I’ve included prayers in the title for completeness, but it’s definitely more of a dark art.
My aircraft is a Turbulent with a VW 1600cc engine. There is no choke, but you can use the accelerator pump to prime the engine instead. There is a Solex PHN-II carburettor and it has two Lucas SR4 magnetos, neither of which has an impulse mag*. However, when I bought the aircraft, one was set with a significant delay and the engineer speculated that this might have been so you could start it without it kicking back i.e. that mag wouldn’t fire until you were nearly at top-dead-centre so you could start the aircraft on one mag, fly it on two, and if the correctly timed one failed, the one with more-than-ideal delay would probably still get you home.
This is probably true, and would doubtless make starting easier sometimes. I probably get more kicks than starts, after all. However, with time I have come to believe that starting is about mixture, mixture and mixture. When the prop starts kicking back, I am usually close to getting it started so I might as well keep the mag timings equal and go for performance. I’ve never experimented with unequal timings.
At one point my philosophy was to start a little rich and carry on turning the prop until the engine started to fire. At this point, I would give a few more pumps of the throttle, wait a bit, then try again. I would try to feel my way around the mixture – if more fuel seemed to help, I’d add a bit more. If not, I’d turn the prop a few more times and hope to get back to the sweet spot. Carry on propping around the sweet spot, and sooner or later one assumes you’ll be lucky enough to have a running engine.
A little while later, I experimented with Easy-Start. It was certainly easier to get a kick with than AVGAS – it’s an Ether that will combust at most stochiometric ratios. However I was warned off it by a hangar-mate that knows about such things. Apparently it dissolves anything organic and so it’s best kept away from all the synthetic diaphragms, seals and tubes that you get in engines. I have been much less keen to use it since then. I can vouch for the fact that it dissolves paint.
Even with Easy-Start, cold weather starts could be very hard. I’d often get a kick or even a turn or two of the prop, but that didn’t always convert into a running engine. Logically enough, I figured that you don’t just need one successful spark to get going. You need a supply of fuel to supply the engine for a few revolutions until the carb, induction system and engine have all settled on a mutually acceptable compromise. That is to say, even with Easy-Start you need to prime the engine the right amount, just as if you’re just using fuel. It’s not enough to have a spark, some fuel and air in the right ratio in the cylinder. You need to have some more of that righteous mix in the induction system to power the next few revolutions. If the throttle is too closed, it causes too much resistance to let the engine start. If it’s too open, it’s even more scary than usual when it finally bursts into life.
The temptation when an engine won’t start in the cold, is to keep at it doggedly until you are suffering from both heatstroke and frostbite. But I think that if you keep turning the engine over you’re just blowing fresh air through the system and it never gets saturated enough with fuel to keep the engine running. If you give up, and come back after a cup of tea, the air in the induction system has had enough time to become saturated with fuel and it’s more likely to start. For the same reason, if the engine kicks back I sometimes leave the warm exhaust gases in the induction system for a while before my next stroke. My hope is that they will vapourise enough fuel to make up for the lack of oxygen and surfeit of carbon dioxide.
Other things I have changed are to replace the carbon-core high voltage wires with copper core. Arguably they stress magnetos less than carbon core and they’re more efficient with a better spark. I think on balance it helps. They do cause more radio interference. One day I’ll get round to fitting some copper braid to shield them. I also replaced the pots and cylinders. My compression is better. The engine runs more smoothly. I’m not sure that it’s much easier to start, but it’s less likely to stop whilst idling. As Ibra pointed out, starting into wind definitely helps, though my usual preference is to point the aircraft in the least expensive looking direction on the airfield. Conversely, taxiing downwind whilst idling too slowly, can cause the engine to stop. Even when running, these engines can be finickity.
My next epiphany was that if I bought a hairdryer and stuck it under the cowling I could heat the engine and carb up enough to vapourise the fuel more easily. In cold weather this was a big improvement. One day it took me 3 hours to get the engine running. The next week, hairdryer in hand, it took me 10 minutes. I was worried that the heat might cause the paint on the cowling to blister, but I need not have feared. A £10 Tesco travel hairdryer can only heat freezing air to luke-warm, but it’s enough. It’s not so practical away from the home airfield. A little heater under the carb with a small battery would doubtless be more efficient, but the LAA might be rightly sceptical of attempts to convert a travel kettle to heat petrol.
In short, after years of experience and deep (and quite possibly misguided) reflection I can sometimes manage to start the engine from cold, by warming it up. What I haven’t worked out yet, is why I can’t get the engine going if it’s equally warm before I start.
