What's new
Van's Air Force

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

Experimenting With Ignition Timing

Toobuilder

Well Known Member
Much has been written on this forum concerning the pros and cons of electronic ignition and more specifically, the level of advanced timing they offer. Plenty of testimony that Lycomings "like" more timing than offered by the fixed magneto, and plenty of testimony that the factory knows best. There has also been dyno testing between the two, but the results have been less than convincing for either side. The current crop of electronic ignitions have fairly limited to zero adjustability and the curve seems somewhat mysterious. In my mind, what is needed is a completely adjustable ignition which you can tweak on the fly. Being able to set power, then move the timing around to see where the engine is happiest should put the issue to bed, shouldn't it?

Well, I have just installed an electronic ignition with that capability and I have done some preliminary testing. First off, the ignition is the new CPI product from Ross Farnham and it offers essentially an infinitely adjustable advance curve for RPM as well as a second curve based on MP. This system replaces one of the Slicks on my Rocket. On the first flight I set up a conservative advance essentially duplicating the 26 degree timing of the magneto and was able to replicate my typical cruise performance. Without touching anything but the ignition advance I was able to pick up a few knots of speed. The second flight was a data gathering flight to see if it was a fluke. The performance repeated.

Attached is a table transcribing the hand written notes from that flight. The method of testing was to get the airplane set up at cruise altitude, on autopilot and with a typical power setting (in this case, slightly LOP). I started at 26 degrees advance (duplicating the magneto timing) as my "baseline". Once stable, I would punch in some advance, let the ship stabilize, and then write down the results. Each change would stabilize in about 30 seconds, so I was able to cycle through the whole test in one 20 minute flight.

This is still very preliminary and some of the individual measurements dont track exactly as one would expect, but the overall trend is obvious. I thought it interesting enough to share. There will be much more testing shortly.

2vdirn8.jpg
 
Last edited:
One change begets another!

Well, looks like your CHTs are kinda cold - now you will have to fab up some way to close off the outlet to bring 'em up in cruise, but keep 'em cool in climb. ARRGHH! ;-)

195KTAS on 11.5GPH is pretty good for that airframe. What is your compression?

I gotta hear more about this system - I expect good things from Ross. Good luck with the testing!

Carry on!
Mark
 
Yes, way too cold now. All that work to keep it alive in the desert summer heat is really hurting me now. It was 16 degrees F this morning on the way to work - two months ago it was 116. This airplane needs a cowl flap!

Engine is a factory stock -D4A5 (260 HP).
 
Good stuff Mike.

May I suggest an experiment? Set up the same test as above (LOP cruise), run the four timing settings, then repeat them with the mixture richened to 100 ROP for the leanest cylinder.

If theory holds, the highest TAS will found with less advance...like 30~32 rather than 35.
 
How low should you go?

Like most of RVers I have been focused on managing the upper limits of CHTs. I have never given much thought to the lower end and was under the belief that 290 -310 were acceptable temperatures for cruise.

You now you have me thinking, "How low should you go?" I haven't been able to find any good data points on the implications of cruising with CHTs below 300 degrees.

What are the implications of cruising with low CHTs and how low should you go?

Data points, experience and opinions are appreciated.


Tks,
MD
 
Sure thing Dan, next time I fly. (Probably Friday)

Along those same lines (less advance needed with a rich mixture), I'm also going to deterrmine if the engine likes less than stock timing at TO power. That one is going to be tough though - flying along at 500 MSL at 99% power while messing with the ignition box is going to require some concentration. Probably best to head out to the coast... Dont need to tangle with power lines at 220 knots!
 
Your results are very similar to mine.

I have a single Bendix Mag and a Lightspeed ignition.

The engine is a IO-540C4B5 (Identical to yours except mine is narrow deck) with ~9.2:1 CR.

I have the optional potentiometer connected to the Lightspeed and can adjust the timing in flight.

I have found it much the same. My baseline timing is a little lower due to higher compression. I start at about 23 deg at max power and advance to around 29 degrees at the same cruise conditions that you tested. Any more timing than that and it starts to drop speed but as long as I am running 100LL it will not detonate with a lot more timing.

