Dual P-MAG Reliability in 2017

[Captain Avgas]

Originally Posted by KRviator

So you've never had, nor heard of, a magneto losing its' timing in flight? A quick Google search shows it is far from rare with traditional magnetos indeed occurring 3 times in a 6-week SDR search

It's not that I'm sceptical, but I think you should state your search criteria and produce the specific case results of your SDR search so that we can ascertain for ourselves if they are valid to the discussion at hand.

[maus92]

Originally Posted by rv6ejguy

We're trying to eliminate as many external switches on the CPI2 as possible. I can't say at this point until software and testing is closer to completion. That's probably at least a month away.

Is the newer CPI2 a completely new system, or is it basically an upgraded controller / interface box / display? Is everything firewall forward the same?

[rv6ejguy]

Originally Posted by maus92

Is the newer CPI2 a completely new system, or is it basically an upgraded controller / interface box / display? Is everything firewall forward the same?

The PCB, case, display will be all new, all FF components the same as before. I can't talk about some of the other new features planned at this time until we're closer to release.

[maus92]

Originally Posted by rv6ejguy

The PCB, case, display will be all new, all FF components the same as before. I can't talk about some of the other new features planned at this time until we're closer to release.

I guess what I'm getting at is there an upgrade path for those who either have installed the current system, or want to install the current version sooner rather than waiting for CPI2?

[N941WR]

Originally Posted by lr172

This is completely incorrect. Auto manufacturers have complete control of their ignition maps and spend 1000's of hours on the dyno optimizing fuel flow (i.e. mixture) and ignition timing. It is also done in harmony with other engine variables, such as temps. Granted, they also factor in reliability and emissions. However, they are highly optimized. They sell cars, in part, based upon HP claims. Trust me, they want as high of a number as they can get.

Timing makes a big impact on performance and even though many here don't want to hear it, it is variable based on many factors, the most profound being mixture. I can't imagine having an EI that cannot be adjusted based upon my mixture setting (ROP vs LOP). It would be like not having a mixture control and just feeding full rich through the flight regime.

I do agree with Bill that if you favor simplicity over performance, the Pmag is a good upgrade option over the mag. However, I would still buy the CPI for cost and reliability.

Just to give one example of the benefits of flexibility, I solved my FI hot idle problem by bumping my idle timing to 35.

Larry

The this is completely incorrect.

While it is true that auto manufacturers spend a lot of time developing timing and fueling maps, they do not tune them to extract maximum performance out of your vehicle. They are looking for emissions, reliability, and performance, not necessarily in that order.

When I was doing the auto racing thing, I replaced the ECU and was able to pickup around 5% more torque and HP by tuning the car on a dyno. Of course it was tuned within a breath of its life, something auto makers are loathe to do for obvious reasons.

Another way to look at it is, why are there so many different aftermarket automotive ignitions available, if the auto manufacturer's ignitions are so good?

My biggest issue with multi component systems is the number of connectors that can fail. That is not to mention the additional complexity required by adding extra batteries and alternators to protect against electrical system issues.

Then you still have the problem one LS user had a few years back. He was on short final when his alternator belt broke, ripping out the wires for the crank angle sensor. Instant glider.

In the end, you as the builder have to weigh the risks, both set up and operational, and pick the system that you feel is best for you.

[rv6ejguy]

Originally Posted by N941WR

The this is completely incorrect.

While it is true that auto manufacturers spend a lot of time developing timing and fueling maps, they do not tune them to extract maximum performance out of your vehicle. They are looking for emissions, reliability, and performance, not necessarily in that order.

When I was doing the auto racing thing, I replaced the ECU and was able to pickup around 5% more torque and HP by tuning the car on a dyno. Of course it was tuned within a breath of its life, something auto makers are loathe to do for obvious reasons.

Another way to look at it is, why are there so many different aftermarket automotive ignitions available, if the auto manufacturer's ignitions are so good?

My biggest issue with multi component systems is the number of connectors that can fail. That is not to mention the additional complexity required by adding extra batteries and alternators to protect against electrical system issues.

Then you still have the problem one LS user had a few years back. He was on short final when his alternator belt broke, ripping out the wires for the crank angle sensor. Instant glider.

In the end, you as the builder have to weigh the risks, both set up and operational, and pick the system that you feel is best for you.

Almost all auto ECUs today use a knock sensor or CPT to advance timing just short of knock at all times for the given fuel octane plus they're taking into account IAT, CLT, MAP and/or MAF, baro etc. to ensure highest BMT at ALL times, whether cruising at part throttle well lean of Stoich or WOT at redline. Since emissions are not measured at WOT/ high rpm in most tests because most car engines don't spend a high percentage of time there, power is paramount under that condition given a reasonable buffer to the edge of detonation which ANY system should have. To suggest otherwise, especially in aircraft, is reckless.

