What's new
Van's Air Force

Don't miss anything! Register now for full access to the definitive RV support community.

Hot start techniques

charrois

Member
Hi everyone. I'm plagued by hot start problems and am wondering what techniques are working for you if you have a engine similar to mine.

My firewall forward is probably nearly as stock as a person can get. I have a YIO-540-D4A5 with AVStar AVX-5VA1 and two Slick mags, left one with a retard breaker (exactly what the factory new Lycoming purchase option from Van's came with). I also have a "SlickStart".

I don't have a fuel return line - the fuel is plumbed as per Van's plans flowing through the fuel filter, boost pump, engine pump, fuel flow transducer, fuel servo, spider on top the engine, and then the injectors. The only deviation from plans here was to place the fuel flow transducer after the engine pump on recommendation from the manufacturer, not before as Van's plans call for. This eliminated a problem I had earlier with erratic fuel flow readings.

Normal starts when the engine is cold work great. I open the throttle about 1/2", boost pump on, mixture rich for around 5 seconds to where fuel flow and pressure stabilizes, mixture ICO, boost pump off, crank the engine and go mixture rich when it starts, usually within a second or two.

If I shut down and start up again fairly quickly (within 10 minutes or so), a procedure that usually works for me is even simpler - all I do is open the throttle around 1/2", leave mixture at ICO, don't run the boost pump or anything, and then crank the engine and go rich when it starts.

But I have no luck at all if the engine has been sitting for longer (30 minutes to a couple of hours). Even if the engine doesn't feel that hot to the touch, I can't get the engine running with either technique for the life of me - it just cranks with no life at all. The only way I've been able to get the engine started is with a flooded start - full throttle, mixture ICO, boost pump on (left on), mixture rich for 2 seconds then back to ICO, crank the engine and then use 3 hands to get the throttle back and mixture forward when it starts. Usually my third hand doesn't work fast enough and the engine surges at high throttle when it starts, which is why I hate that approach. And its made me apprehensive about flying somewhere for just a short stop.

It's always done this; it's not something that just recently started happening.
And on the last annual, I cleaned and gapped all the spark plugs as perfectly as possible, as well as adjusted both magneto timings to be spot on (resulting in the same 90 RPM drop on either magneto during a runup).

Surely there has to be a technique that works better than this? I know the trick is to get colder, non vaporized fuel into the lines without flooding the engine, but I haven't sorted out a way of doing that trick.

I know that people with a fuel return line don't really have the same hot start challenges since they can purge their fuel lines with cool fuel properly. I don't know the intricacies of fuel return line plumbing though - would it be possible to retrofit my fuel system with a return line without having to change everything out?

Thanks for any advice you can provide! I'd love to start being able to fly places for an hour or two without being worried I'm going to be stuck there... starting the engine has become the most stressful part of my flights.

Dan
 
I have IO-360. Might try to run the boost pump at ICO to pressure up the fuel line a few second before hot start. If it doesn't catch the first time I prime for a couple of seconds and try again.
 
Hello Dan

We use nearly the same hot and cold technique as you. We use more throttle on the hot technique (about 1/4 open). For the intermediate starts, we normally try the hot technique first, if that doesn't work then mixture rich, 3 secs of boost pump the mixture ICO. It then seems to start using the hot technique.

Just one data point.
 
Yep I usually do what Dan suggests in the last few sentences. It catches, I wait for it to use up the fuel and then SLOWLY add in mixture. Works every time.
 
The IO 540 was designed to humiliate you at the pumps. Right after you get out of the plane open the oil door to let heat escape. If your doing a quick turn around (hour or less) you will experience the notorious hot start. Over the years, I have tried all scientific and Vodoo techniques with little success. I always seem to revert to a good flooding of the engine to try to force out the hot vaporized fuel then go for a flood start, fuel full throttle and mixture to cut off. When she fires then a coordinated dance between the mixture and throttle. It’s not pretty but never been left on the ramp with a weak battery. At least so far.....
 