During the summer months when summer mogas was available with a 9 PSI RVP I ran 93 octane non-ethanol but had to limit the advance to around 26 deg due to higher compression. Now that the Mogas here has switched to winter (13 PSI RVP) gas I am back on 100LL.
 
...What are the implications of cruising with low CHTs and how low should you go?

Data points, experience and opinions are appreciated...

My understanding is that IC engines are heat difference devices. The more heat that turns to power is more efficiency. Generally, the hotter the better - as long as you stay within the thermal limits of the materials the engine is made of. You may recall the interest in ceramic engines some years ago. That was an attempt to get an engine to survive at much higher temperatures than available today. Consider also the evolution of the gas turbine engine. The basic architecture of jet engines has not changed all that much, but yet we have seen radical gains in output/efficiency with every new generation. This is significantly attributable to increasing TIT (turbine inlet temperature), or more specifically, the ability to keep the hot section parts alive.

Just my opinion, but I would be very happy if I could maintain a CHT of 350-375 in all phases of flight, year round. I also think my engine would be happy as well.

All that said, my sub 300 CHT is not due to the change to this ignition - it was doing this last winter as well. LOP ops and sub freezing OAT will do that to you.
 
Last edited:
I have one bendix and one light speed as well. I can not manually adjust my light speed, but I do have a readout, and at similar flight conditions my advance is 32 degrees. I believe this is the maximum allowed by the lightspeed?
The one difference is that I run at best power settings for 65% or less power settings. I have not found the difference between total gallons used on a trip between LOP and best power settings, at altitude, to be worth the extra time.
Dan has suggested a 100 ROP, why not include one at best power to see if there is any difference? At these power settings you should be within the reccomended Lycoming power range for best power.
 
Intersting information Mike. Thanks.

The EFII system I installed is limited to 30? advance. I believe most EI systems will go to at least 35? which is probably more optimal LOP above 10,000. The CAFE testing years ago seemed to conclude up to 8000 mags were just about as efficient as EI, but up higher they fall off due to fixed timing. Your testing would indicate that may not be true.

There is a good reason timing is not adjustable in flight - it requires attention and if not done properly the consequences could be expensive. So one needs to be careful and know whats going on - like flying LOP.

I like advanced timing down low at low power. The Lycoming probably would be just fine at 37? - clean plugs, no unburnt fuel and LOP. I co not have that feature, 30? is max advance.
 
Timing Map

Toobuilder,


Here is the basic map that my ignition is operating on for your interest. Some of the data was not available at the higher MAP Lower RPM settings because I chose not to operate there. This caused the chart to be jagged on the lower end but you can assume that the curve would extend into these areas on a flat plane.

Timing%252520Map.png
 
Last edited:
...There is a good reason timing is not adjustable in flight - it requires attention and if not done properly the consequences could be expensive. So one needs to be careful and know whats going on - like flying LOP...

The ability to manipulate timing is a powerful tool - but with great power comes great responsibility. Your caution is certainly warranted.

To be fair, this system does have a "lock" feature that will keep the settings fixed once you get them where you want them.
 
That one is going to be tough though - flying along at 500 MSL at 99% power while messing with the ignition box is going to require some concentration. Probably best to head out to the coast... Dont need to tangle with power lines at 220 knots!

Dude! Surely you can recruit a GIB who will cheerfully fly the Rocket while you fool with the ignition box and power levers.

Yeah, yeah, we care about ya'. If you bust your butt we won't get data ;)
 
Dude! Surely you can recruit a GIB who will cheerfully fly the Rocket while you fool with the ignition box and power levers.

Yeah, yeah, we care about ya'. If you bust your butt we won't get data ;)

Gee, where can you find a qualified Test Pilot in the Mojave area to help you out....... :D
 
Lightspeeds can do this too

FWIW, you can adjust timing on a Lightspeed Plasma unit as well by hooking up a 10K pot to the box, although you can only shift the whole curve rather than making custom timing maps. It's documented in the manual which is available on their website. You can display the timing with either the display unit which Lightspeed sells, or directly on a Touch-equipped G3X. Search this forum for the connection diagram which includes the necessary op-amp if you want to display timing on your G3X.
 
