Engine starting

[sharpend]

There is a thread 'Why are aircraft engines so thirsty?'.

So I thought, 'why are aircraft engines so difficult to start?' OK, so they are not, on the whole, fitted with the most sophisticated electronic ignition systems; in fact most are fitted with twin magnetos. But at any airfield in the early morning, especially in winter, one hears the churn churn of an engine waiting to fire up.

Why is this? Surely twin mags + twin plugs should ensure that the engine starts like a car, ie instantly. After all, a car, even an older car with a carb and choke starts just about instantly (unless the battery is weak).

So why do we sit there with no much happening?

[Tupperware Pilot]

So I thought, 'why are aircraft engines so difficult to start?'
Are they????? Don't have any issues starting mine? C-85 hand start..... :ok::)

[flybymike]

Aero engine mixture management is a black art. Wait till you see the fun that can be had trying to start a hot fuel injected turbocharged engine.;)

[Dave Gittins]

Because we stop an aeroplane engine by starving it of fuel with full leaning.

To get it going again, you need to drag the right mixture back into the cylinders.

Used to be just as hard to start my old MGB by pulling the choke out and cranking (anybody remember chokes on cars ?).

[Zulu Alpha]

Car starting these days is all done by computer. I do remember old cars that were very difficult to start.

Most aircraft engines don't have computer controlled injection and ignition and are therefore as difficult to start as old car engines.

I fitted electronic ignition to my last LAA Lycoming engined aircraft. It transformed the starting. I am going to fit it to my current aircraft this winter.

[Rod1]

I stop my aircraft engine by turning off the electronic ignition…

A Rotax aircraft engine has a choke and is exactly the same as starting a 1970’s car. Very easy if you know how and very hard if you have no clue.

Rod1

[Dave Gittins]

I have had very little exposure to Rotax's (only one in a Pipersport). The first time it stopped I thought it had seized.

[BackPacker]

Surely twin mags + twin plugs should ensure that the engine starts like a car, ie instantly.

Like others said, the issue is normally not the sparks. It's getting the right mixture to the cylinders that's tricky.

Sparks can be a problem too though. The magnetos are configured so that the spark happens about 20-25 degrees before Top Dead Center, which means that the flame front, at normal RPMs, reaches the piston shortly after TDC.

However at very low RPMs, such as when the engine is being turned by the starter motor, the flame front will reach the piston well before TDC and will cause the engine to stop, or even backfire. That's why aero engine magnetos are fitted with impulse couplings, which retard the spark by about 20-25 degrees at very low RPMs.

Impulse couplings are normally only fitted to one set of magnetos though, and it's only this set that is activated when you rotate the mag key to the "start" position. But in aircraft that have a separate starter button (not integrated in the Off/L/R/Both key mechanism), there might be a note in your POH that you should only start on the left (or right) magneto, not on both.

[Cobalt]

I would also add ageing engine / maintenance issues into the mix.

In autumn, several of our aircraft (C152) became progressively hard to start, and it became next to impossible in winter - heavy priming, throttle pumping, all the BAD things we had to do to get them going at all... and when really cold even that stopped working and we could not fly.

We leaned on maintenance, they replaced the primer nozzles, and hey presto, 2x primer pumps and they started every time for the rest of the winter. 1x would get it to fire, but would not always sustain it.


Also - is anyone actually following the POH procedure for priming a PA28? Read it an be surprised... I have yet to meet a pilot who uses it, everyone just primes 3 or 4 times and cranks. The POH says crank, if no start THEN prime ONCE and crank again. Works very well...

[A and C]

Slick have an SB that recommends a 500 hour check, the reason for this is to change the contact breaker points and check the internal timing in the mag.

The internal timing has a large effect on the strength of the spark, while on most occasions the Slick mag will run OK up to about 700 hours the efficiency however drops of to the point that the engine won't run at idle on one mag.

It follows that a weak spark will not start an engine very will especially when cold it may well be the mag internal timing.

[tmmorris]

Cobalt,

Interesting. That got me reading the POH for A and C's, which I rent. What you didn't mention (which is in ours, at least) is it says to operate the throttle accelerator pump 2-3 times first.

