What's new
Van's Air Force

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

Custom Timing Curve for Legacy EFII Ignition

crabandy

Well Known Member
I have a "Legacy" EFII Dual Ignition only (no injection) on my Parallel valve standard 8.5 compression O-360 with a fixed pitch Catto Prop. I've flown almost 300 hours behind the canned ignition curve with no problems per say, but I don't believe it is optimum for my situation.
I generally takeoff full power 2100 static RPM and climb at full power ROP and 130 KIAS. After level, accelerate and CHT's below 400 I pull power below 65% and adjust the mixture to LOP. I have to manipulate the throttle somewhat to achieve an EGT spread within 80*. I also burn 91 Mogas in 1 tank during cruise when available.

Here is my canned ignition curve super imposed in purple on Nigel Speedy's graph from his article in KitPlanes Magazine "The effects of ignition advance on cylinder heat temperature, speed and efficiency."

IMG_8002-L.png


I'm thinking my canned timing curve is a bit aggressive with the oversquare condition with my fixed pitch prop during climb resulting in higher engine temps. I would appreciate any input on my proposed timing curve anyone has to offer. I have spoke to and appreciate Robert of EFII who has consistently fielded/answered my numerous questions via phone and email for the past several years.

My current ignition curve is currently controlled by RPM and Manifold Pressure together.

The current RPM curve is shown in the shaded yellow column 30 degrees above 1250 RPM, my proposed timing curve keeps timing at 25 degrees BTDC above 1250 RPM.

IMG_8312-L.jpg


My current ignition curve uses a 4 degree retard at higher manifold pressures tapering down to 0 retard at 24.1, set timing as 30 degrees BTDC above 1250 RPM. With my proposed timing curve of 25 degrees BTDC I would start an ignition advance similar to Nigel Speedy's compromise timing curve shown in the first graph. 1 degree advance starts at 24.6 inches manifold pressure and levels off at 6 degrees of advance at 22.3 inches of manifold pressure for a total of 31 degrees of ignition advance. I'm initially limiting the advance close to what I've been flying the last 290ish hours.

IMG_8313-L.jpg


Robert suggested a value at the low end of the manifold pressure range in case of a manifold pressure sensor failure, I picked 0 degrees to keep timing at 25 degrees.

IMG_8314-L.jpg


Thoughts, suggestions and ideas appreciated!
 
Last edited:
With my proposed timing curve of 25 degrees BTDC I would start an ignition advance similar to Nigel Speedy's compromise timing curve shown in the first graph. 1 degree advance starts at 24.6 inches manifold pressure and levels off at 6 degrees of advance at 22.3 inches of manifold pressure for a total of 31 degrees of ignition advance. I'm initially limiting the advance close to what I've been flying the last 290ish hours.

Andy, consider manifold pressure to be a proxy for altitude, then think about where you typically cruise. In general, flying higher is more efficient, plus the benefits of a LOP advance are more pronounced. Sadly, far too many RV'ers don't routinely climb high (say 9500-12500) because they are fighting CHT all the way up.

I would suggest 23 degree base timing for your parallel valve, then start adding advance at 24.6"MP. You're carbed, so that would probably be 4500 feet or so, WOT and leaning in the climb to maintain best power mixture. Now stretch out the ramp so it doesn't arrive at 31 degrees until a much lower MP, say 20" or less. That should put you pretty close to a best power ROP timing all the way up into the 9.5-12.5K altitude (i.e. least time to climb), after which you go LOP.

23 degrees base is obviously conservative, but I note you have auto fuel on board, and run a fixed pitch prop. It would buy a little detonation margin while oversquare, in case you picked the wrong tank on a hot day, and it will lower CHT without significant power loss (in any). Later, after establishing a baseline, shift the entire map 2 degrees more advanced (25 base) and see if you can tell any difference. I'll bet you can't.
 
I think you are on the right track Andy, but in situ test is the quickest way to get what you want. Two things right off: If you dont have a way to adjust timing in flight, call Ross and see if you can get the dual board programmer. Second, see if your box can be reprogrammed to use the LOP function. This is a VERY handy feature for your application.

With that capability in place, go out and burn some gas looking for some key test points. As a minimum you want to find your "best" performance at:

Cruise altitude, best power mixture
Cruise altitude, LOP
Sea level power, Full rich
Idle

At ROP cruise it's pretty simple to sweep the advance 10 degrees around your current "known good" setting to find your peak TAS. Dont expect to find a very sharp peak when ROP, but you should be able to distinguish a "hump". LOP is the same technique, except the peak is going to be a lot more distinct.