Last edited:
Please be careful! You don?t have a knock sensor like modern automotive engines. The more you advance the ignition timing, the higher is the risk of destroying your expensive engine. If you experience a heavy knocking event, this takes 2 to 3 seconds. You will not be fast enough to cut the power.

The risk depends much on the cooling. Chances of a catastrophic engine failure rise with higher ambient temperature. What works in the winter may fail in the summer. For this reason the timing curves have to be conservative.
You need much more data to calibrate your ignition timing, than you can collect during flight. If you want to go to the limit, you need a test cell where you can change ambient temperature und pressure.
 
I really only intend to push the limits when high and LOP when detonation is essentially impossible. The TO power experiment will not surprise me if the best power setting is less advanced than the timing specified on the data plate. At any rate, the way I have the curve set now my max RPM advance is 26 degrees and all the MP advance is removed by 24 inches, so high power ops matches the data plate.

Your caution concerning OAT is noted. Detonation margins are greatly reduced by high induction air and cylinder head temps. However, this particular series of experiments is just informational. Any advance curves that I develop now will be invalid with the addition of the second electronic ignition. As it is now, only one plug is controling the start of the combustion event and as a result the spark timing has to be earlier to get optimal PCP.
 
Last edited:
I really only intend to push the limits when high and LOP when detonation is essentially impossible. The TO power experiment will not surprise me if the best power setting is less advanced than the timing specified on the data plate. At any rate, the way I have the curve set now my max RPM advance is 26 degrees and all the MP advance is removed by 24 inches, so high power ops matches the data plate.

Your caution concerning OAT is noted. Detonation margins are greatly reduced by high induction air and cylinder head temps. However, this particular series of experiments is just informational. Any advance curves that I develop now will be invalid with the addition of the second electronic ignition. As it is now, only one plug is controling the start of the combustion event and as a result the spark timing has to be earlier to get optimal PCP.

I had timing info operating the Subby engine and as I recall the Subaru ECU set timing at about 24° at take off power. I thought that interesting. At low power cruise in closed loop it would run at 36-37°. That is a critical item in establishing best MPG for the vehicle, same applies to an aircraft engine.

You may find the data plate number is pretty close to optimum. It could well be a compromise considering cooling requirements at that power level. Unburnt fuel going out the exhaust does cool the heads.
 
Last edited:
Unburnt fuel going out the exhaust does cool the heads.

Although there are some fuel molecules going out the exhaust at any mixture setting richer than stoichiometric, there's not enough fuel mass to physically cool anything.

Temperature of combustion is simply tied to mixture, with near stoich being peak. Combustion temperature is mirrored in head component temperatures.

Below from Taylor. I've converted F/Fc to the familiar air:fuel and F/A at the top:

 
...You may find the data plate number is pretty close to optimum...

Roger that, but I suspect that it is more of a compromise. There is a radically different requirement for ignition timing influenced by piston speed, manifold pressure and mixture. It's likely the data plate timing is spot on for one specific operating condition, but determining what condition that is will be the goal. Maybe it's takeoff power at sea level, maybe it's 24 squared... I don't know.

Back in my drag racing days it was said that tuning an engine on the dyno was a good start, but the best dyno available was the dragstrip. That's where the real world performance played out. I now have the opportunity to do the same thing with my particular airplane, my particular induction system, my particular exhaust tuning, etc. keep tweaking things until the performance is optimized and see what the curve looks like. If it ends up being a perfectly flat 26 degrees, so be it.
 
Good stuff

Michael,
Always enjoy reading your post and learning something. Thanks for taking the time to share what you learn!
Cj


Roger that, but I suspect that it is more of a compromise. There is a radically different requirement for ignition timing influenced by piston speed, manifold pressure and mixture. It's likely the data plate timing is spot on for one specific operating condition, but determining what condition that is will be the goal. Maybe it's takeoff power at sea level, maybe it's 24 squared... I don't know.