The Robin I used to fly had no primer and was started exactly as above. I'll have to try it. Though I've found the POH warm start procedure doesn't work...

Tim

[david viewing]

it says to operate the throttle accelerator pump 2-3 times first

I'd go easy on the throttle pumping.

Following a near-disastrous fire to our -161 (rescued only by a quick thinking pilot of another aircraft who discharged is own extinguisher into the cowling) my late partner and co-owner Trevor Shaw looked into what had happened. The renter pilot had pumped the throttle "5 or 6 times" in an effort to get the recalcitrant engine to start.

Trevor, who was a garage owner and race car mechanic, found that any more than 3 pumps had neat fuel running out of the air changeover box onto the inside of the cowling. On the other hand it took a great deal of primer pumping to get any fuel to emerge at all.

The real culprit was the recalcitrant engine which had got progressively harder to start. It turned out that the carb was badly worn and that there several other contributory factors. After Trevor finished with the engineers and they finished with the engine, it became the best starter I've ever known, starting first turn, every time, all the way through the rest of it's 2300 Hr career.

[Pegasus912]

Hi Rod1,

Very easy if you know how and very hard if you have no clue.

I think I fall into the latter category! Never had any problems starting the 912 from cold, full choke and no throttle. Same with hot, except no choke. It's in between that I tend to get into a muddle. Care to offer some advice?

[Ebbie 2003]

My PA28-181starts OK if the manual is followed, but I've seen some epic fails by others trying to get it started especislly when hot.

When hot no one seems to like the idea of the mixture being in the lean cut off position.

It seems to be a need thrre hands situation. Seems mixture in lean cut off, throttle half inch open as opposed to the usual quarter following a couple of throttle pumps, crank it and when it catches smartly xture to full rich - works every time from cold, same deal but full rich, quarter inch and three or four throttle pumps - can prime instead but the Essex primer lock is fiddly.

The engine is a Lyc 360-A4M - must admit this is in the Caribbean so it's never "UK cold"

[tmmorris]

Our PA28-181 POH states for cold start, throttle 1/4 open, not 1/4 inch. (Warm start says 1/2 inch.) Is that a misprint, or have others misread it?

[Ebbie 2003]

It's 1/4 INCH - not 1/4 OPEN

Amazed that it starts at all if it's being cranked with the throttle open that much, if it does it suggests that the throttle linkage has an issue with the 1/4 open being the fix - but that would not explain the 1/2inch on hot.

Maybe we have a solution to some starting problems - errors and misprints in rewrite POHs and chec lists?

[A and C]

It's my aircraft that Tim is talking about, there are no issues with the throttle linkage.

I favour 1/4 inch open for a cold start...... it fires first time for me.

[sharpend]

Any views on left mag (impulse) or both for starting a Lycoming IO 360 in a Bulldog? There seems to be several opposing thoughts re this.

[BEagle]

I favour 1/4 inch open for a cold start......

As I explained to a young lady student "¼ in means just that. You might have been told by some that an inch is perhaps rather different to its true size, but in this case boasting is irrelevant - ¼" means ¼"!".............:ooh:

Bluntie, at the cranking speed of a 'dog, fatter sparks from using both magnetos is probably of more benefit than an impulse magneto alone. But what do the FRCs say?

[tmmorris]

Ebbie, and A and C,

Don't worry, I do in fact use 1/4 inch open. However, this is a verbatim quote from the POH (if I could work out how, I'd post a scan of it):


(Pages 41-42)
D. Starting Engine when Cold
...
4. Throttle Control - pump throttle full open and closed 2 to 3 times then set control 1/4 open.

I guess it's a misprint. The POH is typed, due to the age of the ac.

Tim

[cockney steve]

I'm really going to stick my neck out here, and challenge BEagle's statement.
A normal magneto , if turned by hand , can be felt to break the magnetic"grip"
That instant when it "lets go" is when the lines of flux are broken and the current is induced in the windings,,,hence the points-gap is critical, as this IS the deciding factor in getting the timing spot-on for a fat spark, which will only occur if you have the maximum generated current at the instant the points open and collapse the field....that's what induces the HT voltage in the secondary winding.

