What's new
Van's Air Force

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

Idle Quality versus Ignition Timing

nigelspeedy

Well Known Member
As part of my recent ignition timing experiments I looked at how ignition advance affects idle quality. I had some advice that reducing advance at idle would make it smoother. Well lets see.

For this test I started and taxied to the run up area and did my normal run up. For this I have the mixture full rich and prop full forward and then advance the throttle until I have 1900 RPM, then I lean the mixture until the RPM rises about 50 RPM, so 1950 RPM total, this is where I do the mag check and cycle the prop. Once I had this mixture position set I did not make any change to it for the whole test.

I used the advance shift function of the EICommander to adjust the timing. The first test point was with zero shift and this resulted in 23.8 deg advance at around 670 RPM with the throttle fully back with the engine warmed up.

I then let the engine idle for 1 minute and recorded the RPM at 16 Hz. I followed this with a run up using the throttle until I had 1950 RPM and did a mag check on each timing setting. I repeated this exercise with -1.4, -2.8, -4.2 deg shift which resulted in 21.0 deg, 19.6 deg & 18.2 deg advance respectively.

The results are shown in the graphs below. As I retarded timing from the baseline of 23.8 deg the idle quality decreased as measured by the increase in spread between the min and max RPM over 1 minute and the increased standard deviation. The RPM drop during a Mag Cx also increased as the advance was decreased.

Anyone have data to support a 'less ignition advance at idle = better idle quality' argument?

23.8%20deg_zpssb0w0nko.png
[/URL][/IMG]

21.0%20deg_zpsar3kxt6o.png
[/URL][/IMG]

19.6%20deg_zpsfc7msb08.png
[/URL][/IMG]

18.2%20deg_zpsmspgbaas.png
[/URL][/IMG]

Cheers
Nige
 
Yes

I have an Electroair on one side and mag on the other. There is a huge difference between idling on the mag and on the EI. I have varied the timing of them both from 20 to 25 degrees. The mag idles much better with advanced timing, as does the EI. of co urge the EI is running about 38-38 DBTDC at idle.

Vic
 
Nigel, I'm not sure what is metering your fuel but unless you take AFR into account, I'm not sure you can draw too many valid conclusions here about ignition timing vs. idle quality. All variables must be controlled to get valid data.

As rpm sags, MAP changes, which affects fuel vaporization and any change in AFR below 13 AFR will have a pronounced effect on rpm. Bendix type injection is not known for precise idle metering and carbs can have very poor cylinder to cylinder distribution. One cylinder at 14-15AFR and the rest at 12ish won't make for a consistent idle.

Most engines work well with 20-25 degrees of timing at 1000 rpm. Often the throttle can be closed slightly more to obtain the same idle rpm, resulting in lower MAP, improved fuel vaporization and more even distribution between cylinders which makes for a smoother idle.
 
I'm with Ron on this one.
If you want a lower idle speed, you need to set the idle stop lower. then re-adjust your idle mixture, and also chase the idle quality with the mixture control.
25 degrees fixed timing on 2 mags makes for a choppy idle at 470 RPM with a 2 blade wood prop. Plus, it might die during landing, which would be a bummer.
I haver a Pmag w/ one Mag on both my T-18 and my RV-8. They both idle better at low RPM now, and even lower if I shut off the mag. Both Pmags are set with the jumper 'in' for the lesser advance.
I have not done any special idle testing like you have, and my hat is off to you for doing it, but I don't think you are interpreting the data correctly, and until you test idle timing down to 8 or 10 degrees BTDC @ 450 RPM like a 1950's car, I don't think you have any usable data.
 
Precision Airmotive Silverhawk fuel injection

G'day Ross,

I am not chasing the lowest idle speed record, more after something smooth that won't falter on acceleration.