With the full power setting, you are using a similar technique, but you re not looking for a peak, you are looking for the point where retarding the ignition causes the TAS to drop. So in this case, go fast and record the TAS as a baseline. Then start pulling timing OUT and watch the TAS. Once you get a drop in speed, record your ignition setting, add one more degree to get your former speed back, and this is now your new 100% power setting. My guess is it will be a lot less than even the data plate value your engine has from the factory. Using the minimum advance here assures you are not losing any power, but have as much detonation magin in the bank as you can.

Idle timing is just a nice to have. In my experience, the Lycoming likes a lot of advance. I accomplish this by relying pretty heavily on MP as a controling function. This differs from your scheme in that you are primarily RPM driven with the MP as only a minor correction. For example, I start at a baseline of 15 degrees until 1000 RPM, then ramp up to 25 degrees by 2000, where it remains flat. I've also been zeroing out the MP advance for the lowest few "slots" on the window to guard against sensor failure, but the first usable setting is a whopping 15 degrees advance. This advance reduces as the MP increases and is shaped to intercept my typical 100%, cruise, and idle/taxi power setting (in my case, my engine "likes" ~ 34 degrees). The advance continues coming out until it zeroes, and then goes negative at about 25 inches MP, ultimately retarding 6 degrees at 30 inches MP.

If you were to follow this scheme, you would tailor your MP curve to hit your engines "happy points" at idle, cruise and also at 100% power. your curve would look different from mine, but the methodology is the same.

I have a flight test worksheet made up which will help you gather and organize this data. If you want a copy, send me a PM with your email and I'll get it out to you.
 
Last edited:
Thanks Dan and Michael for the responses!

Michael, I woukd love a look at your test cards.

I’m guessing I have the older boards with the older software with only 250 RPM adjustability, I’ll eventually have to dissect my boxes and see for sure. For now I have 1 programmer and no way to easily swap ECU’s in Flight, I could buy another cable and swap the programmer to individual boxes but kinda combersome.

I thought about in Flight changing the timing on just one ignition kinda like a single EI and Mag setup but I think that would skew my results once both ignitions were using the same timing.

My engine idles very smoothly at 500 RPM even with the Catto on the current timing, once I dial in the upper MP curves I’ll have to spend some time dialing in the lower RPM/MP timing.

Dan, I Fly 9-13K as much as possible. Doesn’t take long to get there and I like the performance and other options the extra altitude allows, no level offs needed but I will always exceed 400 CHT after a hot quick turn.
Your suggested curve does make sense, pretty easy to change 1 base timing value on the ground from 23 to 25 and test fly. I’m thinking of using your suggested 23 degree base timing and this advance schedule, I started the advance at 25" MP and kept it easy roughly advancing 1 degree/1" MP. I leveled off the advance at 7 degrees at 19"MP-7"MP. I assume less than 7" MP the engine isn't running or a manifold pressure sensor failure. This also gives me similar timing to what I have run previously in the lower MP range.

IMG_8323-L.jpg


2 Test flights, first at 23 degrees base timing and the second at 25 degrees. Flights performed as close together as possible for similar atmospheric conditions.
-warm airplane up flying for 15 minutes
- land
-record OAT/altimeter setting/CHT/oil temp
-takeoff full throttle climb out 100 KIAS and 100 ROP to 10K
-Record time to climb/CHT/Oil Temp
-set 20” man pressure and LOP and record KTAS/CHT
 
Last edited:
You might be only a software upgrade away from "current" EM-5 functionality. The biggest bang for the buck is the LOP function - makes changes in a flight test scenario essentially risk free. Send me a PM with your email and I'll get the worksheet out to you.
 
Test flights

Hello Andy,
It might be easier if instead of recording altimeter you just set the altimeter to 29.92. I do all my test flights at pressure altitude and record the OAT. That allows me to easily convert all my data to density altitude. It may not matter if your able to make all comparisons on the same day, but the advantage is you can better compare data over the life of the aircraft.

