What's new
Van's Air Force

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

Unhappy Result on First Engine Start

J

jpharrell

Well Known Member
I?ll get right to the sad punch line: I had a kickback and broke six teeth on the flywheel ring gear on my engine last weekend during the first start attempt.
Here are the details:

  • Aero Sport Power IO-360-B2B
  • 8.5:1 compression
  • Dual PMags with jumper installed to operate on the A curve. I set the timing on the PMags right at TDC, or as close as my eyeball could align the flywheel mark to the hole in the starter case.
  • Catto 3-blade prop
  • Skytec 149 NL starter
  • Andair fuel pump
  • Airflow Performance fuel injection
I used the cold start procedure recommended by AFP for the fuel injection.
a) Purge Valve out (at ICO position)
b) Mixture ? Full Rich
c) Throttle open 1/2
d) Fuel Boost Pump ? ON for 30 to 45 seconds
e) Prime: Purge Valve to RUN position for 5 seconds then back to ICO
f) Throttle to 1/8
g) Propeller Area -- CLEAR
h) Crank engine until it fires then reduce throttle to idle
i) Purge Valve to RUN position​

On the first try the engine fired after about two revolutions and ran for about 3 seconds then died before I could get the purge valve to the run position. I primed again and it did the same thing. I tried several more times with the same result. It really seemed like it was trying to start but was starved of fuel. On the sixth try I heard a clunk and the prop stuttered about a second after I pressed the starter button. Six teeth missing from the ring gear. The starter looks fine with no visible damage and the shear pin seems to be intact.

Obviously I need a new ring gear but the most important thing is trying to understand why this happened and prevent it from happening again. I thought the Skytec NL starter shear pin was supposed to protect the ring gear from damage.

I would like to hear comments on possible causes of this incident. The tooth damage was in several different areas. The photo shows it pretty well (yes, I already pulled the flywheel off). What can I do to make sure this is the last time this happens?

dsc_4833c.jpg
 
Kickback

This sounds like it's not timed correctly, possibly you have the ignition leads plumbed to the wrong cylinders? The two wires closest to the vacuum fitting go the the cylinders closest to the prop, see the manual. It's a wasted spark system so it doesn't matter if you time it to the wrong stroke, as long as you had TDC.
Cheer up though, the ring gear is probably cheaper than the starter, but I would have hoped the shear pin in the starter would let go first.
Good luck.
Tim
 
ring gear

I agree timing probably off. Ring gears are pretty easy to replace.
 
It is certainly possible to suffer a kickback with pmags even if correctly timed, especially when using a lightweight prop. Ask me how i know...
Assuming you have the leads attached correctly and were able to identify TDC correctly, there are a few additional things that will minimize the chances for reoccurrence. First, retard the timing slightly by rotating the ring gear teeth past TDC by several teeth. Second, if the engine doesn't start on first attempt, on subsequent attempts, (master still on) set your switches to ground the p-leads, hit and hold the start button to let the prop gain momentum for a few seconds, then unground the p leads while continuing to hold the start button. Harder to write than do.

Info from brad at emagair. Works for me.
 
Ensure your Ring Gear marks are designed to align with the starter mark and not the top of the crankcase centreline. Ask me how I know ;) albeit got away without any damage.

Diagnosed by someone with better knowledge than me reverting to first principles and establishing TDC with a plug out and checking piston #1 TDC.

I have also had both P-Mags re-time themselves once preventing start - I can only think some unusual voltage patterns? So check timing again.

No idea about the AFP procedure, but seems somewhat complex? I had an RV-8 with AFP, albeit purge valve not fitted. Started fine using standard cold FI procedures i.e. Mix Rich 1/2 throttle, prime for 3-5s (until you get FP), then ICO + 1/8 throttle and start.
 
The P-mags fire at TDC when below 200 RPM's. If your timing is off a little, you can get a kickback.