Back in my drag racing days it was said that tuning an engine on the dyno was a good start, but the best dyno available was the dragstrip. That's where the real world performance played out. I now have the opportunity to do the same thing with my particular airplane, my particular induction system, my particular exhaust tuning, etc. keep tweaking things until the performance is optimized and see what the curve looks like. If it ends up being a perfectly flat 26 degrees, so be it.
 
Michael,

I am very interested in your results. I just built a megasquirt EMS to replace one of my mags. I almost bought Ross' system, but had good luck with MS when I built one for my Porsche and it was half the cost. It also allowed me to build a custom trigger arrangement. I got a sacrificial mag from Gil and modified it with a Hall Sensor and flying magnets. I didn't like the idea of a trigger on the flywheel due to the alt belt and the damage it could do to the pickup.

In addition to the complete flexibility, it also has two advance tables that can be switched with a toggle switch on the panel. What I learned from my tuning experience is that optimum mixtures/best power (12.5:1-13:1) burn faster and require less timing than very rich or lean mixtures. I plan to have an adv table for Peak/ 50* ROP and one for LOP. Just not sure what the delta's will be. Just yesterday I was building my test advance curve. Your results and weasels are very helpful.

I hope to have this installed in a week or two and look forward to sharing our results.

Larry
 
Last edited:
Larry,

I too was/am wary of a critical element of the timing system living out near a belt that can fail at any time. However, the CPI system only mounts the hall pickup out there in harms way, and it is a very robust piece. Ross does not have a dedicated mount for the 540 yet, but I took his advice to make my mount "strong enough to lift the engine with", and I'm confident it will take more than a thrown belt to take this thing out. Additionally, it is protected by the front baffle and I have also added an aluminum conduit to "armor plate" the wiring. My approach goes beyond "belt and suspenders". That said, I am very much in favor of a direct crank pickup for timing accuracy. The way the accessory geartrain rattles around back there it is a wonder magnetos can deliver any kind of accuracy at all.

I have already suggested the "high/low" switch to jump back and forth between more advance for LOP and less for ROP and Ross tells me he's looking into it.

Also, for everyone reading this: these initial experiments are for info only. Any curve I get from this data is going to be largely false once I get the second unit up and running.
 
Last edited:
I think you pick your poison with regards to where you choose to mount sensors. Electronics mounted inside the engine may see 240F on a bad day. Their lifespan could be affected unless rated for those temperatures.

The chances of breaking a well maintained, quality belt on a Lycoming are probably about the same as some catastrophic failure like a separated jug, magneto drive gear failure or dropped valve. People too lazy to pull the prop and change a cracked or old belt once in a while might have seen a few break.

Mike has mitigated the effects of that remote possibility well IMO. The pickups are located inside the flywheel ID (the belt will flail outwards initially with the greatest energy), the mount is very strong and he's armored the cables.
 
Last edited:
The way the accessory geartrain rattles around back there it is a wonder magnetos can deliver any kind of accuracy at all.

Seen any hard data regarding how much they rattle around?

Allow me to play devil's advocate for a moment.

Crank triggering has a lot of merit on traditional automotive engines which drive a distributor from the far end of a very long torsional system. For example, an old-school GM V8 distributor was subject to the combined twist (and twist equivalent) of the crank, the timing chain, the camshaft, the distributor gears, and the distributor shaft. Torsionally speaking, that's a long, soft, flexible system.

The mag drive system on our Lycs is actually much stiffer. The crank gear drives the idler gear, which drives the mag. It's just one tooth mesh from being a direct connection...and those gears have specific backlash limits. The bottom line is that torsional oscillation is pretty much confined to crank twist.

The 540 and 580 sixes have longer cranks, so they are torsionally more flexible, but they also have pendulum absorbers to limit resonant oscillation. The shorter 4-cyl cranks don't twist as much, and those with higher HP ratings also have pendulums.

