What's new
Van's Air Force

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

How to break your starter, again

hevansrv7a

hevansrv7a

Well Known Member
This is really embarrasing but perhaps it will help someone.

The first time, I broke my starter and destroyed the ring gear. That was because, in part, I had mis-labeled left and right mag wires in my wiring kit which eventually resulted in starting on the normal magneto instead of the impulse coupled one which in turn caused kick-back. Sky-Tec rebuilt/repaired the starter for a reasonable fee. I fixed the wiring and thought my problems were over. This was late last year. As revealed below, there was a second problem.

Fast forward to yesterday. Smaller kickbacks. Grinding noises. Finally, away from home, of course, the starter would not engage the ring teeth enough to rotate the prop. A good friend flew to where I was, hand propped my airplane and I flew it home. The starter casting was found broken (the casting) again, in two places. I just knew I had to be at fault, consulted Superior. Yes, it was me, again. Then I got my RA and sent the starter for repair again.

The engine is Superior's IO-360 Plus 180 HP with their standard updraft FI. I had been starting from "cold" with mixture full rich (like I always did in my C-150). It turns out this causes many problems and accounts for my second episode of kick back and starter damage. As an aside, a similar issue was the cause of some "blown out" Ryton sumps. My sump is the standard cast aluminum and survived. Correct procedure is on page 34, I think. Mixture rich, pressurize the system with the aux. pump, turn off the pump, THEN MIXTURE LEAN CUTOFF, then engage starter, push in mixture as it starts. That part in BOLD is what I did not do and should have. BTW, the manual doesn't say to turn off the pump but at least I got that right.

I had been using approximately the correct method for hot starts, but very wrong for cold starts. They were amazed I was able to start it at all. It might have been better if I had not been able to because then I might have re-read the manual. My father always said that if all else failed, read the manual.

All sympathetic and/or educational comments welcomed. Gloating and calling me dumb, well maybe you can restrain yourselves. Thank you.
 
Your proceedure is what I always use to start FI engines on all kinds of airplanes. It's the only proceedure I know that crosses over between all the systems I've worked on. Moving the Mixture to ICO is probably not a complete solution to you problem, but time will tell. I would think that you should just try a little less time on the boost pump. I do know plenty of people that use that suggested mixture to ICO, but I've never done it that way, it seems like a lot of unnecessary lever movage for now reason. You either have enough fuel, too much fuel, or not enough fuel. It would seem that your problem was that you had too much.
 
Advance angle

Any possibility you have too much spark advance? A rich mixture doesn't necessarily cause a kickback, but 35 degrees of spark advance sure will.
 
Thanks, but..

Yukon said:
Any possibility you have too much spark advance? A rich mixture doesn't necessarily cause a kickback, but 35 degrees of spark advance sure will.
Click to expand...
No. It is correct. But your point is clearly relevant.

osxuser said:
Your proceedure is what I always use to start FI engines on all kinds of airplanes. It's the only proceedure I know that crosses over between all the systems I've worked on. Moving the Mixture to ICO is probably not a complete solution to you problem, but time will tell. I would think that you should just try a little less time on the boost pump. I do know plenty of people that use that suggested mixture to ICO, but I've never done it that way, it seems like a lot of unnecessary lever movage for now reason. You either have enough fuel, too much fuel, or not enough fuel. It would seem that your problem was that you had too much.
Click to expand...
I am simply telling everyone exactly what A. one of the experts at Superior told me, B. what it says in the operations manual for that engine. I use very little time on the boost pump, only enough to get a change in audible tone plus .5 - 1.0 seconds. Yes, I had too much fuel, since with fuel pressure and the mixture open to any degree, the fuel is flowing into the intake port.
 
I'm not sure that this applies to your problem but about a year or two a go Superior had a bunch of bad ring gears. They were prone to loosing teeth. After some conversation, two flywheels and 8 starts later they admitted there was a problem with a new venders ring gear. They sent us the other design (obvious difference) and 150 hrs later it's just working fine. The main difference was the good ring gear looked anodized and the bad one looked silver. The silver one appeared not to have a proper radius cut at the base of the teeth. Larry
 
I think mine is OK - and another observation

LARCO said:
I'm not sure that this applies to your problem but about a year or two a go Superior had a bunch of bad ring gears. They were prone to loosing teeth. After some conversation, two flywheels and 8 starts later they admitted there was a problem with a new venders ring gear. They sent us the other design (obvious difference) and 150 hrs later it's just working fine. The main difference was the good ring gear looked anodized and the bad one looked silver. The silver one appeared not to have a proper radius cut at the base of the teeth. Larry
Click to expand...
Mine looks sort of gold color and is a recent replacement.

Also, I should mention that either the starter is a little weak or the engine is really tight. It is not unusual for the first attempt at starting to result in some movement and then the blade stops. Then, the next try goes all the way around. This could be part of the problem, but I don't really know.
 
Correct amount of retard on impulse mag????