How many teeth is your starter ring? If it has 122 teeth then regarding it one tooth is equal to 2.9 degrees. (360 divided by 122)

One other thing, when my engine doesn't catch on the first try, I usually cycle the power to the P-mags. No real reason other than I am comfortable doing that and my P-mags are wired in a non-standard fashion so I can do that independent of the master switch.
 
Last edited:
With a lightweight prop, I engage the starter for a couple of prop revolutions and turn on the ignition system while the engine is already spinning.
 
What Andy said......been there, done that. On my ring gear there are timing marks both on the front *and* the back of the ring gear. It's very easy to use the wrong mark when aligning the TDC mark with the case seam.

It is also a good idea to confirm that the ring gear has been installed correctly.

A good reference: Lycoming documentation on Engine Timing Marks
 
Last edited:
149 teeth

N941WR said:
The P-mags fire at TDC when below 200 RPM's. If your timing is off a little, you can get a kickback.

How many teeth is your starter ring? If it has 122 teeth then regarding it one tooth is equal to 2.9 degrees. (360 divided by 122)

John did mention he is using a Skytech 149 NL starter hence it is 149 teeth = 2.4 degrees
Click to expand...
 
I'm generally critical of 'recommended' start procedures on FI Lycomings, and this is no different. Cranking at 1/2 throttle seems like a very bad idea, no way you can move all the controls quick enough to follow that procedure without having the engine >1000rpm with essentially no oil pressure.

My general flow:

Mixture rich (In your case, purge valve RUN as well)
Fuel pump 3-7 seconds, looking for fuel pressure to show on the gauge generally.
throttle cracked, just off the idle stop
Crank until it fires, and then release immediately, let it either run or die. As soon as it fires, the starter is just weighing it down.

That cold start procedure works in every airplane I've tried it in. Leaving the throttle up halfway seems a recipe for disaster.
 
I totally agree with Steven, your cold start sequence is way too complicated. I have over a 1000 hours with the AFP system with NO purge valve and I have never had a hot or cold start issue.
 
A little more information

I will add that I performed the "pull thru" test recommended by E-Mag Ignitions before the first start where you rotate the prop with sparks plugs out of the cylinders but grounded to the case to verify that the 1 and 2 cylinders fire at TDC and the 3 and 4 fire 180 degrees later. But I admit that I was looking mainly for gross timing issues like miss wired spark plug leads or spark well off the timing mark. I am sure the sparks were in sequence and approximately at the TDC mark but it is possible the plugs could have fired a little either side of the mark by a degree or two. If the P-Mags begin to advance at a low as 200 rpm as stated, is it possible that the timing could have advanced before TDC while cranking causing the kickback? I'd like to hear more from those with lightweight props and P-Mags.
 
What they said, And...

For what it's worth I used the "B" curve and retarded the engine to split the difference between the A and B curve. So the engine is a little more retarded than the A curve during start and slightly more aggressive during full power. I can tell you that since I did this my engine is very easy to start.

Lance
 
judoka5051 said:
For what it's worth I used the "B" curve and retarded the engine to split the difference between the A and B curve. So the engine is a little more retarded than the A curve during start and slightly more aggressive during full power. I can tell you that since I did this my engine is very easy to start.

Lance
Click to expand...

The "A" or "B" curve has nothing to do with starting. As stated above, below 200 RPM the P-mags fire at TDC.
 
Kyle Boatright said:
With a lightweight prop, I engage the starter for a couple of prop revolutions and turn on the ignition system while the engine is already spinning.
Click to expand...

You can modify the RMSD (Run Mode Start Delay) with the Emag EICAD program (or our EICommander) so that the ignitions do not fire until "X" number of revolutions.
 
I agree with Tom Martin and I would cut the boost pump time to about 3-4 seconds instead of 30 or more seconds.
 
Concerning boost pump use - I've found that some move more fuel than others, but 45 seconds is a LOT. My injected Hiperbipe needed a good 10 seconds, while the RV will flood after about 2 seconds. The Rocket is somewhere in between.

EDIT - Completely missed the purge valve part - never used one.
 