Like you, I was having starting issues. Not only hot, but cold too. Every start was painful.
I was using a standard mechanical magneto with retard breaker points and a slick-start. (not an impulse coupled magneto)

I bit the bullet and installed a Surefly e-mag (SIM6L) on the left side.
With the SIM6L, my IO-540 will start hot/cold/warm on the first blade.
I'll go on to mention that I've noticed improved fuel burn and much smoother engine operation in cruise since the magneto can advance the timing in cruise.
Just yesterday, I was indicating 9.9 GPH with 21 inches @4500' MSL. Before the E-mag I could barely get it under 11 without it rough running.
 
Not much question...any of the common EI choices will significantly improve starting, hot or cold, as compared to an impulse Slick. And with dual EI, we typically energize both ignitions to start.
 
As Dan mentioned install a EI. I put in a Surefly because of the ease of installation. Zero hot start problems since.
 
As Dan mentioned install a EI. I put in a Surefly because of the ease of installation. Zero hot start problems since.
I had been using various "OWT" methods on various aircraft for years, with what felt like random success/failure.

I now use Dan's "goldilocks/sweep" method for every start - and so far no trouble. Dual pmags. It really helps to understand how things work.
 
In my experience, the IO-540 starts easier when warm than the IO-360. That said, if you're having trouble with a cracked throttle along with a lean mixture and slowly advancing the mixture when the engine fires, you may have other issues. Perhaps a magneto or servo issue?

You mention that you have Slick mags. How much time is on them? It's not uncommon for them to allow the engine to run fine, but have trouble starting when warm as they age.

Your "been sitting for 10 minutes" start sounds like the correct procedure, and should work after the engine has been sitting a bit too.
 
Slick Mags - Starting

I had hot start problems on my 540 for the first 50 hrs or so. I then discovered that the internal timing on the factory set Slicks was off. Took them off, followed the procedure to time, fixed the E-gap and no more starting issues after that. How do you know if the internal timing is off? I found it with the buzz box when I noticed the difference in the amount of time the points stayed closed varied on the same mag during it's rotation. Reminded me of dwell angle on an old distributor ignition on a car. This "dwell" has a big effect on the strength of the spark at low RPM. The effect is masked when the engine is running at speed. As an aside, I replaced the Slicks at 450hrs with dual Surefly units and the engine starts first time hot or cold now and I get better fuel economy at altitude LOP... :)
 
Thanks for the great suggestions, everyone! I hope to head to the hangar sometime in the next couple of days, do a ~30 minute flight to warm up the engine properly, let it sit for around 45 minutes, and then try some of the techniques mentioned (in particular, I'm intrigued about the "sweep" method in the "Goldilocks" post, which not only makes a lot of sense, but should be relatively easy to do). I'll post back here with my results.

This will be the first time I've ever looked forward to seeing what happens during a hot start :)

My mags have about 300 hours on them. Since they came with the Lycoming, I figured I'd just run with them until they needed to be overhauled, and then perhaps try out one of the EI options. But from some of the comments, it may be worth looking into doing that sooner rather than later; hot starts aside, though my engine runs quite smooth once it's started, there's nothing wrong with improving that further, not to mention lowering fuel consumption.

Dan
 
I had hot start problems on my 540 for the first 50 hrs or so. I then discovered that the internal timing on the factory set Slicks was off. Took them off, followed the procedure to time, fixed the E-gap and no more starting issues after that. How do you know if the internal timing is off? I found it with the buzz box when I noticed the difference in the amount of time the points stayed closed varied on the same mag during it's rotation. Reminded me of dwell angle on an old distributor ignition on a car. This "dwell" has a big effect on the strength of the spark at low RPM. The effect is masked when the engine is running at speed. As an aside, I replaced the Slicks at 450hrs with dual Surefly units and the engine starts first time hot or cold now and I get better fuel economy at altitude LOP... :)

a Point to add here is that new mags will quickly lose their egap (internal timing) in the first 100 hours. Egap changes with point and follower wear and a new mag will often "wear in" in the first 100 hours. Those with new mags would be wise to check the egap at 100-200 hours to maintain optimal low RPM spark energy. I am helping a friend with his 500 hour since new mags and one had drifted SO far that it stopped firing all together due to a severely off egap. Each mag will wear in a bit different, so some will be fine and others will be quite a ways off at 200 hours.