I'd check to make sure your impulse coupled mag has the correct amount of retard (LAG) as there are different ones out there. Your lucky the starter broke and not the case. I've seen a few cases trashed when the kick back was bad enough to bust the case where the starter mounts. :eek: that then involves a complete teardown :eek: ouch. Hope you find the problem. I really don't think it's a mixture issue, timing is my guess.

I'm using a very small 15 lbs battery to start my High Comp Lyco and one thing that I do is hit the starter, let off of it when the prop comes back off compression then hit the starter again when the prop starts going back the direction of rotation, this takes a lot of load off of the starter and your not trying to spin the engine over when a cylinder is coming up on compression. Been working for me for along time.
 
Last edited:
I was always taught to start FI engines at mixture cut-off as well and I've gotten backfires (or what sound like backfires, anyway) frequently on the few times I'd neglected to do that.

Ok, so I don't really know if there's a correlation or if it's just a coincidence in my case but I'm guessing there's a correlation. Does anyone know why this is? I've always just done it like this because that's what I was taught but I don't know why.
 
Mixture likely not the issue

I have a Bendix FI Servo as well, standard sump, Skytec starter. I used to follow their standard recommended COLD start procedure with the start on ICO until my A&P-IA and experienced CFI friend suggested COLD starting with full rich. I find that for cold starts it is a much smoother start up. If anything, I was getting kickback on the ICO Cold starts.

For HOT starts, absolutely yes, you want it to be ICO or you will never get it started.

Agree - check those mags for timing.

Sorry to hear about your starter woes,

Rob
 
I think your problem is your starter. The lightweight Skytec starters are junk, pure junk. Mine is setting up on the shelf of my hangar. After about 20 hours, it sounded like a quarter rattling around inside of an empty tin can. The permanent magnets cracked and came loose.

I tossed in the junk pile and got myself a new B & C starter. Problems solved.

I know all the posts to follow will extoll the longevity of their Skytec starters. Count yourself lucky. Just look at one of these next to ANY OTHER starter and you'll see what I mean.
 
One uninformed opinion from a carb guy

Pardon the ignorance of a carburator guy who has never owned a FI engine, but I have a real hard time imagining that simply leaving the mixture open would cause kickback. I can see how it might flood it, but simple logic inclines me to think that a rich mixure would be less likely to fire and therefore less likely to kick back.

You've been having starter problems. You've even said that sometimes your starter isn't even strong enough to push it through the compression stroke.

Ever kick start a Harley? I can tell you from long personal experience with a '76 Ironhead Sportster that if you don't generate enough momentum to carry it past TDC, the least bad scenario is where the kickback sends you over the handlebars. The idle advance setting on that vintage of Sportster is only about 10 degrees BTDC.

The combustion event on an Lyc begins 25 degrees before top dead center. If your starter can't push against the beginning of the explosion long enough to insure that the bulk of the energy is expended on the opposite side of top dead center -- then it's going to kick back. And the closer to "correct" that the fuel mixture is, the more likely it is to do it.

Personally, I wouldn't start that engine again until I had a starter that could make it spin until the battery runs down.
 
Well, our skytec NL starter works fine, but thats on a standard compression carbed engine. THe Skytec HT work GREAT on fuel injected engines I know. (Experience on TIO-541s)
 
jonbakerok said:
Pardon the ignorance of a carburator guy who has never owned a FI engine, but I have a real hard time imagining that simply leaving the mixture open would cause kickback. I can see how it might flood it, but simple logic inclines me to think that a rich mixure would be less likely to fire and therefore less likely to kick back.

You've been having starter problems. You've even said that sometimes your starter isn't even strong enough to push it through the compression stroke.

Ever kick start a Harley? I can tell you from long personal experience with a '76 Ironhead Sportster that if you don't generate enough momentum to carry it past TDC, the least bad scenario is where the kickback sends you over the handlebars. The idle advance setting on that vintage of Sportster is only about 10 degrees BTDC.

The combustion event on an Lyc begins 25 degrees before top dead center. If your starter can't push against the beginning of the explosion long enough to insure that the bulk of the energy is expended on the opposite side of top dead center -- then it's going to kick back. And the closer to "correct" that the fuel mixture is, the more likely it is to do it.

Personally, I wouldn't start that engine again until I had a starter that could make it spin until the battery runs down.
Click to expand...

Nope, the combustion event occurs during the start at 0 degrees, thanks to the impulse coupler (Mags). All bets are off with electronic ignition, because they seem to have a mind of their own.
 
Mixture setting doesn't make much sense. I agree with Jon. Propeller inertia is a huge buffer for the starter. If your starter/battery/wiring setup isn't generating a reasonable rotational velocity, then you have little inertia to counter any kickback.

Which brings up a new question: metal, wood, or composite prop?
 
Light Prop

DanH said:
Mixture setting doesn't make much sense. I agree with Jon. Propeller inertia is a huge buffer for the starter. If your starter/battery/wiring setup isn't generating a reasonable rotational velocity, then you have little inertia to counter any kickback.

Which brings up a new question: metal, wood, or composite prop?
Click to expand...
The prop only weighs 14 pounds - Carbon Fiber over Maple and because of the P-tips, it's a little shorter than it would otherwise be: 68" maybe? Also, there is very little mass at the tips. I'm beginning to think that is part of the problem. It's just not a very good flywheel.