Yes it does.

N941WR said:
The "A" or "B" curve has nothing to do with starting. As stated above, below 200 RPM the P-mags fire at TDC.
Click to expand...

I retarded the flywheel a few degrees when I initially timed it, so the wider curve does exactly what I said.

Lance
 
Hydraulic lock

Everyone is right here on having you look at the timing, but with 45 seconds of priming there is also a possiblity of a hydraulic lock happening in one of the cylinders.

Vic
 
I've seen it

I've had my engine kickback during start with 2 P-Mags but never had any damage like that. As a precaution, I set my base timing about one tooth retarded. That way on start it will fire after TDC but once running you can make up the advance with the timing curve. There is also a start delay as mentioned that will allow a programmed number of revolutions before firing the plugs. I wonder if there is something else going on here.
 
Some of this discussion has gone off in the weeds with all the discussion of starting procedure and some incorrect assumptions. Check the original post. The purge valve is at ICO when the boost pump is run for 30-45 secs. That does nothing but circulate fuel through the distribution spider - nothing goes to the cylinders until the purge valve is pushed in. Similarly, the original post does NOT indicate starting at 1/2 throttle.
 
N941WR said:
The P-mags fire at TDC when below 200 RPM's. If your timing is off a little, you can get a kickback.

How many teeth is your starter ring? If it has 122 teeth then regarding it one tooth is equal to 2.9 degrees. (360 divided by 122)

One other thing, when my engine doesn't catch on the first try, I usually cycle the power to the P-mags. No real reason other than I am comfortable doing that and my P-mags are wired in a non-standard fashion so I can do that independent of the master switch.
Click to expand...
Sounds as if your starter is incompatible with starter ring. Count the ring and compare that number with the part number and specs of the starter. This is the likely cause.
 
Ring Gear

The ring gear has (or had) 149 teeth and the starter is a 149 NL so there is no issue there.

But to explore a different direction, I am using a momentary contact push button switch rated at 16 amps for my starter. It is ACS part number 15-131, just like 15-132 but momentary contact style. Is it possible that the switch bounced or stuttered causing the starter to try to re-engage while the ring gear was turning? Anyone hear of a starter switch problem that caused ring gear damage?
 
Timing

As several have suggested, I have concluded my problem is at least partially related to my lightweight prop. I found the following quote on the Emagair website under "Tips and Tricks".

Low Mass Props: Kick-Back Precaution

Lightweight low mass props, by their nature, have less flywheel effect on the engine. This allows the engine to decelerate much quicker as the starter motor pulls against each compression stroke, and as the engine approaches TDC - where the plugs fire for startup. Under the right conditions, the combination of low mass prop, weak or undersized battery, corroded terminals, cold engine, long cable runs, tired starter (or any combination) can cause the engine to come to a (virtual) stop by the time it reaches TDC. At this point there is a 50/50 chance the engine will move forward or backward (a so-called kick-back) when the plugs fire. While the ignition may be performing as designed, the results are the same. Fortunately, there is an easy defense.

Simply re-time the ignition with the engine parked two or three degrees "AFTER" TDC (instead of "AT" TDC as you normally would). AFTER TDC means (from the TDC position) rotate the prop in the normal direction of travel to your revised target timing mark. Note: This will also slightly retard the operating advance range, but this small amount is not likely to be noticeable for most flyers. If desired, the shift can be nulled with the EICAD interface.​

I will certainly try that when I reset the timing.
 
That's a better explanation than mine!

Good write up John. That's what I did, but I used the more aggressive curve so I didn't lose anything on the top end. In fact I still gained a little since I split the extra 4(?) degrees that the "B" curve provides.