The auto ignitions are orders of magnitude more sophisticated and precise than a Pmag, CPI, LS or anything else in aviation short of Lycoming's IE2 or perhaps the Rockwell system on the Rotax 912iS, which they easily match or exceed.

Any advance in timing will almost certainly affect knock margins and/or emissions. The aftermarket does not need to meet any emissions spec and most say "for off road use only". I don't see many aftermarket ignitions out there for the cars of today outside of Link, Haltech and Motec which are intended mainly for performance / race use at higher boost levels where factory ECUs are not mapped-that longevity/ warranty card again.

No auto OEMs use integrated ignition controllers these days because they found out long ago that long term reliability is compromised by heat and vibration when attached to the engine. The Pmag integrated design was a marketing decision to make installation easier IMO (which it undeniably is), not something based on best engineering practice.

No auto OEMs today use gear driven timing components because it's less precise and more prone to failure. These are reasons why Pmag mandates 100 hour inspections- to make sure the bearings, gears are ok and the electronics have not been overtemped or got wet. Pmag requires blast tubes for cooling as a result of their design decision and they use non-water proof screw type connectors externally. Really, in 2017? You don't see this stuff on modern auto ignitions.

Poor belt maintenance and questionable installation likely caused the LS incident in Australia. Anyway, we can all learn from that and protect crank sensor cables with armor as we recommend. No way a thrown belt will take out our sensor or cables.

The internal generator in the Pmag has caused issues as outlined in their own documentation. http://www.emagair.com/service-notes/ http://www.emagair.com/downloads/

More reliable than a backup battery? Doubtfully. More spinning, vibrating stuff inside to go wrong and cause damage to other critical parts.

[Toobuilder]

Edit, Ross beat me to it.

[Toobuilder]

Originally Posted by N941WR

...Another way to look at it is, why are there so many different aftermarket automotive ignitions available, if the auto manufacturer's ignitions are so good?....

Please define an "automotive ignition" so that we have context. Are you discussing something from the 1960's or today? In the 60's you had a true stand alone system: Distributor, amplifier, coil... Today, with the ECU handling many functions, receiving signals from many sensors, its hard to segregate an "ignition system". You can certainly buy aftermarket coils and fat spark plug wires just like in the old days, but a "system"? Please define.

Originally Posted by N941WR

...My biggest issue with multi component systems is the number of connectors that can fail....

Please cite actual failure rates "caused" by the use of "Multi component systems". Not what "could" happen, but what actual experience shows "does" happen. Compare that to the engineering challenge of cramming all that stuff into one box to satisfy a marketing niche and see where that gets you? Would you like to compare the history of "magnet misalignment " with Ross' product line? How about "lost timing events"?

I'll grant P-mag the benefit of the doubt that their issues are largely resolved (with frequent, repetitive inspections), but we all know that that has been a long, uphill battle - a battle LARGELY the result of the marketing decision to package it all together - But there is no way you can pound the drum that a "distributed" system is inherently less reliable. No way. To do so flies in the face of logic and absolutely overwhelming evidence to the contrary.

It's time to put that bogey man to rest.

[KRviator]

Originally Posted by Captain Avgas

It's not that I'm sceptical, but I think you should state your search criteria and produce the specific case results of your SDR search so that we can ascertain for ourselves if they are valid to the discussion at hand.

No problems, Google "Magneto losing timing lycoming" and it is the 2nd from the bottom. Of if you prefer, you can login to the CAsA SDR Portal and search using "Magneto" with no date range yourself. But the initial 3 I found were using Google.
From CAsA's Flight Safety magazine, 01August14-22Sep14:

Lycoming IO540K1A5 Magneto/distributor?magneto unserviceable. SDR 510019711 RH magneto timing found to be approximately 7 degrees advanced. Magneto had 94 hours TSN. P/No: 6350. TSN: 94 hours

Lycoming IO540K1A5 Magneto/distributor?magneto unserviceable. SDR 510019832 RH magneto timing moved to 4 degrees advanced. Magneto had approximately 50 hours TSN. P/No: 6350. TSN: 50 hours.

Lycoming O320 Magneto/distributor?magneto unserviceable. SDR 510019703 Magneto losing timing. Timing reset but changed again so magneto replaced. P/No: 4370. TSN: 477 hours.

However, using the SDR Portal, you find the following using no date range and the search term 'magneto':

Originally Posted by The CAsA SDR Portal

18 August 2016: On newly overhauled engine after relatively short time in service, magneto timing was advanced by more than 5 degrees. According to Slick SL No. 4300/6300-74-20-001, this amount of timing drift required immediate action. Magneto was therefore removed and returned for warranty claim.