I verified that my mags are timed together and that the right mag with the impulse coupler is firing at zero degrees advance. I verified that my mag timing is 25 degrees as called for by Superior.

I will be getting the starter back from Sky-Tec Monday, so I'll have more to report Monday night or Tuesday night. I am very grateful for all the posts and PM's on this. Some have suggested the starter may be inadequate and that will certainly be the next place to look if changing starting technique does not do it.

I am as surprised as anyone about the starting technique, but it makes sense after I consider it for a while. I don't recall every having a kick-back when using "hot start" technique. Also, I remember when we were first trying to start the engine that it started before we expected it to, with the mixture in ICO position. It's amazing the things I've ignored (that event, the instructions, etc.). My dad always said we get too soon old and too late smart.
 
Perhaps try retarding the mag timing a degree or two to make the event just a touch later. If something is a bit screwy with an impluse coupling that might show it up. This is a shot in the dark, but remember that there are 20? impluse couplings and 25?, is it possible that you might have the wrong impluse on your Left mag? IF that happened theoretically your spark might be going off at 5? before TDC instead of at TDC.

Easy check for this would be to remove all your bottom (or top...) plugs and turn the prop by hand, noting the exactly location when the impluse clicks, and seeing if that is actually at zero...
 
How an impulse coupling works

osxuser said:
Perhaps try retarding the mag timing a degree or two to make the event just a touch later. If something is a bit screwy with an impluse coupling that might show it up. This is a shot in the dark, but remember that there are 20? impluse couplings and 25?, is it possible that you might have the wrong impluse on your Left mag? IF that happened theoretically your spark might be going off at 5? before TDC instead of at TDC.

Easy check for this would be to remove all your bottom (or top...) plugs and turn the prop by hand, noting the exactly location when the impluse clicks, and seeing if that is actually at zero...
Click to expand...
Actually, the impulse coupling does not retard the timing by a fixed number of degrees. It delays the timing by a fixed time delay. Read post #4 at this link.

Also, see the interactive Flash magneto tutorial. You can see how the impulse coupling works by clicking on the "components" icon at the top, and then selecting "operation".

The impulse coupling time delay is only correct if the starter spins the prop at the assumed speed. If the starter spins the engine too fast, the spark will fire past TDC. If the starter spins the engine very slowly, or isn't able to get it past TDC, the spark will fire before TDC. That impulse coupling lag angles on some engines vary depending on which starter is installed. More info.

Maybe the cause of the kickback is because the engine isn't spinning fast enough.
 
I think Kevin hit the nail on the head and explains why the kickback on my engine stopped after replacing the Skytec starter with a B&C starter.
 
More questions than answers

Using a device borrowed from an IA/DME that has lights and a tone, I verified once again that my timing is correct - certainly not over-advanced. And, with all respect to Kevin, no matter how slowly I turn the prop, the impulse coupler fires at exactly the same spot which is as close to zero as I can measure with a probe into the lower sparkplug hole on #1.

When I rotate the prop manually the impulse makes its first click/thunk noise about where 25 degrees would be and then "ignites" at zero. Since my starter is out for repair, I'm using a fixed mark on a rigidly mounted vertical stick to identify zero and a digital angle indicator to measure advance.

Monday PM the starter should be back on and the IA/DME will be visiting to help me go over the whole system again. I'll let you all know if I find anything. I've also written to B&C to ask what the kick back protection is on their product.

Thanks again, everyone. You have given me a lot to think about and to check on. My original purpose was to help somebody else, but this is OK, too!
 
hevansrv7a said:
Using a device borrowed from an IA/DME that has lights and a tone, I verified once again that my timing is correct - certainly not over-advanced. And, with all respect to Kevin, no matter how slowly I turn the prop, the impulse coupler fires at exactly the same spot which is as close to zero as I can measure with a probe into the lower sparkplug hole on #1.

When I rotate the prop manually the impulse makes its first click/thunk noise about where 25 degrees would be and then "ignites" at zero. Since my starter is out for repair, I'm using a fixed mark on a rigidly mounted vertical stick to identify zero and a digital angle indicator to measure advance.
Click to expand...
Interesting. I wonder if maybe different brand mags have impulse couplings that work in a different way. I.e. is it possible that brand X have impulse couplings that provide a fixed amount of retard in degrees, and brand Y have impulse couplings that provide a fixed time delay. I don't claim to be an expert on mags - I only know what I have read in various places, and it is always hard to separate fact from fiction.

heavens7 and Randy Pflanzer - what brand mags do you have?
 
Impulse

hevansrv7a said:

I verified that my mags are timed together and that the right mag with the impulse coupler is firing at zero degrees advance. I verified that my mag timing is 25 degrees as called for by Superior.
Click to expand...

I assumed that the impulse coupler would generally be fitted to the Left Mag, but your note uses the term "right mag". Did you really mean right side, or did you use the term "right" to mean the "correct" mag.

Just checking.