Lance
 
Kudos to Aero Sport Power

I just wanted to report on the excellent support I have received from Aero Sport Power regarding this incident. I called them Monday morning to ask for advice and Doug immediately arranged to have a new flywheel painted to match my engine and shipped to me. The new part, smelling like fresh paint, was on my front doorstep on Wednesday night when I got home from work. That is all the way from Canada to SoCal. I just have to send my damaged flywheel back to them and my cost is basically the price of just the ring gear alone plus shipping. It would have cost me more to have a local engine shop remove and replace the ring gear on my flywheel. Thanks to Doug and Margaret and Aero Sport Power for helping me get this repaired.

Now I just have to install the new flywheel and reset/double check the timing. My plan is to re-time with a bias of 2-3 degrees after TDC and try again.
 
Update and More Questions

This is an update to my original post. I need some help understanding these P-Mags.

I installed the new flywheel this morning and got the prop re-installed. As a check, I performed another ?pull thru? test described by the Emagair installation manual to see if the timing was still set at TDC. As a reminder, in this test the spark plugs are out but grounded to the engine and the prop is rotated by hand while observing the spark timing relative to the timing marks. However, this time a watched very carefully as the TDC mark slowly approached the hole in the starter. To my surprise the #1 spark plug fired about 1.5 teeth (about 3.5 degrees) ahead of TDC. I repeated this several times with the same results except the two P-Mags would sometimes fire at different times by as much as half a tooth (1.2 degrees) in spite of being set exactly the same. At this point I was thinking the P-Mags were just set advanced by 1.5 teeth and this was the probable cause of my kickback.
I decided to reset the timing to 1 tooth after TDC which is about 2.4 degrees retarded. So I rotated the prop to 1 tooth after the TDC timing mark and followed the P-Mag instructions to re-time them. After doing so, I performed the pull-thru test again and found that the #1 spark plug fired ? tooth before TDC. What? There is that same 1.5 tooth offset advanced from the set point again.

So then I rotated the prop to 2.5 teeth after the TDC timing mark and retimed the P-Mags. This time I found that the #1 spark plug fired 1 tooth after TDC. Again, there is that 1.5 tooth offset.

I haven?t found anything in the P-Mag documentation about an offset between the set point and the spark during the ?pull thru? test. Maybe this is an intentional timing offset to account for latency in the P-Mag electronics, or maybe there is a problem with my P-Mags. I don?t know. I plan to call Emagair on Monday but I thought maybe some of you P-Mag owners may have encountered this before and can provide an explanation. I was just not ready to try another restart today and risk another ring gear until I can explain this behavior.

Anybody out there seen this before?
 
Still looking for P-Mag help

Anybody out there with P-Mag experience that can give me inputs on the post above?
 
jpharrell said:
Anybody out there with P-Mag experience that can give me inputs on the post above?
Click to expand...

I haven't seen this behavior with Pmags before. Are you sure that you're ONLY turning the prop in the direction of in-flight rotation when setting the timing (counterclockwise, if you're standing in front of the plane looking at the prop)? Bumping the prop back and forth to get the timing mark aligned just right can cause this problem because the gears don't (and can't) mesh that tightly together.

I'm guessing you're doing it the correct way but just wanted to throw it out there.
 
What I would consider at this point is to forget timing by the ring gear. Pull the #1 plug and find TDC by using something to find the top of the piston stroke. Then compare true TDC with the ring gear/starter pin. Set your PMag timing using the TDC piston stroke. Then check grounded spark plugs.
Hope this helps.
 
cjbouldin said:
What I would consider at this point is to forget timing by the ring gear. Pull the #1 plug and find TDC by using something to find the top of the piston stroke. Then compare true TDC with the ring gear/starter pin. Set your PMag timing using the TDC piston stroke. Then check grounded spark plugs.
Hope this helps.
Click to expand...

^^^ This is very good advice. You can use a piston stop and "flower pot", or simply use a dowel through the spark plug hole.
 