23 December 2015: During a 100 Hourly Inspection, the magneto to engine timing had drifted 4 degrees which was corrected during the 100 Hourly Inspection. On the post 100 Hourly ground run, there was a 200 rpm magneto drop noted on the magneto that had the timing drift (the 200 rpm drop was not there before the timing drift on the pre 100 Hourly ground run). As this magneto drop is beyond limits the magneto was replaced with new, was timed to the engine and a ground run was carried out and found serviceable.

11 November 2015: Left hand engine wouldn't start. The magneto to engine timing was checked (as we have had previous magneto problems) and it was approx. 20 degrees advanced. The magneto was removed and bench checked and was founds approx. 20 degrees advanced with a weak spark below 1,000 rpm.

03 July 2015: LH magneto timing out of adjustment

29 May 2015: LH magneto timing drifting. Investigation found points cam loose. Further investigation found the rotor shaft cam slot had been machined oversize.

05 May 2015: Magneto has a history of the timing setting moving. See attachment for history details.

08 September 2014: RH magneto timing moved to 4 degrees advanced. Magneto had approximately 50 hours TSN.

03 September 2014: LH magneto faulty. Timing was found to be 8 degrees advanced.

02 September 2014: LH magneto faulty. Timing was found to be 10 degrees advanced.

13 August 2014: RH magneto timing found to be approximately 7degrees advanced. Magneto had 94 hours TSN

08 August 2014: Magneto losing timing. Timing reset but changed again so magneto replaced.

20 May 2014: Magneto losing timing settings after a few hours of flight

01 May 2014: Magneto timing advanced 6 degrees from20 degree mark. Found during inspection iaw Slick SB2-08B.

26 April 2014: Magneto losing timing settings after a few hours of flight.

04 April 2014: Magneto losing timing settings after a few hours of flight.

04 April 2014: Magneto unserviceable. Timing advances in use.

25 March 2014: Magneto losing timing settings after a few hours of flight.

13 March 2014: LH magneto timing slowly advances by 5-6 degrees between inspections. Magneto is correctly retimed at each inspection but then slowly advances again.

13 March 2014: RH magneto timing slowly advances by 5-6 degrees between inspections. Magneto is correctly retimed at each inspection but then slowly advances again. See also SDR 510018760 for similar defect on LH magneto.

11 March 2014: Magneto losing timing settings after a few hours of flight.

05 November 2013: On Engine Run-up performance check, right magneto 50 RPM drop. Timing 5 Degrees out.

02 May 2013: RH magneto failed to stay correctly timed. Suspect internal timing slipped due to unknown internal problem. Magneto had been fitted to a newly rebuilt engine.

01 February 2010: Magneto had excessive play in gear shaft bush causing gear to jump teeth and affect timing.

08 August 2008: Magneto internal timing shifted. Timing moved from 20 degrees BTDC to 12 degrees BTDC.

As you can see, PMags are not alone in the lost-timing issue and I dislike the fact that at least two qualified pilots do not understand basic magneto operation enough to understand they too suffer from timing issues - and they can be a relatively common occurrence too - that can affect engine operation, instead trying to point the finger at PMag as being the only possible culprit in this regard. Now, before anyone suspects I am affiliated with Brad or his company - I have no interaction or association with PMag/EMagAir or any of their employees, other than having two PMag L114's installed on my engine at manufacture by ECI.

These SDR's are a relatively new concept from CAsA, and only capture those that are reported, and only then in Australia, which has a relatively small GA fleet compared to the US, so it is certainly within the realm of possibility an American pilot would expect to experience such a phenomenon and be expected to react accordingly. Commentary here suggests this may not be the case, and if this is indeed true, I would encourage a mag-check to be incorporated as part of your IA's following engine roughness or failure and be practiced as such, at the earliest opportunity. As I said earlier, it was part of the spam-can checklists I was trained under and that is going back close to 20 years, and it is there for a reason.

[Malndi]

Originally Posted by rv6ejguy

Poor belt maintenance and questionable installation likely caused the LS incident in Australia. Anyway, we can all learn from that and protect crank sensor cables with armor as we recommend. No way a thrown belt will take out our sensor or cables.

I'm on the fence and watching the debate, however the risk of the crank sensor and wires being damaged by alternator belt failure is a key concern for me. Particularly given that I wasn't far from that crash and know the occupant who was very badly burnt. Would like to know more about how the crank sensor and cables can be protected, but didn't find the answer on your website. Can you elaborate here.