Hawkeye Hughes
Skyote RV-3
 
Right as in Correct

REHughes said:
I assumed that the impulse coupler would generally be fitted to the Left Mag, but your note uses the term "right mag". Did you really mean right side, or did you use the term "right" to mean the "correct" mag.

Just checking.

Hawkeye Hughes
Skyote RV-3
Click to expand...
Hey, HH, good initials (N17HH for reference). I meant "correct" and you are right that right is ambiguous. It was wrong before, but now it's right, on the left. That's why I said "again" in the start of the thread; that's how I broke it the first time.
 
I have Slick Mag w/impulse coupler on the left. LightSpeed EI on the right, which is OFF when I start.
 
Yeah, but..

Yukon said:
Nope, the combustion event occurs during the start at 0 degrees, thanks to the impulse coupler (Mags). All bets are off with electronic ignition, because they seem to have a mind of their own.
Click to expand...

You're right. Forgot about that. But I think my original point is still valid. Even if it fires at exactly 0 degrees, it still has to be turning fast enough to make sure the energy gets expended on the far side of 0 degrees. By definition, if it kicks back, either something is causing it to fire too early or the starter isn't doing the job. He's already eliminated all the fire too early scenarios, while at the same time he's said that sometimes the starter can't push it past TDC. Gotta be the starter. From his last post, it sounds like a new one is on the way.
 
For whatever it's worth:

Both Bendix style (Now TCM Alpha) and Slick mags have fixed degree impulse couplings. I'm had both in my hand teaching people how they work at various times in the past two months. Both have a spring, a catch and a counterweight. When the the catch catches, which causes the spring to wind. At the fixed degree point, the catch gets released by a tab on the coupling body and the coupling then releases the wound spring and spins the mag. I don't know if there are 'timed' impulse couplings, but I've never seen one on an aircraft.

Right vs. Left mag: Main thing you need to verify, if you have a key switch, is that you mag with the impulse is wired to the L terminal on the switch. The right terminal gets grounded out and the left alone is used for starting. If you have individual toggles, obviously check your p-leads to make sure they are going to the correct mag. It is standard for the Left mag to be the impulse one, and for you to start that way, but it's not required. If you have the impulse in the right, you need to either: use toggle switches or wire the key switch backwards.

Right vs. Left mag is always determined by the view from the accessory case of the engine. In other words, sit in the pilots seat, and the one that is on your side, is the left, the one on the passengers side is the right.

A few questions. Skytecs tend to spin the engine pretty quickly, but if the electrical system or battery is marginal, they go to lagging pretty fast.

Which skytec are you using?
Which battery do you have?
Where is it mounted?
Is you engine high compression at all?
Does the starter lag on the compression stroke much when it's turning?
 
osxuser said:
For whatever it's worth:

Both Bendix style (Now TCM Alpha) and Slick mags have fixed degree impulse couplings. I'm had both in my hand teaching people how they work at various times in the past two months. Both have a spring, a catch and a counterweight. When the the catch catches, which causes the spring to wind. At the fixed degree point, the catch gets released by a tab on the coupling body and the coupling then releases the wound spring and spins the mag. I don't know if there are 'timed' impulse couplings, but I've never seen one on an aircraft.
Click to expand...
I wonder how long it takes from the moment the spring is released, until a spark is generated? There has to be some time delay here, as it is a mechanical device. I wonder if this delay is long enough to be significant. If the cranking speed is 100 rpm, every 1 millisecond of delay represents 0.6 degrees of crank rotation. It must take several milliseconds for the spring to get the mag spun up to speed.

Maybe the best way to describe the impulse is it delays the spark by X degrees plus Y milliseconds. The actual angle that the crank is at when the spark fires will depend on how much it rotates during the Y milliseconds the impulse device takes for the spring to spin the mag up and generate the spark.
 
Bus Bar

Heaven,

I can't be 100% sure this really matters, but do you have 2 pieces of copper bus bar material between the master contactor and the starter contactor?
Test ran mine with 1, and it seemed to engage a little slower than I would have liked. Now with both pieces installed as per plans, engages quickly.
 
Kevin Horton said:
I wonder how long it takes from the moment the spring is released, until a spark is generated? There has to be some time delay here, as it is a mechanical device. I wonder if this delay is long enough to be significant. If the cranking speed is 100 rpm, every 1 millisecond of delay represents 0.6 degrees of crank rotation. It must take several milliseconds for the spring to get the mag spun up to speed.

Maybe the best way to describe the impulse is it delays the spark by X degrees plus Y milliseconds. The actual angle that the crank is at when the spark fires will depend on how much it rotates during the Y milliseconds the impulse device takes for the spring to spin the mag up and generate the spark.
Click to expand...
Yep, that could be, but if when the crank is still Y=0, the mag fires at TDC, then any faster rotation of the crank would bring the spark later... :cool:
 
Interim update

Thanks to everyone who has been so helpful. Here's an update. I hope this ends up helping others, too.


  • Reinstalled the now repaired (again) Sky-Tec starter.

  • Re-timed both mags; they were about 2 degrees advanced beyond the 25 that's specified by Superior. They are now almost perfect. Of course, I thought they were OK before.

