 17 degrees (like it should). From this observation it would appear that the turn co-ordinator is indeed aware of the speed of the aircraft.

[Kliperoo]

DFC is very correct.
If you can find a picture of the internal workings of the Turn Coordinator, all the better. If not imagine a tyre sitting up with the gimbal system attached to the holes in the wheel and springs below this. Crude I know :rolleyes:
OK... the gyro spins (say away from you) and an aircraft enters a left turn...the turn is felt as a "push" on the gyro (thinking of the tyre its felt at the 3 o'clock position - now looking at the tyre from the side) through precession (because the gyro is spinning) the system feels the push at the top (12 o'clock). This pushes the gyro (tyre) down and to the right, and through gears the needle goes left.
I hope that made sense.
Now, as the aircraft speed is incresed it takes more initial push for the precession to show the same amount turn on the dial. The only way for you to increase the precession is to increase the angle of bank (which is what you have to do with an increased speed for rate 1).
So it does not sense the speed, but rather the new force and resultant precession as a result of the increase angle of bank for the speed.

[Contacttower]

So it does not sense the speed, but rather the new force and resultant precession as a result of the increase angle of bank for the speed.

But that does suggest that the instrument works throughout the speed range, and I took DFC's post (perhaps wrongly) to mean that it didn't.

[Kliperoo]

I see where youre coming from now...

yes you could make that argument that it does work "through a range of speeds" with the fact that it will show rate of turn at varying speeds - and good enough to use in IMC...but if it showed up as an answer to a question on an exam or during a flight test, Id be weary at picking it without considering all the other options. :)

It will only show the proper rate at one airspeed though, as DFC stated. However, for say a C152 or PA28 the speeds they go are slow enough to ignore the error - with the error also depending on mechanical errors like friction, wear and tear, et al.

Remember too that (and I assume this is the same with the JAR's - correct me if Im wrong) when IFR your turns should always be at rate 1 or 30* of bank, which ever is the least, so above 230 knots you will always be 30* and below it will be rate 1.

But if you can get to 230 in 152 without becoming an unintentional parachutist...youre a better pilot than I. :}

[Henry Hallam]

" It will only show the proper rate at one airspeed though, as DFC stated."

This is really not correct. A properly operating turn coordinator directly measures *rate of turn* through gyroscopic techniques. It is independent of airspeed. If you have the aircraft stationary on the ground, put it on a turntable and turn it at standard rate, 3 deg/sec, the turn coordinator (or turn and slip) will show a Rate 1 turn.

[IO540]

Forget the compass during the turn - it reads rubbish.

Only a fresh graduate of the CAA/JAA "must sort the men from the sheep" exam system will remember which way the compass errors go, and few people will be able to remember this nonsense under any sort of workload in a real partial panel situation.

Times turns are the only way - 3 degrees per second. Enter and exit the turn with a rapid roll to/from wings level, to minimise errors.

The turn coordinator reads Rate 1 irrespective of airspeed !!

[Henry Hallam]

"The turn coordinator reads Rate 1 irrespective of airspeed !!"

Right - it is the whole point of a turn coordinator after all!

[rsuggitt]

My method (avoids hard caculations).... set the ADF card as you've described. Pretend it's the face of a clock (analogue !) . When you have a second hand on a clock it sweeps past each 'major' division every 5 seconds. So, read round ADF card to from 'north' to your required heading and count seconds for each major division, 5 seconds, ten seconds, 15 seconds, etc. [Easier done than described]. Then double it. That's the time required to turn at rate1.

[Contacttower]

My method (avoids hard caculations).... set the ADF card as you've described. Pretend it's the face of a clock (analogue !) . When you have a second hand on a clock it sweeps past each 'major' division every 5 seconds. So, read round ADF card to from 'north' to your required heading and count seconds for each major division, 5 seconds, ten seconds, 15 seconds, etc. [Easier done than described]. Then double it. That's the time required to turn at rate1.