* I understand that impulse mags on SR4s have a nasty failure mode, where the failure of the impulse coupling can stop both mags from working.
Ignition timing can be very important for starting so one comment might be that when you dual plug any engine that was previously run on one plug, the optimum ignition advance (for highest power) is less. So if you were to run the engine with split mag timing, you might need more advance on the ‘active’ mag to make maximum power. Conversely if you’re running on dual ignition with equal timing you very likely need less advance than with the single ignition car engine installation.
The way to check power is with full throttle static rpm, looking for small changes, as the propeller load increases quickly with increased rpm.
Warm starts are always an issue for any hand propped engine because it’s hard to know how much to prime it. If it’s dead cold you’d probably have a routine based on experience, if it’s hot (just shut down) you’d likely not prime at all, but in between who knows?
If the Solex is set up like the car installation, i.e. downdraft above the cylinders, then the accelerator pump should be OK as a primer. If it’s a side draft carb under the engine a primer injecting fuel nearer the inlet valves might well help starting from cold.
You need “prime assist”
In the other thread I mentioned putting your nose in the exhaust to smell if it’s air, smoke or fuel and get a better idea of the priming situation
For very cold starts, easy-start may help but you really need a heater, there is no other work around
For hot starts, when it get stuck, I think you need to flood the whole thing down with cold fuel (close throttle and prime with acceleration pump) untill it comes out of the exhaust then purge the fuel (open wide throttle and wind the prop), then start from scratch as all the vapour locks were kicked out of the fuel lines
In the thread that spawned this one, Flying – Aches, pains and aeroplanes, Ibra wrote:- “Tubs I used to fly had WW1600cc with Leburg Dual Ignition Kit (electronic ignition not an impulse mag)”
On that theme this AAIB report could be of interest to some as it was concluded that mis-wiring of the Leburg sensors caused an EFATO when one of the triggering magnets was shed in flight probably due to a poor Araldited connection.
I understand that later versions of Leburg use a pre-made wiring harness which avoids the issue, that and a more reliable method of fixing the magnets.
That is good know, I guess wiring & earthing is very important in electronic ignitions Leburg or FADECs 
there are lot of flimsy parts and huge lack of redundancy in Turbs that can spoil someone’s day but I never had an issue with electronic ignition: as long as one can understand their flashing volts lights and make sure battery is well charged before flight as the aircraft has a very “weak alternator” (fuel endurance with VFR reserves is 1h anyway)
The aircraft originally never had an alternator, I only installed one to feed the new electrical backup fuel pump (the engine driven mechanical did gave up on one flight)
Fuel endurance with VFR reserves is 1h anyway
I get a lot more than that, again with a 1600cc engine. Which part of the fuel pump broke?
Trying to smell when the mixture is right sounds an excellent idea.
You should get more endurance with tailwind or if not going anywhere
I don’t recall exactly if diaphragm or some sealant ring that failed
Not going anywhere you should be looking at 5+ hours endurance. Even at 12 lph you have nearly 3 hours (without reserves).
My fuel pump, it was a cork gasket.
How many useable USG you have? I recall max 6.5USG, 3h is feasible as record preferably overhead the airfield, 5h no way? there is no mixture to lean it 
I have flown did a UK tour (Lands End & Scotland) with other 3 pilots, 2*1h30 leg max everyday, every evening: the word was “knackered”
Every Turbulent is different. According to the internet, the Stark Turbulents have 39 L tanks and most other sources mention either 10.5 US gallons or 8.8 Imperial gallons, both of which are 39 litres. I recall that mine held 36 when I filled it from empty. I looked through the logbook and my longest flight so far has been 2 hours 55 (Blackbushe to Welshpool slowly avoiding clouds and airspace) and I landed with plenty of fuel left. I’ve never fancied working out what the min usable fuel is.
Other aircraft seem to have about 95% usable fuel so if the Turbulent had only 6.5 gallons (24 litres) usable out of 10.5 it would be doing something terribly wrong. Unless of course your fuel is limited by weight rather than volume? Even then, if you pull the throttle back you can stay up at low power levels. I’ve seen 6 Litres per hour, though I wasn’t going anywhere fast. The original 25hp Turbulent reportedly cruised at 7.5 liters/hour.
I wonder whether you could get a nice rich and even mixture by rocking the propeller back and forth rather than turning the prop in the same direction, which will dilute any vapourised fuel with fresh air. I know there are some engine accessories on other aircraft that can be damaged by turning props backwards, but I don’t think there’s anything on an average Turb for which it would matter?