  • Discovered the engine came with two impulse coupled mags and I'm only using the left one. Originally I was inadvertently using the right one. Shouldn't have mattered. This changes a lot; the first broken starter was therefore not caused by my switching the wires. So now I'm looking for the same cause for both breakages.

  • Superior recommend using both mags for starting and all I have to do for that is to remove the little jumper on the back of the Van's ignition switch that's on the right-side magneto terminal. Waiting to do that until...see below.

  • Checked where the impulse coupled event takes place. One is slightly before TDC and the other is slightly after TDC and the spread between them is 5 or more degrees. The mags are specified as a lag of 20 degrees which means that if perfect they would click at 5 degrees BTDC. I assume this design incorporates an assumption about rotational speed during starting (see Kevin Horton's post about that). The relationship between the pawl and the body of the mag is fixed by a drive key on the shaft. The lag is not adjustable, therefore. But see below.

  • Asked Superior and they explained that the E-Gap being different will cause the timing lags on impulse mags to vary because the body of the mag is rotated a little differently to compensate for that when setting the timing. I won't know until tomorrow if these numbers are within acceptable ranges or if I have to reset the E-Gap. While there is no direct adjustment of the impulse lag, the E-Gap has the same effect. It is clear that the timing changed since first flight and normal wear between the cam and the (soft) follower is a good candidate for blame. Changing the E-Gap changes the timing, just like on the old fashioned cars with points.

  • My A&P/IA/DME advised me to get a different starter. Got a long reply from T. Hedding at B&C, some of which was on point and helpful. I'll probably buy one.

  • Will be checking my battery and starting circuit voltages just to be sure, but the battery is the Concorde 25XC and has only 70 hours so it will be a surprise if it is the problem. The connection between the master relay and the starting relay is a thick cable that came with Van's wiring harness, not a copper bar. There is no measurable resistance in the circuit when there is no load. All the same, it is common for my first attempt at a cold start to stall out on the compression stroke and then to need a second twist of the key to "start" position.
 
Here's a good way to check your electricals. What we want to do is check the voltage drops across each connection from the battery posts to the starter. For this you need an analog voltmeter, since the voltage will vary and it is hard to read a rapidly changing DVM. First off, disable the ignition, EI or mags. Then with the assistance of a friend or two, connect the VM red lead to the battery POST, and the black lead to the connector on the post and enegize the starter, noting the VM reading. Now leaving the R lead on the battery POST, connect the B lead to the cable next to the connector and note the VM with the starter on. Now move the B lead to the cable next to the next connector and repeat. Now the B lead to the adjacent connector. Continue this process all along the plus run with the last reading at the starter input post. Now you have all of the voltage drops across each connection, and the total drop from the battery + to the starter. Now you need to repeat the same thing, starting with the B lead on the battery negative terminal, and measure to each connection with the R lead, with the final reading from the negative post to the starter case. If you have any significant voltage drops on any connection, they will immediately become apparent, plus you will see the total drop from battery to starter.
 
Wild Goose Chase

hevansrv7a said:
Thanks to everyone who has been so helpful. Here's an update. I hope this ends up helping others, too.


  • Reinstalled the now repaired (again) Sky-Tec starter.

  • Re-timed both mags; they were about 2 degrees advanced beyond the 25 that's specified by Superior. They are now almost perfect. Of course, I thought they were OK before.

  • Discovered the engine came with two impulse coupled mags and I'm only using the left one. Originally I was inadvertently using the right one. Shouldn't have mattered. This changes a lot; the first broken starter was therefore not caused by my switching the wires. So now I'm looking for the same cause for both breakages.

  • Superior recommend using both mags for starting and all I have to do for that is to remove the little jumper on the back of the Van's ignition switch that's on the right-side magneto terminal. Waiting to do that until...see below.

  • Checked where the impulse coupled event takes place. One is slightly before TDC and the other is slightly after TDC and the spread between them is 5 or more degrees. The mags are specified as a lag of 20 degrees which means that if perfect they would click at 5 degrees BTDC. I assume this design incorporates an assumption about rotational speed during starting (see Kevin Horton's post about that). The relationship between the pawl and the body of the mag is fixed by a drive key on the shaft. The lag is not adjustable, therefore. But see below.

  • Asked Superior and they explained that the E-Gap being different will cause the timing lags on impulse mags to vary because the body of the mag is rotated a little differently to compensate for that when setting the timing. I won't know until tomorrow if these numbers are within acceptable ranges or if I have to reset the E-Gap. While there is no direct adjustment of the impulse lag, the E-Gap has the same effect. It is clear that the timing changed since first flight and normal wear between the cam and the (soft) follower is a good candidate for blame. Changing the E-Gap changes the timing, just like on the old fashioned cars with points.

  • My A&P/IA/DME advised me to get a different starter. Got a long reply from T. Hedding at B&C, some of which was on point and helpful. I'll probably buy one.