That's exactly mine as well...great minds and all that ;).

[Nathan Parker]

Remember that the turn "co-ordinator" is calibrated for a specific speed and only shows rate 1 if you have 1g, ball in the middle and this speed. However, for the IMC rating simply peg rate 1 and fly in balance.

I disagree with this as well. A TC senses yaw rate; anytime you're getting 3/degrees per sec, it will indicate a standard rate turn.

[Fright Level]

Not speed dependant at all. A rate one turn will produce 3 degrees a second turn. In a minute, you will have turned 180 degrees. The faster you go, the larger the radius of turn, but you would have still only turned 180 degrees. One of the reasons published holds have a max speed rating to keep you in the protected area.

[Contacttower]

TAS/10 + 7 works at these low speeds


Does that not work whatever speed you're doing?

[DFC]

I am surprised indeed.

The turn coordinator is only accurate at 1g. The more G the less accurate the indication.

The higher the airspeed, the more bank is required in order to complete a rate 1 turn.

The more bank in a balanced turn the higher the g loading.

In general this is only theoretical since standard turns are at rate 1 or 25 deg angle of bank whichever requires the less bank.

The reason why we use the simple calculation of (airspeed (in knots)/10 ) +7 to obtain an angle of bank for rate 1 rather than simply select rate 1 on the turn coordinator (when using full pannel) is becuase of the inaccuracies of the turn coordinator.

Use the simple equation and the figures measured by contacttower and you will see the errors.

I would laso like to warn anyone about following the advice of "Enter and exit the turn with a rapid roll to/from wings level, to minimise errors. from IO540

Firstly, the turn coordinator shows roll as well as turn and that is why it is used in basic aircraft for a limited panel back-up i.e. you do not have to wait for a rate of turn to develop before discovering that your wings are not level i.e. any roll will cause movement of the indicator. Now please imaging what the turn coordinator does when you apply a large rolling moment.......how easy is it to smoothly find wings level then? Remember also what your instrument instructor taught you when doing unusual attitude recoveries - nose low and banked.

The timed turn will work well enough no matter how slowly you enter the turn provided that you enter and exit the turn at the same rate

Start the watch, start the roll to acheive rate 1. Let's say that you acheive a steady rate 1 at oh 5 seconds. When the time is up you roll out at the same rate and thus you will be back wings level 5 seconds after the required time. The turn lost while making the entry is compensated for by the extra time at the end.

Finally the single speed is the one where you can turn rate 1 in balanced flight with zero roll moment i.e. ailerons perfectly neutral (think about climbing and descending turns here) and only 1g load factor.

Don't think that I have come across that speed often. :)

Regards,

DFC

[Contacttower]

The turn coordinator is only accurate at 1g. The more G the less accurate the indication.



I see what you mean DFC, but most turns below 30 degrees are pretty much 1g though. Generally speaking the turn co-ordinator is not a very accurate instrument and the little sum mentioned earlier is much more reliable (when you have the AH at least :E).

[tmmorris]

But if the AI/AH is working, chances are the DG will be, so the timed turn won't be necessary!

I guess that if you do as your instructor taught you and keep all turns gentle, you won't exceed 1g by enough to make any difference - how accurately can you hold that rate 1 turn on the TC anyway? It's a bit like flying a PAR, when the talkdown controller (bless her cotton socks) says things like 'turn right one degree, heading 003'...

Tim

[IO540]

Does that not work whatever speed you're doing?

No. The exact formula for the bank angle required for a rate 1 turn involves a bit of trigonometry, and the (TAS/10)+7 is just an approximation which works for small angles, say below 30 degrees.

The usual place to look for aviation formulae is here (http://williams.best.vwh.net/avform.htm).

[Nathan Parker]

The more G the less accurate the indication.

Repeating it doesn't make it true. Do you have any evidence?