The sixes have rubber magneto drive blocks to lower the natural frequency of that part of the system. Those do detract from timing accuracy when deflected, but that would only be at one resonant RPM range. The fours (there may be an exception or two) don't need them. Not that it really matters; when we replace the mag with a low inertia EI trigger, we change the natural frequency anyway.

Now consider the mode shape of the crank oscillation. In our torsional system, the prop is a huge inertia. As such, it can be considered torsionally stationary, in effect an anchor against which the crank oscillates. The end nearest to the prop oscillates very little, while the far end oscillates somewhat more. The usual argument for triggering at the prop end is thus its torsional stability.

However, our real interest is making the spark arrive with accuracy as it relates to piston position. If we place the trigger at the prop end of the crank, the pistons at the far end are still subject to whatever crank twist may be present in the system. The spark merely arrives with "perfect" timing to find that the pistons at the accessory end are arriving early or late. In reality, perfect accord between spark and piston position is only obtained for the pistons physically closest to the trigger position. It doesn't matter which end we trigger from; the pistons at the other end of the crank will always be out of sync by whatever degree of crank twist is present.

So, in reality, I'd argue that the only difference between front and rear triggering is the limited lash of one idler gear.

One more detail. Flame speed is more or less finite. Double the RPM and spark accuracy must be twice as accurate. Triple it, and spark accuracy is three times more critical. We are running low RPM engines. Just how accurate does it need to be?

 
Last edited:
Seen any hard data regarding how much they rattle around?

Only when assembling a brand new engine and giving the driven gear a twist - then watching the effects of the tolerance stackup resolve through the geartrain.

I know Bill has some opinions about this too, concerning the drive system he built for development of the EI commander.


...One more detail. Flame speed is more or less finite. Double the RPM and spark accuracy must be twice as accurate. Triple it, and spark accuracy is three times more critical. We are running low RPM engines. Just how accurate does it need to be?

I agree - it's a nit. But it is still an improvement. Whether it is enough of an improvement to justify a distributed ignition system is up to the consumer. If I gave the impression that this factor was an important discriminator, that was not my intent. I personally like the direct measurement, that's all.
 
The gear lash is certainly there. The book spec is 0.004" to 0.015", with a service limit of 0.020". The crank and mag shaft gears are the smallest at about 1.6" D at the tooth contact, which means the idler would be 3.2"D.

So, treat the crank as fixed. A worst case 0.020" lash between crank gear and idler would allow the idler to rotate 360/[(3.2 x 3.14)/0.020], or 0.7 degrees. The same lash between the idler and the mag gear would be twice that, so 0.7 + 1.4 = 2.1 degrees. Woggly shaft bearing fits might make it a little worse.

I doubt the system actually oscillates through the full lash dimension in operation, unless it gets into a resonant range for one of the modes. Mostly it would just ride with the driving and driven teeth in normal contact. Even if it did rattle the full lash dimension, at the middle of the 0.004~0.015 spec it would be 1 degree total at the trigger shaft. That's not as good as a magnet on the ring gear, but it ain't awful for a 2700 RPM engine.

Just putting things in perspective...
 
I am not worried about slop from lash. I will see no more slop or inconsistency than I get with the timing on my current mags. Further, most of the slop is during changes in RPM. Under steady state, I doubt there is much movement and I am comfortable that my timing won't drift more than 2-3* from initial setting. Your results posted early show there is not much to be lost from being off a degree or two, as long as I stay away from aggressive timing and detonation risk. I am confident that any drift that I see will be no worse than that drift every lycoming user sees between annuals as the mag timing drifts due to point wear.

Larry
 
Larry,

I too was/am wary of a critical element of the timing system living out near a belt that can fail at any time. However, the CPI system only mounts the hall pickup out there in harms way, and it is a very robust piece. Ross does not have a dedicated mount for the 540 yet, but I took his advice to make my mount "strong enough to lift the engine with", and I'm confident it will take more than a thrown belt to take this thing out. Additionally, it is protected by the front baffle and I have also added an aluminum conduit to "armor plate" the wiring. My approach goes beyond "belt and suspenders". That said, I am very much in favor of a direct crank pickup for timing accuracy. The way the accessory geartrain rattles around back there it is a wonder magnetos can deliver any kind of accuracy at all.