You just need the slick timing tool (piece of 063 AL), a VOM and 60 minutes of time. You will also need a buzz box for re-installing the mag.

Larry
 
Last edited:
I had the chance today to do two hot starts, about 45 minutes after shutdown when the engine had been running for around 30 minutes each time. And I have to say that I have made some progress (I got the engine started both times without resorting to a flooded start and the resulting lever dance), though I think there is still some room for improvement.

I tried the "sweep" method of the throttle set for what would be around 1100 rpm, mixture ICO, and then slowly advancing the mixture towards rich. The engine would catch and then quit after a second or so. Bringing the mixture back to ICO and then cranking again while advancing the mixture didn't work (slowly bringing it all the way to rich, and back to lean over 20 seconds or so), though I found if, after the initial brief catch and quit, I added a touch of boost pump for around a second while at ICO before cranking, and then doing the sweep from ICO to rich again, it caught and then ran (the first time I attempted a restart, I had to do this twice - it hiccuped and died again the first time).

If I understand what's going on properly, this seems to make some sense. After a hot start, the fuel vaporizes and as it expands, pushes the fuel through the injectors into the cylinders, making them too rich, so starting at ICO means that as the fuel is used during cranking, it eventually reaches the proper mixture and starts. But I assume there is still vapour in the fuel line, which is why it promptly quits afterwards.

I noted that before start, my fuel pressure was around 25 psi (left over from the prior shutdown), but as soon as I started cranking and advancing the mixture, the pressure dropped way down to just 4 or 5 psi. It doesn't look like the engine pump makes much pressure when being turned at a slow speed from the starter.

In trying to understand things, is there a reason to leave the boost pump off during a hot start? At ICO, wouldn't the cylinders be getting no fuel regardless of whether the boost pump is on or not? And so long as the mixture is advanced slowly, would not the stoichiometric ratio be reached all the same at some point? I'm thinking that using the boost pump to keep some pressure in the fuel system during start might be an idea. I just don't know if it's a good one.

If it didn't take a good hour to repeat the experiment each time, I would have had a chance to do some more experimenting... but I did find that after an engine stumble after a second or so of running, it seemed that I had to repressurize the fuel lines with a brief burst on the boost pump at ICO to get it to re-start.

Any ideas? At least I'm a LOT closer - I did get the engine to start both times eventually without resorting to a flooded start. But the process is still not as smooth as I expect it could be.

I know from people's reports that getting an EI like the Surefly would help. I imagine a hotter spark at a more appropriate timing during start would help initiate combustion with more margin for error on the fuel/air ratio. But I thinking getting the right combination of events happening to get the ratio close in the first place would help, no matter the ignition source.

Thanks again for all the insight!

Dan
 
Last edited:
I noted that before start, my fuel pressure was around 25 psi (left over from the prior shutdown), but as soon as I started cranking and advancing the mixture, the pressure dropped way down to just 4 or 5 psi. It doesn't look like the engine pump makes much pressure when being turned at a slow speed from the starter.

Actually, engine RPM has little to do with pump pressure. The engine rotation merely rotates a cam, raising the pump diaphragm against a big spring, decreasing chamber pressure and allowing fuel in through the inlet check valve. Further rotation releases the spring, which pushes the fuel out the exit check valve. Note (1) RPM doesn't change pump displacement, (2) its total capacity is much, much greater than fuel flow at start, and (3) the spring provides fuel pressure even when the engine is stopped, assuming the pump pushrod isn't on the pump cam lobe.

If pressure is low, it's because the pump or lines contain vapor. The pump cannot effectively maintain pressure if fed vapor, or if pumping against vapor trapped in the line between the pump and the servo.

In trying to understand things, is there a reason to leave the boost pump off during a hot start?

None. Run it. It simply pressurizes the feed into the engine driven pump and the line to the servo. It does not significantly change pressure or flow downstream of the servo's ball valve, a feature of the Bendix-style system. Metering is based on differential pressure, and is more or less independent of the feed pressure, at least above 15 psi or so.

Thanks for the report. Let us know how the boost pump experiment goes.
 