The reason why we use the simple calculation of (airspeed (in knots)/10 ) +7 to obtain an angle of bank for rate 1 rather than simply select rate 1 on the turn coordinator (when using full pannel) is becuase of the inaccuracies of the turn coordinator.

The reason we use this approximation is to get an idea of what bank angle will produce a standard rate turn. When you use it, the turn coordinator will indicate a standard rate turn.

[Islander2]

FWIW, this is what is said about turn indicator (not coordinator) errors in my (CAA-flavour) CPL course notes, as produced by the sadly now defunct Professional Pilot Study Centre:

"The rate of turn indicated will only be correct when the aircraft is flying at the true airspeed for which the instrument has been calibrated. Providing that the actual TAS is within 100kts of the calibrated TAS the error in indicated rate of turn should not exceed plus or minus 5% of the actual rate of turn. The instrument over-reads if TAS is higher than the calibrated airspeed."

[Fright Level]

Providing that the actual TAS is within 100kts of the calibrated TAS

That should cover Vs0 to Vne in almost any plane fitted with a turn indicator.

[Contacttower]

But if the AI/AH is working, chances are the DG will be, so the timed turn won't be necessary!



Are yes...but I like to be anal about my turns in the hold.;)

[MikeJ]

Pprune is well up to strength on this! Irrelevences galore.

The original question is in partial panel IMC flying. You have no attitude or DI info, they are covered up. No information is available on angle of bank (or pitch). This is to simulate loss of the suction pump, which in the standard old style systems causes loss of both AI and DI.

As IO540 says, just put the TC onto rate 1 marker for the number of seconds for the required turn, at 3 deg/sec. I've always found it easy. I've never heard of a 'g' correction, but at rate 1 in a light aircraft it is negligible anyway.

[Vee One...Rotate]

Personally I prefer the UNOS (undershoot/overshoot) method i.e. non-timed. Seems to be more accurate and require less fine-tuning after roll-out. The timed method relies on your standard rate turn being spot on during the whole turn whereas monitoring the compass throughout the turn (as in the UNOS method, described above) gives you an instantaneous cue to roll out.

I often combine the two - I use the UNOS method and roll-out. If the heading is slightly off I'll correct with a short timed turn one way or the other.

Each to their own!

[IO540]

The reality of timed turns is that they are suprisingly accurate.

One isn't going to be doing a 360, normally. Typical turns will be on ATC direction, say 20 left or right etc. That is about 7 seconds. You don't need a stopwatch for that - just count to 7 at about the right speed, and you will roll out on the '20' as accurately as anybody could have flown the desired heading anyway.

I never did timed turns before the IR, and every one I did was spot on, and most of the time (for the slight turns) I wasn't even using a stopwatch to time them.

Try it one day. It really works.

[Islander2]

Pprune is well up to strength on this! Irrelevences galore.

Apologies for the thread drift.

MikeJ, that raises an interesting issue. However, please don’t take my comments here as a personal attack as I can assure you they’re not intended as such.

Frequently when a thread here broadens from the original post, there’s a queue of people ready to step in to criticise. But why, and for what purpose? Generally speaking, the broadening of a topic allows it to be more thoroughly explored and that, quite often, can be truly educational. Surely that’s a plus? In any event, doesn’t the ‘expanding discussion’ process merely follow normal real-life human intercourse … or are these forums (which in their entirety could be described as ‘irrelevances’ with about as much justification) supposed to create a virtual interplay that’s quite different from established social norms?

Closely related to this is the desire of posters here to close down threads that, presumably, espouse views that differ either from their own or from the accepted norm. A classic example in the last couple of days has been the ‘RT at the hold’ debate. Yes it got heated, and yes maybe one or more of the protagonists portrayed themselves as a cross between a playground bully and a foul-mouthed lout. But, close the thread down because some posters were questioning whether an ATCO’s views and CAP413 had considered all the possibilities? Why? Some of those doing the questioning combine powerful intellect with considerable flying experience. What is it that compels people to step in and attempt to ‘control’ those contributions? Indeed, in the case of the RT thread, what is it that motivated those people actually to PM BRL urging him to cull the thread?