I have already suggested the "high/low" switch to jump back and forth between more advance for LOP and less for ROP and Ross tells me he's looking into it.

Also, for everyone reading this: these initial experiments are for info only. Any curve I get from this data is going to be largely false once I get the second unit up and running.

I also liked the robustness of Ross's setup. I also would have put a conduit over the wiring, as it is the weak link. I suspect your set up will survive a belt failure.

I had a unique problem that kept me from the CPI system. I have the old Generator Flywheel and it doesn't have the little "shelf" where Ross' system has you mount the magnets. That installation provides a good bit of protection from the belt. I would have had to mount the magnets outside of the belt path (using the holes on the outer edge of the flywheel), leaving me highly exposed, as the hall sensor would have been 2" from the belt in a direct path and also far from the block (a long lever). Not a good combination. I didn't have the appetite for buying a new flywheel. I also would have had to fabricate a custom hall sensor bracket.

With all of that said, I am no safer than you are. My franken-ignition is built on a Slick mag setup which had been known to explode for no reason from time to time.

In the end, I would have swallowed the risk of the crank timing mechanism, as I had a redundant system (mag). I am not so sure I would do it with a dual ignition system. Just seems to risky. I do not agree with the other poster that a belt failure can be eliminated through prudent maintenance. I don't care who's name is on the label, it is a piece of rubber and could possibly been formed in crappy off-shore plant. I have had a LOT of problems with rubber based auto products in the last 10 years. Most be something about virgin rubber not making the trip to Asia well:) Or they are sneaking in lower cost compounds to reduce their cost.

Larry
 
Last edited:
...

I know Bill has some opinions about this too, concerning the drive system he built for development of the EI commander.
...
While developing the EICommander we realized early on that testing on a flying aircraft with an expensive engine wasn't really the best way to go. So, we cut the back half off of my old O-290 and tried spinning the crank shaft with an electric motor. It turned out the gear noise was so loud we couldn't stand next to the thing. Strike one.

We then milled up our own "accessory case" to spin two P-mag simultaneously, using the idler gears with the electric motor in place of the crankshaft.

In both tests we found there is about 1.5* of slop in the drive train, even with new gears.

That is why the EICommander doesn't report any timing divergence below 2.0 degrees.

For those who aren't familiar with the EICommander, it can monitor and control P-mags. One of its primary functions is to notify a pilot if the ignitions do not fire at the same time, along with telling them when they fire, and the condition of the coil packs and ignition harnesses.
 
Last edited:
I guess I have better luck than some folks here with belts. 10 years on the one in my RV before I changed it during the big overhaul and 17 years on the ones in our shop car, probably the original ones (1989), maybe 6000 hours.

As a point of interest, how many people here have seen a Lycoming belt less than 5 years old/1000 hours break? Curious as to how common this might be.
 
I guess I have better luck than some folks here with belts. 10 years on the one in my RV before I changed it during the big overhaul and 17 years on the ones in our shop car, probably the original ones (1989), maybe 6000 hours.

As a point of interest, how many people here have seen a Lycoming belt less than 5 years old/1000 hours break? Curious as to how common this might be.

I lost a belt on a 182 10 years ago. I rented from a fellow owner and did not know the maintenance history. It could have been neglected. Replacing that belt at a remote airport was the first time I had the cowl off.

Larry
 
I forgot to mention before, when you add the 2nd EI installation, be sure to dial back the timing a bit. Dual plug installations need less actual timing to reach the same effective timing as single plug installations (learned this on my porsche). With only one EI, the more advanced flame front will travel past center to meet up with the other, less advanced flame front; vaguely averaging the two timing settings. Once both are EI and advanced to the same point, the flame fronts will again meeting in the middle and your effective timing will be greater than the effective timing with one EI and one mag, assuming you left the advance tables the same. This is theoretical, as I do not have experience doing it, by my understanding of the dynamics tells me that you will see it.