In case you need it the simplest formula for Density altitude:

DA = PA + [120 x (OAT - ISA Temp)] and ISA Temp = 15 - [(PA / 1000)] x 2

DA - Density Altitude
PA - Pressure Altitude
OAT - Air Temp in degrees C
ISA Temp - standard atmospheric temp at altitude in degrees C

NOTE: These formulas do not account for relative humidity, but they will get you the same results as your trusty old E6B.
Marvin
 
Last edited:
Dan,

Great post. From the GAMI dyno I can suggest that the 23 degrees you speak of is equivalent to what a magneto produces at the spark plug. So, in saying that, with a magneto being a compromise, it could be another degree or so, especially on a fuel with less latency.

You may want to take that number down a bit. Pressure sensors are the way to go, but it is expensive to test.

As a guided guess you are on the right track, but could be a bit more conservative.

All the best.
 
Dan, I fly 9-13K as much as possible. Doesn?t take long to get there and I like the performance and other options the extra altitude allows, no level offs needed but I will always exceed 400 CHT after a hot quick turn.

Yep, that's why I suggest delaying the addition of advance until a lower MP, and then ramping it in more slowly. The current timing map (in purple, displayed in your first post) starts advancing too early, and hits 30 BTDC while you're still in a climbing flight phase where best power is desired.

Why would a vendor deliver the system with advance cranked up to 30 BTDC at 24"? I'll speculate, and say it's hot-rodder thinking (i.e. power before all else). Power claims sell the merchandise. In reality we should be more interested in the "all else" part.

Let's look at dyno numbers from a good high performance engine shop

Yep, 30 BTDC was the best timing for max power...at an RPM suitable for Reno, or Red Bull.

BTDC HP-RPM

20 217 @3050

25 228 @3082

30 232 @3082

33 223 @3004

At 2700 RPM? No power difference between 25 and 30 degree timing. Running 30 just added mechanical stress, and higher CHT.

BTDC HP

20 200

25 210

30 210

32 205

Here's the good part. Take a look at 25 BTDC vs 30 in the RPM range (underlined red) where a fella might operate while climbing with a fixed pitch prop. Below 2500 RPM, less timing makes more power.

2qxrwr6.jpg


BTW, I've not posted it here, but power at 20 BTDC timing wasn't a lot less than 25 in the 2000 to 2500 RPM bracket.

Anyway, that's the theory. Go fly and get us some data.
 
Thanks Marvin, I?ve done my previous test at pressure alt as well using 29.92. In my ?mind? seems easier for airspeed to pre-calculate Density Alt and and fly all test at the same DA, but PA it is.

Thanks for the graphs/explanation DanH.

Another question......Wondering if burning oil (3-5hrs per at) retards timing?
 
DA vs PA

Actually,
If your precalculating and saving your data in DA, that is a good way to do a comparison.
 
I fired up the programmer and found out that I do have the EM-5 with an older software version, with the current software I can have the LOP functionality.

IMG_8331-L.jpg


I watched some videos online about the controller and found after scrolling through the menus and verifying my current settings made me comfortable operating the controller. I left the low RPM timing set at:
500--10*
750--15*
1000--20*


I Changed Higher RPM 1250 and up to 23*. I input the advance schedule below.

IMG_8335-L.jpg


A couple flights later left me more confident in experimenting with the timing and a little head scratching. I flew the same manifold pressure advance schedule at both 23* and 25* RPM timing above 1250 RPM.
Test flights were take off, set 90 KIAS engage autopilot on IAS/HDG to be established prior to 2000 PA where I timed the climb to 7000 PA. LOP was at 7500 and 21" map. Not enough data by any means to average out my results but here's my initial observations. 25* ran better throughout the range with:
-better idle
-10 seconds less time to climb 1000-7000 PA
-Approximately 10* less CHT
-Smoother and further LOP
-(From memory) Stock EFII curve was similar CHT's and way better LOP

I can only adjust 1 ECU at a time with my current boxes, I decided play with the timing of 1 ignition while on the ground for the lower RPM's. My airplane has always had a dead spot from 700RPM to 1000RPM when slowly applying throttle. I notice the dead spot when adding power to taxi from idle and when power is all the way back in the pattern and I need to add a smidge of power in the pattern. I always assumed that I wasn't moving the throttle fast/far enough to properly engage the accelerator pump of my carb. I played with the timing from 10-40* and found 750 RPM and up that 25* timing with the 7* manifold pressure advance (32* total) seems to give the smoothest running idle. The dead spot between 700-1000RPM seems to be gone. The timing advance increased my idle 200ish RPM and things were too hot to adjust it back to 500 and play with timing there.