At cranking-speed, the rotation can be too slow to generate a sufficient primary current....so we have an IMPULSE START.

This is basically a clock-spring in the couplung between the drive and the magneto"innards"....a Pawl engages the spring-mechanism and continued rotation winds the spring until the pawl hits a trip-mechanism which releases the magneto to be "flicked" round by the spring unwinding. The rapid twirlthrough the break-point gives a full spark. the trip-mechanism is set to operate at a point well retarded from the normal "run" timing. The Pawl has a centrifugal weight, such that, should the engine fire, the Pawl retracts and the spring , uncoiled, acts as a solid drive in it's normal "run" advance timing.

If you crank with a "normal" Mag engaged, It will fire aroung the same time as the impulse starts winding up on the other mag...It could fire a "good" mixture and stop the piston dead then reverse it's direction (a "Kick-back" ) OTOH, the impulse timing is late-enough that by the time the cylinder charge is burning and expanding, the piston and crank have gone over TDC and thus the expanding gases push the piston on the right side of TDC.

Note the impulse -coupling is NOT SPEED DEPENDENT (except to disengage) This means it will ALWAYS (unless broken!) be engaged on a stationary engine....no matter how slowly you turn the prop by hand, the coupling will flick the mag round at it's release -point.

IT THEREFORE STANDS THAT IF THE MAG IS "LIVE" IT WILL SPARK.

A broken ground-wire will give this situation, that's why a running "drop-test" is important, it shows the mag is working AND the cut-out is also working!


Summary...Start on impulse and switch to "both" as soon as the engine runs.

Always be aware that a slight movement of the prop , could trip a spark and start an engine.

Sorry Mr. Beags...If I'm wrong, please tell us!

[phiggsbroadband]

Hi, we once had a Petter stationary engine, on a water pump, that had an Impulse Magneto. After it had run for a bit to get it warm, if we needed to re-start it, all we had to do was crank the flywheel over slowly (even taking 10 seconds from BDC to TDC.) and when it reached TDC the Impulse would occur, and the engine would thump into life.

The voltage produced by a coil with a contact breaker is proportional to the rate of change of current (di/dt.). So if the points open instantaneously, there is a very large voltage or back-emf produced.

(I have forgot the actual formula but suspect it is something like
dv = L di/dt where L is the Inductance.)

[Piper.Classique]

Cockneysteve, thank you for that very clear explanation. Now can you tell me how shower of sparks magnetos work? The ones with a buzz box that feeds off the battery while the starter is engaged.
Many thanks if you can!

[sharpend]

Hi Beags,

The FRCs (RAF) say use left (impulse) as does the BAe manual. But I have been told by a Lycoming chap to use both. I cannot think what the disadvantages are for using both; if we get a good spark from the impulse mag, then surely the No 2 will have no effect, good or bad.

Mind you, my engine turns very slowly as I need a new starter motor. I'm going for the lightweight SkyTec which has good reviews. That said, it can overheat if used for too long.

Now, having read Cockney Steve's explanation, I think I'll try starting with just the impulse mag, though I have not had an examples of kick-back using both.

[Ebbie 2003]

I think it is ambiguity or a missing " - I think it means 1/4 inch - and as has been said 1/4 inch cold start no problem.

Still think one should follow the procedure in the manual and will have no problem.

Any need to depart from the manual's procedure does suggest something in the airplane is non-standard.

[cockney steve]

@ Piper Classique.... I am not familiar with the technical side of "shower of sparks" magneto, but it's very simple to do with coil ignition. The points could operate a transistor to hold an astable oscillator "off"...the points open, transistor switches oscillator "on" oscillator , in turn, feeds it's output pulses through the primary of an ignition-coil thus you get a shower of sparks which switches off when the points close again.