Attachments

  • Regulator Drawing 1000w.jpg
    Regulator Drawing 1000w.jpg
    41.1 KB · Views: 197
Last edited:
I tried the "sweep" method of the throttle set for what would be around 1100 rpm, mixture ICO, and then slowly advancing the mixture towards rich. The engine would catch and then quit after a second or so. Bringing the mixture back to ICO and then cranking again while advancing the mixture didn't work (slowly bringing it all the way to rich, and back to lean over 20 seconds or so), though I found if, after the initial brief catch and quit, I added a touch of boost pump for around a second while at ICO before cranking, and then doing the sweep from ICO to rich again, it caught and then ran (the first time I attempted a restart, I had to do this twice - it hiccuped and died again the first time).

Dan

WHen I hot start, I give the engine a second of boost pump at full rich to get fuel in the lines, then start at ICO and throttle for 1100-1200 RPM. I tend to push the mixture in quickly as it catches, so often the engine quickly dies. I then leave everything the same (mixture still full rich) and the next press of the start button, she fires over in one blade. About 1 in 10 times that 2nd start attempt doesn't work and I go back to ICO and do it a again.

Food for thought.

Larry
 
Your description of vapor was good. If the mechanical pump provided no pressure then you might imagine vapor back down the line. In this case, you might tick the electric pump (ICO) to get some liquid back in the line from the electric pump to the mechanical/servo. Then, start at ICO and move to ~half mix at firing with a slow advance from there.

Where vapor forms will certainly depend on the temps in that section. If the fuel trapped between the boost pump and servo is at a temp to vaporize but trapped as a liquid, then it might flash as soon as the MIX releases allowing some running, but then stopping as the line refills. The boost pump will (obviously-duh) assist this process. I like to just hit the boost at ICO before the start process.

My IO360M1B acted just like your OP on hot starts.

Do you have a pump shroud and a blast tube?
 
Before you spend a wad of money on other ignition systems, try this after it's been sitting for 30 minutes to 3 hours. mixture firmly cut off. Then one or two seconds of boost pump to pressurize the supply to the servo. Since the mixture is cut off you are not flooding the engine. Boost pump off.

then throttle about 1/4 open.
Then crank and after about 1-2 blades move the mixture lever forward about 1/2 way(essentially like Dans sweeping description). The fuel pressure in the servo then flows into the cylinders and you get the right air to fuel ratio and a quicker smoother start, without the massive over flooding and extended cranking to clear and lean.

The slick start boosts starting voltage about 300% and is almost as good as an EI for starting. I installed a second 6393 retard breaker (starting mag) with a second slick start.

2 slick starts will light a fat or lean mixture with ease.
 
Last edited:
When I hot start, I give the engine a second of boost pump at full rich to get fuel in the lines,

If the mechanical pump provided no pressure then you might imagine vapor back down the line. In this case, you might tick the electric pump (ICO) to get some liquid back in the line from the electric pump to the mechanical/servo.

Then one or two seconds of boost pump to pressurize the supply to the servo. Since the mixture in cut off you are not flooding the engine. Boost pump off.

Interesting...the three responses all seem to suggest switching the boost pump OFF before cranking. I do it too, but it's mere habit. Anyone have a better reason?

Running the boost pump while cranking does pull a few amps. It does not change fuel flow as compared to a fully functioning, vapor free engine-driven pump. The mixture knob controls flow.
 
Interesting...the three responses all seem to suggest switching the boost pump OFF before cranking. I do it too, but it's mere habit. Anyone have a better reason?

Running the boost pump while cranking does pull a few amps. It does not change fuel flow as compared to a fully functioning, vapor free engine-driven pump. The mixture knob controls flow.

My reason is that wear in the servo may allow some flow that I don't want at ICO. At 200-300 RPM, I expect the mech pump will struggle to make enough volume for pressure to hit 25, though I have never really looked closely to prove that. I noticed my newly overhauled servo back from Don would hold 15-20 PSI for hours after shutdown via ICO. A couple months later it doesn't. Stands to reason that wear over time might allow some bleed through with full pressure from the boost pump.

this is likely splitting hairs, but uin the end, I don't feel that i need the boost pump to start the engine, so why use it. As soon as the engine catches and reaches Idle RPM, the mech pump should do fine. As mentioned, I do a shot of boost with full mixture before starting to get liquid fuel in the lines, so vapor issues don't concern me, though the rough running for the first 15-30 seconds would indicate there is still some vapor remaining.