TDC

You can quickly tell if the engine is at TDC, on the 4 cylinders all 4 pistons stop briefly at TDC, no piston movement occurs for 1-2 degrees of crank travel and you can easily feel the crank go "loose" at that position. In other words the crank will rock easily back and forth 1-2 degrees because the pistons are not moving. Before or after that your dragging 4 pistons around also.
To find TDC with a piston stop takes a procedure and a protractor, not just a stop or a dowel. And it's not likely you'll find the ring gear support mis marked anyway. Locate it at TDC via the index marks, check to see if the crank goes "loose" there. If it does move on.
For now, maybe set both PMAGs to 5 after TDC, see if it starts and runs ok. Then you can use an automotive timing light on it using the marks on the back of the flywheel and the case seem to see where it's actually firing. A timing light with a variable advance knob will make this easy to see exactly where it's firing.


Tim
 
jpharrell said:
This is an update to my original post. I need some help understanding these P-Mags.

I installed the new flywheel this morning and got the prop re-installed. As a check, I performed another ?pull thru? test described by the Emagair installation manual to see if the timing was still set at TDC. As a reminder, in this test the spark plugs are out but grounded to the engine and the prop is rotated by hand while observing the spark timing relative to the timing marks. However, this time a watched very carefully as the TDC mark slowly approached the hole in the starter. To my surprise the #1 spark plug fired about 1.5 teeth (about 3.5 degrees) ahead of TDC. I repeated this several times with the same results except the two P-Mags would sometimes fire at different times by as much as half a tooth (1.2 degrees) in spite of being set exactly the same. At this point I was thinking the P-Mags were just set advanced by 1.5 teeth and this was the probable cause of my kickback.
I decided to reset the timing to 1 tooth after TDC which is about 2.4 degrees retarded. So I rotated the prop to 1 tooth after the TDC timing mark and followed the P-Mag instructions to re-time them. After doing so, I performed the pull-thru test again and found that the #1 spark plug fired ? tooth before TDC. What? There is that same 1.5 tooth offset advanced from the set point again.

So then I rotated the prop to 2.5 teeth after the TDC timing mark and retimed the P-Mags. This time I found that the #1 spark plug fired 1 tooth after TDC. Again, there is that 1.5 tooth offset.

I haven?t found anything in the P-Mag documentation about an offset between the set point and the spark during the ?pull thru? test. Maybe this is an intentional timing offset to account for latency in the P-Mag electronics, or maybe there is a problem with my P-Mags. I don?t know. I plan to call Emagair on Monday but I thought maybe some of you P-Mag owners may have encountered this before and can provide an explanation. I was just not ready to try another restart today and risk another ring gear until I can explain this behavior.

Anybody out there seen this before?
Click to expand...

Yes, I have. I have 1-P Mag and one std mag on a new carbureted TMXO O320 (Mattituck/ECI) engine. I was having a hard start problem. Checked the timing several times. Ran the pull thru test. The plugs would consistently fire just before TDC by a slight amount. Like you I even retarded the timing a half tooth plus after TDC. I was concerned that the plugs were firing before TDC and the mass of the prop ( alum Sensenich) and the starter were the only forces pulling the prop thru. I talked to Brad at EMag and decided to send the unit back for a checkout. E-mag said it was working fine and the delay that I was experiencing will only happen when you pull the prop thru slowly. Obviously you pull it thru slowly to check the precise time the plugs fire versus TDC line on the flywheel passing the timing hole on the starter frame. He said that when the starter turns the crank a faster speed the plugs will fire at TDC if that's what you set it at. I think this important bit of information should be in their manual. I do remember in their manual somewhere to make sure the plugs fire at or slightly after TDC, especially holds true with a low mass prop.

Jim
RV9a
45 hrs
 
tim2542 said:
To find TDC with a piston stop takes a procedure and a protractor, not just a stop or a dowel. And it's not likely you'll find the ring gear support mis marked anyway. Tim
Click to expand...

Not likely, but I've seen two ring gear supports "modified" with larger index holes for the larger lug - in both cases the builder chose the wrong position. I've also seen a brand new aftermarket ring gear support with the timing marks off by ~8* front and back

TDC with the piston stop and flower pot (protractor) is a very accurate way of determining TDC and the "procedure" is just that - 5 minutes. Once verifying the ring gear support timing marks via piston location you never have to do it again and can use the marks on either side of the ring gear support; I prefer the backside and crankcase CL.
 