As a kid, I had a "shock-coil" -basically it's a transformer, (as is an ignition coil)
the primary connected in series with a buzzer/electric bell....connect to a battery and it operates as follows...
solenoid coil of buzzer/bell and primary of HT coil (series!) energise...solenoid pulls spring-loaded armature (a bit of soft iron) towards it...the primary circuit is arranced so a break in the wire is connected one side to a fixed contact point and the wire continues from a contact -point attached to the armature... thus, when the armature moves towards the coil, it breaks the circuit, the magnetic field collapses and the armature springs back to complete the circuit again...

Distance from coil to armature, strength of spring (and one or both contacts can be sprung as well!) and supply voltage all affect the rate of oscillation....
Meanwhile, the pulsing on-off current is passing through the shock/spark coil primary winding.... every time the coil is energised, it magnetises it's core and every time the current stops, the magnetic flux breaks down and induces a voltage into the secondary.

this is necessarily a shortened explanation, i've deliberately left out the stuff Phiggs alluded to, such as quenching capacitor across the points.

I suppose a magneto could be arranged with a modified form of impulse-coupling, such that when the coupling "flicks" the mag, the rotating part of the Mag has a one-way clutch with a single engagement-point and sufficient mass that the impulse is able to spin the armature multiple times until it slows sufficiently for the drive to "lock " again that would give a brief shower of sparks, but remember the Mag has a permanent magnet and the armature is "braked2 every time it's poles pass the poles of the field magnet.

Physics....the energy for the spark has to come from somewhere. :8

I must start going out!


Forgot to say, an electronic ignition system can trigger a shower of sparks and , to prolong plug-life, revert to single-spark under benign running conditions.

@Phiggs...yep, I had an air-cooled single "Tuff-Tuff " with a Wipac impulse mag....hand cranked it would "clink....clink.....clink as the impulse fired....if you cranked hard enough, you could exceed the impulse cutout speed:) just pulling the flywheel to the "Clink" would often set it going.

we used to drag it out and set it going on the workshop floor...as it had a speed-govenor, it would scuttle about as you revved it, and judicious jiggling of the throttle would produce backfires of vivid blue flame from the exhaust stub....a length of pipe could be held with one end near the stub and be "tuned" like a trombone :} Amazingly, I found time to work as well! Think it was a couple of quid, non-running at an auction. best value entertainment ever.

[Onceapilot]

cockney steve was correct. The "buzzer" or "shower of sparks" system uses switched LT from the aircraft battery in the primary winding of the magneto to generate a series of sparks in the starting period when the magneto is turning slowly and so generating little LT itself. Cheers

OAP

[A and C]

I would advise only using the left mag, part of the job that the impulse unit does is retard the spark to just after TDC thus all the effort of the first power pulse is to send the piston in the correct direction.

The right mag will always fire at (depending on the O360 variant ) 20 or 25 degrees BTDC and try to send the piston in the reverse direction if it fires at the slow cranking speed. So at best having the right mag on during start will do nothing at best and at worst it will appose the efforts of the starter motor.

[Maoraigh1]

It's many years since I had a starter separate from the mag switch - both on before key works starter. I can see the point when handswinging, although I've done it with both on.

[tmmorris]

On installations with a joint mag/start switch (e.g. PA28) does pressing the starter ground the right mag?

[jxk]

both on before key works starterWhen the combined starter/magneto switch is turned to start the right mag is grounded so that it's just the left mag that is live. Releasing the switch puts both mags on-line. It's all in the design of the switch and that's why they have all those terminal connections on them and why they ain't cheap.

[BEagle]

cockney steve & A and C, you are both correct and I was wrong. I didn't realise that the right magneto fired quite so early - so particularly with sharpend's low cranking speed, left magneto only is indeed the correct selection.

[sharpend]

Hi Guys, thanks for all the advice. I tried starting as per the RAF FRCs yesterday (left mag only) and it appeared to start better. Only trouble is, problems do come in twos or threes! When I press the starter button for the first time, the engine will not crank, but there is a whirring noise from the bulkhead. Second push activates the starter, but the engine cranks very slowly. Battery voltage is fine

Luckily the engine normally fires after a couple of turns of the prop.

So my diagnosis:

1. Starter relay knackered.
2. Starter knackered.

ps recently replaced the bendix.