Not normally one for psychology, yet on PPrune I’m increasingly finding these questions interesting!

[Contacttower]

I agree Islander2, I don't see why people have a problem with thread drifts (as long as the original question has been answered, which in this case I think it has). With reference to the RT thread I was slightly disappointed that it was closed down...if people don't like a thread they can just ignore it...I don't see why they feel the need to PM the mods.

[Nathan Parker]

FWIW, this is what is said about turn indicator (not coordinator) errors in my (CAA-flavour) CPL course notes, as produced by the sadly now defunct Professional Pilot Study Centre:

Ok, I consider that a data point. I'd still prefer to have something technically oriented, rather than pilot training oriented. (The FAA stuff has historically been of dubious accuracy.)

I'm wondering if the issue could be that the steeper the bank, the more the turn rate is produced by pitching, rather than yawing? For instance, a 90 degree banked turn would mathematically be entirely pitch, rather than yaw, and might possibly indicate no turn rate in the turn coordinator.

[tmmorris]

That certainly sounds plausible.

Tim

[PoppyPilot]

Hi everyone - I had my lesson on compass turns on Friday afternoon following a classroom session with my instructor . We went over both methods ie UNOS and the timed turn. He said he didn't mind which one I used and suggested I try both before making a decision. I was rubbish at both, but was surprised to find slightly more accurate using UNOS. I thought I would find the timed method by far the easier of the two! Not entirely sure how much time is allocated to this in the air but boy will I need some practice. Now as I said in my original question has anyone got any other hot tips for the IMC! Thanks all.

[flyme273]

angle of bank TAS/10 + 7 for Knots ( + 5 for mph)

[MikeJ]

Poppy,
Just a further thought on limited panal turns. I thought about the alternative of timed v using the mag compass whilst flying yesterday.
As you know, IO540 and myself came out strongly for the timed method.

You have no direct information on angle of bank (or pitch). To fly reasonably accurately requires constant attention to the remaining primary instruments. The standard layout puts the TC at the lower left, and the ASI above it, with the Alt to the right of the blanked off AI.

The mag compass is to your right, above all other instruments, and often, as in my aircraft, roof mounted. It is outside the field of view when flying on the primary instruments. In short, you cannot see the TC and the mag compass at the same time. The more you turn your head to follow the mag compass, with all its wobbles in turbulence, as well as the previous mentioned errors, the less accurately you will be flying the aircraft in maintaining rate of turn and altitude. The timed method obviously only works if you retain the rate one turn accurately, but this I have always found easy if the mag compass is totally ignored until it will have settled 10 secs after levelling out.

In my aircraft, and I think most, the clock is within the field of view of the primaries, and I don't bother with the stopwatch. My examiners, in all the multitude of renewal tests I have taken, have normally asked for 60deg or 90 deg turns, ie 20 or 30 secs, not 20 deg which as IO says can be done simply by counting to youself. I wouldn't trust myself to count to these longer times. But given the instruction on the heading to finish on, you can always take a few seconds to know the time required, decide clock positions of the second hand for start and finish times, and start and finish the turn accordingly. As IO says, it really does work, provided you attend to keeping the TC on the rate one mark, and don't take sneaking glances at the mag compass.

It means being well practiced in simply maintaining the rate one turn, before expecting accuracy in turns to a new compass heading. Certainly for me it means that I can maintain much better altitude accuracy than if ones eyes are moving to the mag compass during the turn.

Whilst the test gives wider tolerances on accuracy when on limited panel, I have always set myself the achievement of the same tolerances of full panel flying, and really do not think I could do this if I attended to the mag compas during the turn. (And whilst I'm IMC rated, I also set myself the IR tolerances.)


Just a thought.

[PoppyPilot]

Thanks Mike, that makes alot of sense. I will try that next time!