Larry
 
Last edited:
Dan, cool image - very helpful! One question - where does the mixture lever play in your diagram?
 
Dan, cool image - very helpful! One question - where does the mixture lever play in your diagram?

"Metering Jet", shown as a fixed orifice in this simplified illustration. In reality it is a variable area passage, i.e. a disk or drum valve. As valve area is reduced, fuel pressure on the metered side of the fuel diaphragm is also reduced, and unmetered fuel pressure closes the ball valve
 
I’m no expert but I use the boost pump on, ICO and mixture sweep procedure.This seems to give me the most consistent hot starts. I have tried a second or two of mixture rich before cranking with the same procedure otherwise and tended to end up with a flooded engine. With a return line purging vapour this would probably work - I don’t have one.
 
Dan and others. When you start it on ICO and it runs for a few seconds and dies out are you still on the starter or do you waiting until it just dies out completely before starting again (this time with the mixture sweep)?
 
Dan and others. When you start it on ICO and it runs for a few seconds and dies out are you still on the starter or do you waiting until it just dies out completely before starting again (this time with the mixture sweep)?

In my case, at ICO and the starter on, as soon as it fires I let go of the starter and move the knob to rich and see if it stays running. If it dies, and it usually does, I leave the knob rich and hit the starter again. It usually fires on one blade. If it doesn't fire in 2 blades, I pull the knob to ICO without releasing the stater until it fires and back to rich; This time it rarely dies. It really depends how hot things are under the cowl and how much fuel boiling occurred. It seems somewhat variable to me, as each hot start is a bit unique.

Sometimes even after it starts, it is quite rough and need to run at 1200 RPM for a minute to keep it running as it gets cool fuel in the lines.

Larry
 
Last edited:
Dan and others. When you start it on ICO and it runs for a few seconds and dies out are you still on the starter or do you waiting until it just dies out completely before starting again (this time with the mixture sweep)?

The trick is to bring the mixture in BEFORE the engine dies. There is a natural cycle where the engine fires and is snorting and rough (indicative of super fat, flooded mixture), followed by a clean out where the mixture leans to more normal levels and speeds up, then it starts to fall on its face as it goes back to lean. This cycle lasts about 3 seconds, but is very distinct if you know what to listen for. Bring the mixture in just as it goes over the hump between "normal" and "too lean". 99.999% of the time I can keep it running on the first fire. No exageration - in 20+ years of flying Bendix Lycomings, I can count on one hand the times I have had to go for a second attempt at a hot start.
 
In my case, at ICO and the starter on, as soon as it fires I let go of the starter and move the knob to rich and see if it stays running. If it dies, and it usually does, I leave the knob rich and hit the starter again. It usually fires on one blade. If it doesn't fire in 2 blades, I pull the knob to ICO without releasing the stater until it fires and back to rich; This time it rarely dies. It really depends how hot things are under the cowl and how much fuel boiling occurred. It seems somewhat variable to me, as each hot start is a bit unique.

Sometimes even after it starts, it is quite rough and need to run at 1200 RPM for a minute to keep it running as it gets cool fuel in the lines.

Larry

Ya mine seems to like this RPM as well. Many times I have to keep RPMs up a bit all the way to the runway or just deal with the looks of others as I taxi by with my engine slightly rough. haha It clears right up at any amount of power.

The trick is to bring the mixture in BEFORE the engine dies. There is a natural cycle where the engine fires and is snorting and rough (indicative of super fat, flooded mixture), followed by a clean out where the mixture leans to more normal levels and speeds up, then it starts to fall on its face as it goes back to lean. This cycle lasts about 3 seconds, but is very distinct if you know what to listen for. Bring the mixture in just as it goes over the hump between "normal" and "too lean". 99.999% of the time I can keep it running on the first fire. No exageration - in 20+ years of flying Bendix Lycomings, I can count on one hand the times I have had to go for a second attempt at a hot start.