  • Will be checking my battery and starting circuit voltages just to be sure, but the battery is the Concorde 25XC and has only 70 hours so it will be a surprise if it is the problem. The connection between the master relay and the starting relay is a thick cable that came with Van's wiring harness, not a copper bar. There is no measurable resistance in the circuit when there is no load. All the same, it is common for my first attempt at a cold start to stall out on the compression stroke and then to need a second twist of the key to "start" position.
Click to expand...

Hey Guys, I've been following this discussion since my new, (and as of yet uninstalled) engine has a SkyTec starter. This evening I came across an article on page #261 of the "24 years of the RVator" about wiring a SkyTec. Basically, it states that permanent magnet starter momentarily acts like a generator once the engine starts. This generator effect holds the solenoid in the engaged position. They recommend a different solenoid. Any chance this is the culprit?
 
Hmm, I'm not going to claim to be an expert on this nuance but... I was under the distinct impression that the impulse coupling is suppose to advance the timeing TO TDC not 5? BTDC. I'm going to try to confirm that tomorrow with a couple lycoming's I have around on airplanes. Do the mags say on the dataplate 25? or 20?? If they say 20?, that is where they should be timed to the engine at. All very interesting, I really think there is something else going on here besides the Skytec... I've installed too many of those and seen them function perfectly to buy into that.
 
more

osxuser said:
Hmm, I'm not going to claim to be an expert on this nuance but... I was under the distinct impression that the impulse coupling is suppose to advance the timeing TO TDC not 5? BTDC. I'm going to try to confirm that tomorrow with a couple lycoming's I have around on airplanes. Do the mags say on the dataplate 25? or 20?? If they say 20?, that is where they should be timed to the engine at. All very interesting, I really think there is something else going on here besides the Skytec... I've installed too many of those and seen them function perfectly to buy into that.
Click to expand...
The plate on the mag says 20 degrees. Superior says 25 for the mag at normal RPM (static). That leaves 5 BTDC. I confirmed this in one conversation last night with Superior. There's another conversation scheduled for this morning. I always thought the zero was the correct setting, too.

Caveman said:
Hey Guys, I've been following this discussion since my new, (and as of yet uninstalled) engine has a SkyTec starter. This evening I came across an article on page #261 of the "24 years of the RVator" about wiring a SkyTec. Basically, it states that permanent magnet starter momentarily acts like a generator once the engine starts. This generator effect holds the solenoid in the engaged position. They recommend a different solenoid. Any chance this is the culprit?
Click to expand...
When it happened it was before the engine started, but in the first event, the one that destroyed the ring gear, it actually ran backwards a little. I wonder what kind of solenoid they mean? I've never noticed any problem with disengaging, just with hesitation and kick-back.

If the Sky-Tec is the problem, I think it's only a problem in combination with other factors including the very light prop with little mass at the tips. But it's just a theory at this point. Nothing I've found so far is by itself enough to cause the problem. That's often the case with aviation, though, isn't it?
 
Ok, well here is my take. On the O-360A1A, O-320 A2B, O-290G that I work on regularly and every other Lycoming i've maintained in 5 years as an A&P, the 25? timing position only works if you have 25? retard impulse couplings. On some IO-360's there are 20? impulse couplings, and those get set to 20? BTDC. Obviously I am not say to go against what you engine builder says, BUT I am pretty sure this is your problem. On the O-290 that I currently work on, with a light 3-blade wood-composite prop, the prop tried to kick back a couple times, diagnosis found the mag switch wired backwards to the non-impulse mag. So I have seen what you describe, and it almost certainly seems to be the same timing related issue. Just less severe because of only 5? off instead of 25? like the O-290.

Here comes the real question though. Superior wants the mags timed at 25? BTDC, and your mags have a 20? retard on the impulse, are you sure you have the right mags? Normally the 20? ones go on things that have Turbos and stuff like that. If your engine was super high-compression, then retarding the timeing a little might make sense (timing them to 20? instead of 25?) but if you retime against superior's guidance now, you are likely to lose a bit of power.

What I'd start with is retiming to 20? and seeing that the kickback problem goes away. If/When it does, then start looking for a set of 25? retard mags. That way you go with Superiors guidance, and don't have the problem any more.

Does anyone else out there have 20? retard mags running at 25??
 
Most 320 and parallel valve 360's that are timed to 25 BTDC use either the 4371 (20 degree lag) or the 4373 (25 degree lag) mags. Some are approved for both and some for one or the other. Common experimental aircraft engines, the O-360-A1A, the O-320-D2A and the IO-360B series are all approved to use either the 4373 or the 4371. I can't say that I have ever seen a difference in starting characteristics, using either on a 25BTDC timed engine. Maybe with a light weight prop, it would make a difference but with aluminum prop either CS or fixed pitch I can't say that either performs better then the other and have never seen kick back from either mag when installed, timed, wired and working correctly. . I don't want to start a war here but I think you guys are barking up the wrong tree, unless having a fiberglass prop makes a big difference with the applicability of either lag angle mag; either the 4373 or the 4371 should work fine on any 25 BTDC timed engine. Or at least they have in my experience.
See latest revision of Lycoming SI 1443 for Slick mag applicability.
Good Luck,
Mahlon
"The opinions and information provided in this and all of my posts are hopefully helpful to you. Please use the information provided responsibly and at your own risk."
 