[Maoraigh1]

Or a bad connection. First push heats it up - expands - better connection, but still poor, turns starter slowly. Very cheap to check the connections, clean, retighten, and try again.

[cockney steve]

Maoraigh1 Or a bad connection
a good and logical diagnosis, especially as he's had a new "bendix" recently (starter off)
Check battery cranking-volts! A £5 multimeter is OK....12V battery should be 13-14V dropping to~10V cranking....24V battery, ~26 resting, ~20 cranking....any lower and suspect a dying battery or see Maoraigh's post.
If it's pre-engaged....check it's the correct pinion, right number of teeth (YES, REALLY !) and properly engaged in the operating fork. you can check on the ground with a pair of jump-leads and battery....connect to the bug terminal and the casing...bridge the small terminal to the big battery terminal and the pinion should slide out with a smart CLUNK. the other big terminal is connected to a short-stout lead that goes into the carcase....connect the other big terminal (or the jump-lead direct to this and the starter should spin merrily......if you can blow out with an airline, it's good to get rid of all the carbon-copper from brush-wear.

Inertia starters DO use a Bendix...this is a quick-thread sleeve on the armature-shaft that engages a female component fixed in the pinion....a bit like a conjourer pulling out the table-cloth, stsrter spins and pinion slides along the quick-thread until it reaches the end, they then both rotate as a whole.

Cracked sleeve or "star" (female part) or dirt/oil/grease can stop it working. again count the teeth! easy to fit wrong one and it will bind on the ring-gear.


DID THE PROBLEM ONLY START AFTER REPLACING THE PINION?BENDIX?

Due to computer crashing, above was somewhat shortened.

having re-read some of the thread, it seems the starter has been sluggish a while. Is it Pre Engaged?.....that is, it has an integrated solenoid with a direct feed from the battery...there is a small terminal, which , when energised, pulls in the solenoid....this moves a forked-arm to push the pinion into mesh with the engine's ring-gear (flywheel) It also has a heavy contact which joins the 2 big contacts on the solenoid end-cap one is the battery, the other is the starter windings. these internal contacts wear but can be replaced usual symptoms, erratic starting...sometimes nothing, sometimes normal, sometimes slow.

If you are on permit, AIUI it's OK to DIY...check the brushes are long-enough....too worn and the pressure springs or tail-wires hit stops and the brush cannot press firmly against the rotating Commutator (copper segments on the armature)...broken or jammed brushes, same effect...worn bearing bushes will allow pole-shoes and armature to foul each other....very rare! excessive carbon-dust from brush-wear can track down a lot of current to earth..always worth blasting innards with an airline. A field-winding that's tracking to earth...tricky to find, high current consumption is a giveaway but you'll need a meter for ~200 amps! (excess battery-drop whilst cranking is another giveaway and the motor will get hot.
Thrown solder on the commutator segments....caused by using a starter too long, especially with a "low" battery...bad or open windings on the armature + low speed and overheating.

All can cause low cranking speed.
many modern starters have permanent-magnet motors, they're smaller, lighter and run at high speeds,so incorporate (usually) epicyclic reduction gears.

Overheating these motors ,weakens the magnets...starter gets sluggish AND overheats....self destructing,if abused.....the old field-coil starters are much more robust in this sense, but see "thrown solder" , "partially -shorted field windings".....yer don't get owt for nowt!

[Brian Abraham]

all we had to do was crank the flywheel over slowly (even taking 10 seconds from BDC to TDC.) and when it reached TDC the Impulse would occur, and the engine would thump into life.On some early engines (WWI SE.5 fighter was an example) the pilot wound the handle of a booster magneto, which provided the necessary spark, and the stationary engine would burst into life. Should that fail it was then necessary to resort to hand swinging the prop.