TooBuilder has some good advice about playing with the timing "in situ." I plan on playing with timing on 1 ignition and recording values and tweaking as I go. Good news is I get to do a lot more flying, about that time and money......

If I find the best timing value for advanced timing with 1 ignition is there a ballpark change in timing that will occur when both ignitions are set to the same advance? For Example if my engine performs best at 25* on 1 ignition adding the second ignition should roughly change the value by how much?
 
Last edited:
The ballpark difference I experienced was 3 degrees between a magneto and the single CPI at "optimum". I'd expect you will see less than that
 
A couple flights later left me more confident in experimenting with the timing and a little head scratching. I flew the same manifold pressure advance schedule at both 23* and 25* RPM timing above 1250 RPM.
Test flights were take off, set 90 KIAS engage autopilot on IAS/HDG to be established prior to 2000 PA where I timed the climb to 7000 PA. LOP was at 7500 and 21" map. Not enough data by any means to average out my results but here's my initial observations. 25* ran better throughout the range with:
-better idle
-10 seconds less time to climb 1000-7000 PA
-Approximately 10* less CHT
-Smoother and further LOP
-(From memory) Stock EFII curve was similar CHT's and way better LOP

Interesting.

Climb is with mixture parked on full rich, or leaning to maintain best power through the climb? What was climb RPM?

Is 21" your WOT manifold pressure at 7500?

I would expect the stock EFII schedule to be a little better LOP at 21" as compared to the 23+5 schedule, because timing is 28 vs 30. It's hard to imagine how the EFII schedule can be way better than the 25+5 schedule, as they're the same timing, 30 BTDC.
 
In my opinion, ignore LOP performance until you have the capability to manually select that condition. Tuning an overall curve for LOP will always mean you are too advanced for best power mixture. The opposite is true of course. Best power timing will be too retarded for LOP. It's the fundamental flaw in the "one size fits all" approach to timing.

You can certainly play with the LOP advance as a knowledge exercise, but call Ross and get a reflash before you try to execute a new curve
 
Interesting.

Climb is with mixture parked on full rich, or leaning to maintain best power through the climb? What was climb RPM?
90 KIAS, leanest cylinder was leaned to 100ish ROP and climb RPM 2100ish. I never fly this configuration, just wanted to start somewhere. I need to do multiple climbouts at the different ignition advances, usually 130 KIAS.

Is 21" your WOT manifold pressure at 7500?
Throttle was pulled back several inches, to even out EGT’s for LOP. My mixture distribution at full throttle is horrible, 150-200 degrees between richest and leanest cylinder.

I would expect the stock EFII schedule to be a little better LOP at 21" as compared to the 23+5 schedule, because timing is 28 vs 30. It's hard to imagine how the EFII schedule can be way better than the 25+5 schedule, as they're the same timing, 30 BTDC.
The advance schedule I plugged in has 4* of advance at 21” MAP, 23* base gives 27* total advance and 25* base is 29* total. 20” MAP would’ve given me 30*. Splitting hairs for sure, but 7500 and 21-22” was a good setup LOP prior. As I think about it more, I left the mixture LOP for the descent and it was smooth all the way down. Below 18” MAP gave me 30* with a 23* base timng and 32* at 25* base timing.
 
Last edited:
In my opinion, ignore LOP performance until you have the capability to manually select that condition. Tuning an overall curve for LOP will always mean you are too advanced for best power mixture. The opposite is true of course. Best power timing will be too retarded for LOP. It's the fundamental flaw in the "one size fits all" approach to timing.

You can certainly play with the LOP advance as a knowledge exercise, but call Ross and get a reflash before you try to execute a new curve

I agree the LOP switch is the best way to get the most out of ROP and LOP and the software flash is in my future. Meanwhile I?ve got a X/C trip this weekend which should give me lots of opportunity to play with timing in cruise at various power settings and altitudes.
 
Sure. Cross country time is a good opportunity to play. Lots of time to let it really stabilize after an adjustment. Even with one ignition with the adjustment you should be able to get very close to your final number.
 
So my 20 hours of flight time playing with timing using 23-25 BTDC for normal flight ops gave me consistenly lower CHT's on climbout, but inconclusive results overall with advance at lower manifold pressure settings. More on that later...