The only thing I don't like about this method on my plane is it usually results in some pretty nasty engine shake while it gets its life figured out. :)

I usually just stay on the starter and don't touch the mixture until it fizzles out on the first catch, then I slowly add mixture and it fires right up everytime. I don't know if I'm doing my starter any favors though when the engine gets up to 500 RPM. I should probably just let go of the starter when it catches and wait until it comes to a stop if I want to continue this method. I tried finding a max RPM for the starter but no dice.
 
"Metering Jet", shown as a fixed orifice in this simplified illustration. In reality it is a variable area passage, i.e. a disk or drum valve. As valve area is reduced, fuel pressure on the metered side of the fuel diaphragm is also reduced, and unmetered fuel pressure closes the ball valve
Thank you, sir, all clear!
 
I should probably just let go of the starter when it catches and wait until it comes to a stop if I want to continue this method. I tried finding a max RPM for the starter but no dice.

You shouldn't do that. While the bendix gear can deal with some of it (driven gear turning the drive gear), I don't believe it is designed to do that on a regular basis.
 
You shouldn't do that. While the bendix gear can deal with some of it (driven gear turning the drive gear), I don't believe it is designed to do that on a regular basis.

Ya I did a bit more research on the matter. It would appear that many of the starters have a pinion catch/release mechanism for this very reason (when engine spins faster than starter). It essential disengages the starter gear from the shaft and lets it free spin. However, I don't think my starter has this feature so I will modify my method to avoid any problems.
 
Last edited:
Ya I did a bit more research on the matter. It would appear that many of the starters have a pinion catch/release mechanism for this very reason (when engine spins faster than starter). It essential disengages the starter gear from the shaft and lets it free spin. However, I don't think my starter has this feature so I will modify my method to avoid any problems.

It doesn't really disengage. It works in concept like your ratchet; Torque applied in one direction and free spinning in the other. However, imagine attaching your ratchet to your drill spinning 800 RPM. How long would it take to wear out the ratchet's mechanism. It is designed predominately to apply torque in one direction and not spinning in the other. I don't design these things, so really don't know whether or not excessive free spinning has any negative effect, but without that knowledge assume that it does. I do know that it is pretty common knowledge in the engine world that it is best to release the starter drive once the engine is producing it's own rotation.
 
Last edited:
It doesn't really disengage. It works in concept like your ratchet; Torque applied in one direction and free spinning in the other. However, imagine attaching your ratchet to your drill spinning 800 RPM. How long would it take to wear out the ratchet's mechanism. It is designed predominately to apply torque in one direction and not spinning in the other. I don't design these things, so really don't know whether or not excessive free spinning has any negative effect, but without that knowledge assume that it does. I do know that it is pretty common knowledge in the engine world that it is best to release the starter drive once the engine is producing it's own rotation.

Ya disengage probably wasn't the best word. Disengage from the torque of the engine (and its speed) is what I meant. Your logic makes sense. My gut feeling is that mechanism will last a long time but I would rather not find out. I'll just change my technique a bit. Thanks for knowledge!
 
As another followup, I tried my idea of doing the sweep technique with the boost pump on during a hot start. It didn't work as well as I'd hoped - the engine still would catch and die after the first attempt. And since by then the mixture was partially advanced with the boost pump on, I'm guessing that the cylinders would get partially flooded before I could bring the mixture back to ICO. Whatever the reason, it would take 3 or 4 attempts to get the engine started this way.

What seems more reliable is to pressurize the lines with the boost pump and mixture at ICO (so the cylinders don't get flooded) and then with the boost pump off, sweeping slowly from ICO to rich until it catches. Almost always, it will quit after the first attempt (I presume there is still vapour in the lines). But if I re-pressurize the lines and try again, usually it starts and stays running the second time. If not that, the third.

It may not give warm and fuzzy feelings to my passengers when the engine takes a couple of tries to get started, but at least it seems to be pretty consistent and I avoid the lever dance by doing a flooded start.
 
Back
Top