First let me say I checked the accuracy of the TDC mark on the flywheel relative to the actual position of the #1 piston and the mark lines right up with the TDC position. I could see the top of the piston looking through the upper spark plug hole with a light shining in the lower spark plug hole. So I am confident that the TDC mark is an accurate representation of piston TDC.

I spoke to Brad at Emagair this morning and he confirmed what jimbo said about the spark occurring early in the "pull thru" test. He said the "pull thru" test is not very accurate at very low prop RPMs. It is just intended to be a confirmation of correct wiring, not exact timing. He assured me that the P-Mags will fire at the correct crank angle when the prop is spinning faster.

However, he reiterated the problem with lightweight props and recommended setting the timing 3-4 degrees after TDC and removing the jumpers to go with the "B" curve operation. That way the engine will start with 3-4 degrees of retard and run with up to 35 degrees of advance relative to TDC.

So that is what I plan to do and will try to start the engine again in that configuration.
 
John,
I cannot imagine the frustration of having to go through this on a first engine start. It seems to me that chasing a few degrees of timing here are futile. I have the IO-320 with dual P-Mags and Catto prop. When I first started flying I had mine timed too far advanced and I have never had a kickback much less a broken tooth on a ring gear. I was B curve and two teeth advanced. I would exhaust all other possibilities to prevent a re-occurrence.
 
MarkW said:
John,
It seems to me that chasing a few degrees of timing here are futile./QUOTE]

Its not at all futile. Not sure why you would think that. He is planning on following the manufacturer's advice that makes perfect sense, and its advice that has worked for others in the same situation (pmags with light weight prop), including myself.
Click to expand...
 
Maybe I could have worded that different. I don't believe retarding the timing is wrong, just not what caused the problem. I would hate to see him retard the timing a few degrees as Brad recommends and then give up on looking for other possible causes. Major timing issue maybe but not 2-3 or 4 degrees.
Erich, did you strip teeth on the ring gear?
 
I agree that a few degrees is probably not the culprit, if the engine is not cranking "vigorously" enough (prop should spin rapidly with no hesitation on the compression stroke) that is when a kickback usually occurs. Slow cranking = kickback.
 
I have an IO-360B1B with 9:1 compression ratio, dual p-mags, and a Whirlwind 20RV prop. I had occasional kickback problems, typically on occasions requiring a 2nd start attempt. My ring gear is missing one tooth as a result of one episode.

Per emagair, the combination of their ignition with a lightweight prop, high compression, and low starting rpm can result in kickbacks due to the spark firing at TDC. Makes perfect sense to me. The easiest thing to change is to retard the timing by a few degrees. Secondly, if a 2nd start attempt is necessary, keeping the p-leads grounded for a few spins of the prop during start up insures your RPMs are maximized before the first spark event, minmizing the potential for kickback. This is not necessary for first start event due to the Run Mode Star Delay, which accomplishes the same thing. RMSD does not occur for subsequent start attempts unless you power down, which I prefer not to do. Your choice though.

I agree that slow cranking can be the problem. Obviously a run down battery will lower your starting RPMs and increase your chances of kickback. Other than having a well maintained battery however, you don't have much control over how fast your prop will spin when starting. The battery was not the issue in any of my episodes. A few degrees of retard in the timing CAN make a difference. Emagair says so, and so do I, because it worked.

Erich
 
Last edited:
Success

I thought I should report back with the outcome of the second attempt to start my engine. After replacing the ring gear and resetting the static timing to 1.5 teeth (3.5 degrees) after TDC I cranked the engine up Saturday with success on the first try. Oil and fuel pressures were in the expected ranges and the temperatures were climbing as expected during the three minute run.

A link to the video is below.:)

https://www.youtube.com/watch?v=X4rcYkZ6oIY
 
You must log in or register to reply here.
Back
Top