I'm sure you'll see this in your testing.

Larry
 
Last edited:
Yes Larry, I'm anticipating pulling the timing back when I add the second system. I briefly alluded to this in post 25.

With the current setup (single ignition), the engine seems to like 36 LOP and about 33 ROP in cruise. I'll go out on a limb and predict 33 and 30 with dual ignition. Testing will tell all, and you will read about it here.
 
Thanks for all of detailed data here. I will model my two tables for testing based up on your experience here. I am curious what your complete advance table looks like. If it's easy to share.

The weather is going to be warm this weekend and I hope to get mine installed. We'll have low ceilings, so no testing unfortunately.

Larry
 
I'm running pretty conservatively here, as my "RPM advance" is all in (25 degrees) by 2250, and my MP advance is zero above 24 inches. The only thing I've really had a chance to play with is the total advance at my typical cruise settings. So my "advance table" is hardly optimized through testing (it may be spot on for all I know, but that would just be luck).

Just be careful using the data I'm presenting here. It's not intended to be a model.
 
Cable Protection

We took the concerns about cable vulnerability to heart here and added 10-32 threaded holes to the Lycoming CPI and EM-5 Hall sensor brackets. These will allow users to easily bolt a robust shield to the brackets to protect the cables from possible flying belts. I should mention that the crank triggered LightSpeed cables have a similar vulnerability.

360hall2_zpsydj6s0un.jpg


360 Bracket

540hall3_zps3tesqe6e.jpg


540 bracket

lyc540pix4_zpsxzjvltri.jpg


Dual element sensor installed on 540 case.

The bolt spacing and support web are different on the 360 and 540 cases, requiring different sensor mounts. Both mounts can take single or dual sensors (the red part).
 
Last edited:
Previously, regarding crank-triggered vs accessory case triggered systems, I wrote:

I doubt the system actually oscillates through the full lash dimension in operation, unless it gets into a resonant range for one of the modes. Mostly it would just ride with the driving and driven teeth in normal contact.

I've since been playing with an EI triggered at the accessory case mag hole. The straightforward way to confirm the actual timing delivered to the engine is to set up a stiff pointer at the case split, and use a timing light.

First, a caveat. Baby, it's cold outside, even in Alabama, and the 390 blows a mighty wind, even at part throttle. The combination doesn't lend itself to slow, deliberate measurement. Right now, about all I can stand is 30~60 seconds draped across the top of the engine while a buddy mans the throttle. However, it is appears that within one particular narrow RPM range (call it "low teens" until I can get a warmer day), the timing light shows some jitter, about 2 degrees, that is probably the result of torsional oscillation in a resonant range.

At all other observed RPM's, the timing is rock solid and razor sharp under the strobe. The only thing bouncing around is my bifocals in the prop blast.

Call it a preliminary observation, subject to a future check. I'm not concerned with it because empirical data (existing Lightspeed and Electroair mag hole triggers) says the oscillation probably doesn't generate much oscillating torque, and the rotor in this case is both low mass and robust. Two degrees of lash means zip in terms of engine operation at this RPM and MP. Without elaborate equipment, I can't tell you if the 2 degrees is simply the full range of freeplay in the gearset, or if some small crank oscillation is also involved. My best guess is it's all gear lash (one degree positive to one degree negative), as it seems to match what Bill observed with his EI Commander bench rig.

So, on a preliminary basis, the original statement appears to be true. The gearset tends to ride in stable tooth contact, except in a resonant range, the result being accurate timing everywhere but that resonant range.

It will warm up quickly here in the coming months. For comfort and safety, I'll bolt the timing light to the engine case, with the pointer, and set up a remote trigger for a future check. If nothing else, it will allow holding on with both hands while making an observation at full power.
 
Its been a while but I am finally running the dual CPI ignition AND found a nice day to test.