Yep! wild goose chase

hevansrv7a said:


I wonder what kind of solenoid they mean? I've never noticed any problem with disengaging, just with hesitation and kick-back.

If the Sky-Tec is the problem, I think it's only a problem in combination with other factors
Click to expand...


They give a part #ES 24021
The article was 1st written in 1998 with a suggested fix on the original solenoid (removing a jumper) and then updated in 1999 to suggest the new one due to possible damage to the mag switch.

Sounds like I led you astray then, I apologize. They noticed a sound like the key was not released on engine start that alerted them to this issue.
 
Today's Progress

Today's progress: There is no single cause; a number of factors seem to have accumulated.

I received generous, knowledgeable help from: Fred and Evan at Superior, Tim at B&C, Gene at Sky-Tec, Skip at Concorde and someone at Unison.

Much to my surprise, my Concorde RG25XC battery is not strong enough. This was determined by watching an analog VOM while cranking. Initial cranking voltage was around 5 volts and then went up to 10V as the blades turned a few revolutions. The good news is there is no significant voltage drop between the battery and the starter or between the battery and the ground. It should not drop anywhere near that low on the first blade/ compression stroke.

The Sky-Tec troubleshooting guide, version 3.0 that came with my starter has some scrambled information and is thus incorrect for what voltage reading means what diagnosis on the starter. Made a note.

Concorde suggested an overnight charge at a 10A rate with my Die-Hard automotive charger which they say is a good charger for this battery. Skip thinks it's sulphated. If that does not restore it they will replace it ASAP. Good folks!

The Sky-Tec starter(permanent magnet) takes more initial current than a serial-wound even though it takes less current after it gets going. Combine a higher initial demand with a weak battery and you have the problem I've been seeing of the blades stopping. It's my feeling that this condition contributes to the chance it will kick-back. A B&C starter would work better in this situation, but the battery is the problem, maybe not the starter.

The 20 degree lag spec. is not totally determinative of the difference between where the impulse spark happens and where the normal spark happens. It's just the geometric difference between things inside the mag. The E-Gap (the dwell for you old time shade tree mechanics) and the position of the mag on the engine taken together determine where the impulse mag fires on starting. One of mine is firing a little after TDC and one is firing a little before. Unison strongly recommended only using the one that fires later. My regular timing is set at 25 BTDC on both. So even though Superior recommends using both, unless I can get the other one ATDC, I will only use the one.This is controlled with the jumper on the back of the switch.

Unison also said that Lycoming (analagous to Superior) allows plus or minus two degrees on the 25 degree advance specification. He suggested if necessary retarding the spark setting to 23 BTDC to ensure the impulse fired ATDC. He also suggested I could use another model magneto (mine are 4371) with a 25 degree lag. That's a bunch of $$ of course.

Gene at Sky-Tec suggested that with a very light prop like mine that a separate starter switch would allow getting the engine spinning before lighting up the magnetos. With my Van's ignition switch I can't do that. I could, though, rig an additional switch just for starting. The downside of this is I would need three hands for a good start. However, I can ensure the engine is in ICO with no prime and use hot start technique with a cold start. I don't know if that will work, but it would help prevent kick-back.

The WX looks good for tomorrow, so I will probably fly it again. I hope all this brings the problem under control.
 
hevansrv7a said:
Today's progress: There is no single cause; a number of factors seem to have accumulated.

I received generous, knowledgeable help from: Fred and Evan at Superior, Tim at B&C, Gene at Sky-Tec, Skip at Concorde and someone at Unison.

Much to my surprise, my Concorde RG25XC battery is not strong enough. This was determined by watching an analog VOM while cranking. Initial cranking voltage was around 5 volts and then went up to 10V as the blades turned a few revolutions. The good news is there is no significant voltage drop between the battery and the starter or between the battery and the ground. It should not drop anywhere near that low on the first blade/ compression stroke.

The Sky-Tec troubleshooting guide, version 3.0 that came with my starter has some scrambled information and is thus incorrect for what voltage reading means what diagnosis on the starter. Made a note.

Concorde suggested an overnight charge at a 10A rate with my Die-Hard automotive charger which they say is a good charger for this battery. Skip thinks it's sulphated. If that does not restore it they will replace it ASAP. Good folks!

The Sky-Tec starter(permanent magnet) takes more initial current than a serial-wound even though it takes less current after it gets going. Combine a higher initial demand with a weak battery and you have the problem I've been seeing of the blades stopping. It's my feeling that this condition contributes to the chance it will kick-back. A B&C starter would work better in this situation, but the battery is the problem, maybe not the starter.

The 20 degree lag spec. is not totally determinative of the difference between where the impulse spark happens and where the normal spark happens. It's just the geometric difference between things inside the mag. The E-Gap (the dwell for you old time shade tree mechanics) and the position of the mag on the engine taken together determine where the impulse mag fires on starting. One of mine is firing a little after TDC and one is firing a little before. Unison strongly recommended only using the one that fires later. My regular timing is set at 25 BTDC on both. So even though Superior recommends using both, unless I can get the other one ATDC, I will only use the one.This is controlled with the jumper on the back of the switch.