I removed my ECU's and sent them to Ross at SDS for a new software flash for the LOP switch functionality. While waiting for my ECU's I added 1 wire to each ECU connector joined at a switch to provide 12 volts for the LOP function.
I put a lot of thought into making the LOP switch fit my normal checklist/cockpit flow, I put the switch next to the fuel pump/throttle where it seemed to flow the best.

IMG_8668-L.jpg


I'm lucky my local awards/trophy shop is easy to deal with, they were able to make a panel sticker for my new switch to match the other stickers.

IMG_8654-L.jpg


If starting over I would probably place it right next to the mixture, but it does stand out a little bit and is next to the throttle/mixture. I had thought about putting the LOP switch by the ignition switches, but I tend to ignore that side of the switch quadrant as the master/ebus/alts/ignitions are always on, I think this spot ties it more to throttle/mixture settings as it should be.

IMG_8740-XL.jpg


I had never verified the timing pickup magnets in the flywheel, Robert from EFII had drilled and installed them on my original installation. Referencing the SDS instructions and Lycoming service instruction 1437, I needed to make a 10 degree mark on the back of the ring gear and make a pointer denoting the case split.
I used a cloth tape measure to measure halfway between TDC and 20 on the ring gear and made a mark. The Lycoming tool looks like it just squares off the ring gear so I did the same with my carpenters square and measured back to the case split and drew a line on my plenum and made a pointer from an old hacksaw blade.

IMG_8764-XL.png


IMG_8747-L.jpg


IMG_8744-L.jpg


I set up the timing light, chocked the airplane and it was almost easier than checking the timing on a car engine. Not sure why I never verified prior, but my position/ timing was correct. I verified by changing the magnet position programmed on the ECU's.

IMG_8745-L.jpg


I input my RPM timing and manifold pressure retard/advance schedule and did a series of runups. I only have 1 programmer so I can only view/interact with 1 ignition on the fly. I tend to leave the programmer on my 88 ECU as it is wired to the battery bus and I don't have to turn anything else on to play with programming. Everything seemed well with the ignitions and a quick glance at the programmer showed my expected timing of 23 degrees (25 BTDC via RPM and 2 degrees of retard at 29 inches manifold pressure) upon full power for the takeoff roll. When I leveled off at 3000 AGL and pulled power I didn't see the expected advance as manifold pressure was reduced. My LOP switch did not apply the programmed 7 degrees of advance either.

Back on the ground I found my 71 ECU was performing as programmed, It was late Friday evening and I put it away sent a couple of emails and pondered the problem. A couple of days, a discussion with Ross and a little more trouble shooting found a bad manifold pressure sensor. My system has a separate manifold pressure sensor for each ECU, easy to swap sensors and verify the bad sensor. My manifold pressure sensor failed "high" as in it was always reading 30.1. I picked up a replacement at NAPA, it was a stock GM/ Delphi sensor.

Makes me wonder how long the sensor has been bad, I should have caught it as soon as I started using the programmer. I am a slow learner and learn by doing (making mistakes).... But I spent most my time in the RPM/manifold pressure setting windows and only more recently using the gauge pages of the programmer that show ignition timing/manifold pressure/RPM. Could've been dead from installation, with the proper timing curve its very hard to tell anything is amiss when the engine is happily purring along.

Why didn't my LOP switch work?
There is a hidden setup menu in the SDS ECU's, one of these settings voids the LOP switch above 25.1 inches manifold pressure. Nice little safety feature for sensor or Pilot failure.

I'm looking forward to more flying and hopefully more definitive results with my timing.
 
Last edited:
Good write up and photos Andy.

For those not so familiar with the dual ECU, 4 cylinder setup, each ECU runs one set of spark plugs at all times but only one ECU is running the injectors at a time so if the backup ECU MAP sensor failed and is pulled high by the ECU (to allow the engine to make near full power), you would not notice any big running changes if you were running on the primary ECU.

Having the dual access programmer is a big plus with dual ECU as you can easily toggle between ECUs with the flip of a switch instead of unplugging cables and rebooting. Gauge 1 mode allows you to see MAP, RPM and the temp sensor outputs plus any sensor error codes for quick diagnostics.

It's important to verify initial timing with a timing gun as Andy shows here. Don't rely on base settings as many people have found these 2-4 degrees advanced over reality which can be dangerous, especially when operating on mogas. You can change the timing offsets by using the Magnet Position Window with your timing gun hooked up. This is a one time thing, once done, your total timing (rpm timing +/- MAP advance/ retard) will always be displayed correctly in the programmer.
 
Back
Top