[cockney steve]

"in the old days"....some cars were equipped with "shower of sparks"ignition....I'm sure a search for "trembler-coil ignition" will throw up something!
Cars into the 1930's routinely incorporated a toothed ring in the centre of the steering -wheel protruding from the hub, radially, were three levers that had ratchet-engagement with the toothed ring....they were hand-throttle, mixture and advance-retard.

a popular trick was to leave the car fully-retarded before switching-off...upon return, switch on ignition and advance timing....invariably a 6-cylinder would start..a 4-cylinder a bit more of a gamble, but it saved the hand-cranking.
Quality cars of the period often had both coil-ignition and Magneto.
not much new in the basic internal-combustion engine, just refinement of old ideas Modern materials and techniques help, too.:)

[Piper.Classique]

Well, that was all very interesting. Thanks, all.
BTW, when we replaced our starter on the Cub the manufacturer's docs recommended a dry silicone spray on the bendix every 25 hours. Comes from caravan shops if you want to get some.

[cockney steve]

@Piper Classique -Are you saying that your starter is of the inertia-type? IE the pinion slides along a helix (quick-thread) and not pre-engaged? In that case, you'd have a separate solenoid (just a bloody great relay, really) in the lead from battery to starter.......I did have a car where you pulled a knob to start....the knob connected to the solenoid via bowden cable.....main contacts have to take a couple of hundred amps,plus , when you actuate it, you're switching an inductive load...so the solenoid contacts have a hard life.

Another "gotcha" for the inertia starter.....the engine firing gives a flick which reverses the pinion and throws it back along the Bendix, where it's arrested by a big cush-spring.......but what stops it vibrating back along the shaft?... there's a light control spring that sometimes breaks.....when it does, the pinion can keep hitting the revolving ring-gear on the engine.........so, if you hear an intermittent metallic tinkling, check the starter!

[A and C]

Selecting the start position on the starter key earths the right mag.......... Sorry being an American aircraft I should say it grounds the right mag.

[The Flying Pram]

I did have a car where you pulled a knob to start....the knob connected to the solenoid via bowden cable...The original Fiat 500 was like that - a small lever on the floor which you pulled up to start the engine. Another one next to it was the choke. The legendary "Little Grey Fergie" tractors had and extension of the gear linkage which pushed the solenoid in, but only from neutral so you couldn't start in gear!

[FANS]

It's another reminder of why GA is so antiquated and difficult.

When we have checklists that make a Tristar's look basic, ancient equipment, 1970s technology, it's little wonder that GA is knackered.

Yes, I'm all for the simplicity of things but there comes a point when we need to update the whole of GA and bring it in the 21st century.

We keep banging on about the safety of everything, but in reality we're flying knackered machines which take longer to get ready for departure than an A320.

Navigating using a clock and wobbly DI!

[sharpend]

Yes, but that's all the fun :ugh:

[thing]

Don't know if Shapend is being ironic but I think it is half the fun. If you want efficiency in getting from a to b then get on a scheduled airliner. I think there's something interesting in old GA stuff the same as some people find traction engines interesting. I don't want some super duper efficient machine that you have to press a few buttons to do what takes you ten minutes with a checklist on a 40 year old Cessna.

[Piper.Classique]

@Piper Classique -Are you saying that your starter is of the inertia-type? IE the pinion slides along a helix (quick-thread) and not pre-engaged?
Yes, that's what we've got. Its working fine, thanks. Yes, separate solenoid etc. We also have a switch to isolate the starter motor if we need to hand prop with the shower of sparks mags. Always wondered why american engines turn the wrong way, makes hand swinging from in front a bit of a pita.

[A and C]

The only reason that GA is so complicated is that a lot of the people who are running it are obsessed with turning it into a black art in order to make themselfs look important.

You need to learn from those who keep it simple.

[FANS]

If you're flying club aircraft, you've got to fly to their rules which includes the 20 page checklist.

I totally agree with the tractor attraction, but my key point is that if I want to drive a 1970s Ford Escort great, but I'd also like to drive a modern day, easy to use car if I choose.

At the moment, I want to have the option to drive a new, easy to use machine or a 1970s designed Piper that takes twice as long as an A320 to get ready for departure.

[Flyingmac]

I've come across poor starting Lycomings where the primer only went to one cylinder. Joining up two more transformed the starting.

[Desert185]

A and C:

The only reason that GA is so complicated is that a lot of the people who are running it are obsessed with turning it into a black art in order to make themselfs look important.

You need to learn from those who keep it simple.

:ok: :D