Unison also said that Lycoming (analagous to Superior) allows plus or minus two degrees on the 25 degree advance specification. He suggested if necessary retarding the spark setting to 23 BTDC to ensure the impulse fired ATDC. He also suggested I could use another model magneto (mine are 4371) with a 25 degree lag. That's a bunch of $$ of course.

Gene at Sky-Tec suggested that with a very light prop like mine that a separate starter switch would allow getting the engine spinning before lighting up the magnetos. With my Van's ignition switch I can't do that. I could, though, rig an additional switch just for starting. The downside of this is I would need three hands for a good start. However, I can ensure the engine is in ICO with no prime and use hot start technique with a cold start. I don't know if that will work, but it would help prevent kick-back.

The WX looks good for tomorrow, so I will probably fly it again. I hope all this brings the problem under control.
Click to expand...
Glad you are getting down to answers. The skytec troubleshooting guide is gold if you understand it. One solution for the starter button would be placing it right above/around/on your throttle/mixture/quadrent. That way you have a hand on the throttle engaging the starter, and the other hand to flip ont he mags. I like that idea from the standpoint that you can pick which mag you flip on first, and it's a conscience decision every time, as opposed to being forced into the same one every time.

Good luck!
 
Progress: Improved or Fixed

It is not possible to prove a negative, so I cannot be sure it's fixed, but things are working better.

I did charge the battery using Die Hard 50A charger set on 10A overnight and it read 12.8 volts after about 10 minutes off the charger. Just before starting the engine, my EIS said 12.4 volts. This is not necessarily directly comparable, so I don't know if the voltage dropped. I was not able to reach Skip at Concorde today so I don't yet know about getting a new battery or if the old one is now OK.

I note in passing that the expert at Concorde says that Die Hard charger is excellent for his battery. I'm glad I didn't spend more money for a specialized charger.

I used the recommended starting method and even modified it a little to ensure safety.



  1. Throttle 1/4 of travel, mixture ICO.
  2. Pressurize system to about 16 pounds. (Just a couple seconds pumping)



  3. Push mixture quickly to full rich and quickly return it to ICO.
  4. Engage starter while pushing mixture in smoothly (2 seconds?)
  5. This worked to start engine with no kick back and the prop did not hang up on the first position as it often did in the past. It does take more cranking than before.
I think that this technique accomplishes about the same thing as a separate starter switch and only requires two hands.

We then flew about 30 minutes, had lunch, dawdled and then prepared to return home. The WX was near 90 and sunny. I followed the above procedure except that I omitted #3. It tried to start on the first try but died out. I repeated from #4 and it started OK.

No grinding, no kickback, no blade stops while starting. Life is good. The pie at Napoleon Cafe at 3NP is worth the trip.

In the end, I believe, subject to new information later, that it was a combination of factors: weak battery, light prop, wrong technique, slightly advanced timing (2 degrees more than specs) and a starter that wants more power for the startup moment than the other type. I don't think any one of these would have done it but all of them together clearly did. Many replies have noted that they used the same starter or the full rich start technique and never had a problem. This may help explain that.

Thank you all again for all the help. I note that many more folks read than replied and that is good. I hope that the RV community has thus gained. Case closed for now.
 
Update - New Starter & Plugs

For a while all was good. Then I started getting two inter-related problems again, but not as severely.
1. Starter stalled at first compression stroke. Which led to..
2. Starter teeth grinding against ring gear - possibly mild kickback.

I got sick of it. I knew the battery was not at fault. I knew the wiring was not at fault. I was already pre-positioning the prop for maximum momentum as the rotation begins before the compression point. I don't want to change the ring gear again. Not much left but...

Today I replaced the Sky-Tec with one from B&C. Thanks to Randy and others for this suggestion. I haven't flown since installing it but quick test in hangar with ignition shorted out is very encouraging. I had to make a small change in the link between the alternator and the starter but otherwise it was a clean swap. The new starter is a little heavier but the added weight is way up front, so that's good.

Last week I installed Autolite Fine Wire spark plugs. This was more for problems that I was having in trying to run LOP with low power settings, but it helps a little with starting on my standard magneto(s), too. It did help with LOP operations, too.
 
How has the B&C starter worked out for you with your 9:1 compression Superior engine? Any issues with how the B&C engages your ring gear teeth? Any kickbacks?

Chad
 
I have had the same B&C starter on a 9.3:1 IO-320 for 26 years and 2,800+ hours and never had a problem with the same starter or ring gear. Built 5 experimentals and used 5 B&C starters - their alternators and regulators too.
 
I've always had great luck with B&C starters as well, but I have a friend with an RV-7A. He had a B&C mounted on his Superior IO-360 and he kept breaking ring gear teeth. After replacing it twice, he switched to a Skytec and had no further troubles. He was beginning to wonder if the B&C for whatever reason didn't like the Superior engine.
 
You must log in or register to reply here.
Back
Top