The earlier data I collected included CHT and EGT data among other things, and I think its safe to say their behavior is consistent whether you are running two EI or a single. So for todays test I ignored those and focused on power (read as TAS). Todays flight confirmed some suspicions and also brought a suprise or two.

The flights today are meant to establish a baseline for the ignition curve across a variety of operating conditions. These include:

100% power
Cruise at best power (8,500 MSL)
Cruise at Peak EGT (8,500 MSL)
Cruise at 50 LOP (8,500 MSL)

I'll cut to the chase and just provide the numbers for each condition rather than put up a chart like last time.

Best power cruise:

14.5 GPH, 23 MP, 2300RPM I explored 25 to 31 degrees advance and it was fairly flat turning 200KTAS at 25 degrees, peaking at 201 @ 27/28, and dropping to 198 @ 31

Cruise - Peak EGT

12.0 GPH, 23 MP, 2300RPM I explored 27 to 33 degrees advance and it was also fairly flat turning 197KTAS at 27 degrees, peaking at 199 @ 30, and dropping to 196 @ 33

Cruise - 50 LOP

10.5GPH, 23 MP, 2300RPM I explored 29 to 35 degrees advance and it showed the more expected positive response to advance turning 187KTAS at 29 degrees, peaking at 194 @ 33 and dropping to 190 @ 35

100% Power

This was the fun part. Because I'm close to Death Valley, I flew over to Furnace Creek airport (-210 MSL) and flew the flat, desolate valley floor at a safe 200+ AGL, but maintained 0 MSL with all the knobs full forward.

25GPH, 31.1 MP, 2700RPM I explored 20 to 28 degrees advance and the suprise of the day was the perfectly flat power delivery turning 211KTAS at every single ignition setting.
 
Last edited:
Very interesting data. The LOP and WOT low level speed runs, most of all. Have to sift all that in my mind for a bit.

Thanks for doing the tests and posting this info.
 
Well, I suspect the 100% power setting needs a dyno to confirm. Flying at 211 knots is probably right up against the drag wall and a swing of 5 HP probably would not even show up on the ASI. I had hoped to see just 1 knot somewhere in there to pick the sweet spot, but nope - flat.
 
Last edited:
Well, I'm late to this party but hearty congratulations and thanks to Michael and all the others who have and continue to, contribute to this important discussion. Fascinating data, I have to say.

Probably not what Ross wants to hear, but I've been waiting on a 6 cylinder P-Mag for so long that I don't see any harm in waiting longer. Am I at least right that as a minimum, its manner of mechanization would at least eliminate flywheel "wiring and installation" damage concerns from the equation?

Note: Dan...amazing knowledge. Never had a clue about all the oscillatory elements you discussed. Please keep it up---I for one, love this stuff.


Lee...
 
Well, I'm late to this party but hearty congratulations and thanks to Michael and all the others who have and continue to, contribute to this important discussion. Fascinating data, I have to say.

Probably not what Ross wants to hear, but I've been waiting on a 6 cylinder P-Mag for so long that I don't see any harm in waiting longer. Am I at least right that as a minimum, its manner of mechanization would at least eliminate flywheel "wiring and installation" damage concerns from the equation?

Note: Dan...amazing knowledge. Never had a clue about all the oscillatory elements you discussed. Please keep it up---I for one, love this stuff.


Lee...

Many folks have been waiting years for their 6 cylinder P-mags and many will continue to wait for them to be ready. That's everyone's choice and we don't expect to take over the whole market. In the meantime, we are shipping lots of CPIs for fours and sixes every week. Most folks buying like the lower price, faster, more flexible programming and lower weight.

With the added threaded holes in the Hall mounts for protecting the cables with armor, few people have any concern about belts damaging them.

You'll see another coil mounting option for the 6 cylinder CPI soon and we'll be working on Hall mounting options for the small flywheel engines and dualmag engines when time permits. Not sure when that will be. We're in a constant backlog situation at the moment with EFI and CPI orders.
 
Last edited:
Back
Top