I agree that in any experiment it is easier to correlate cause and effect if you only change one variable. In this case the throttle, prop and mixture lever were all in the same position for each test. Only the timing was deliberately varied. I agree that as a result of changing the timing the average RPM changed a little, perhaps 30 RPM and this could lead to changes in AFR. But I think if I had adjusted the throttle after the timing change to keep the RPM constant I would have folks arguing that moving the throttle violated the test results.

Do you have any ideas on how to structure a test with mechanical fuel injection to optimize idle quality, and what metric would you use to rate quality? I think I have up to -12 degrees of timing shift available so I could easily repeat the test and vary timing from ~24 down to ~12 deg BTDC.

Cheers

Nigel
 
G'day Ross,

I am not chasing the lowest idle speed record, more after something smooth that won't falter on acceleration.

I agree that in any experiment it is easier to correlate cause and effect if you only change one variable. In this case the throttle, prop and mixture lever were all in the same position for each test. Only the timing was deliberately varied. I agree that as a result of changing the timing the average RPM changed a little, perhaps 30 RPM and this could lead to changes in AFR. But I think if I had adjusted the throttle after the timing change to keep the RPM constant I would have folks arguing that moving the throttle violated the test results.

Do you have any ideas on how to structure a test with mechanical fuel injection to optimize idle quality, and what metric would you use to rate quality? I think I have up to -12 degrees of timing shift available so I could easily repeat the test and vary timing from ~24 down to ~12 deg BTDC.

Cheers

Nigel

I realize you were going for smoothest idle, not lowest idle. My point was you don't know what the AFR was doing as the rpms changed and mechanical injection doesn't generally meter fuel very precisely at low idle especially. Your data generally says that with pretty large variations in rpm.

I'd rate quality as a function of lowest mean and peak variations in idle rpm but I don't think mechanical FI does a consistent enough job in AFR control to get valid results. Only way to tell if that parameter was well controlled would be a wideband on each primary exhaust tube and I don't think you want to do that to your airplane.
 
Last edited:
...I have not done any special idle testing like you have, and my hat is off to you for doing it, but I don't think you are interpreting the data correctly, and until you test idle timing down to 8 or 10 degrees BTDC @ 450 RPM like a 1950's car, I don't think you have any usable data....

Interesting you bring up the ?old car? example, as this is pretty much how I have the curve set on my airplane. Up to 1300 RPM I have a flat 15 degrees of advance. To that is added up to 15 degrees of MP advance. The RPM (?mechanical? advance, in 1960?s automotive terms) ramps up with RPM and is ?all in? with 25 degrees by 1900. The MP (or ?vacuum? advance in automotive terms) is ?all in? at low power and starts ramping DOWN at about 14 inches; zeroing out at 24 inches, then going slightly retarded as MP increases beyond that.

What this means for my idle timing is that I have my base RPM advance of 15 degrees during idle and taxi, to which is added ANOTHER 15 degrees due to my low MP. As a result, I?m usually running on the ground at about 30 degrees and it is happy. I have not performed a dedicated test to see where the engine will idle fastest, but I have toggled the advance while on the ground and observed a noticeable idle speed change. It would be very simple to start at a low advance and simply toggle the ?up? arrow key on the CPI box to add advance until the engine peaks and then starts to labor again. I'm not sure if the highest idle compares to "best" however.

One thing to keep in mind is that my decision to emulate the mechanical and vacuum advance of an automotive ignition was based upon my familiarity with 60?s cars. It seemed a ?safe? way for me to learn this system and in no way am I suggesting this is the ?best? way to do it. There are many options for setting up the advance on the CPI and I may move another direction in the future. But what I have works for me for now.
 
30 Deg at Idle?

Hi Mike,

30 deg at idle really opens up what I thought was best advance at low rpm. I have been told that as little as 9 deg is best. So back to the experiment, I think I will try wider range of advance settings and after each change adjust the throttle to maintain a constant RPM, say 750, to simulate I had adjusted the idle speed screw. If I keep the mixture lever constant and the RPM constant this might be a better test. I think I should be able to get from around 12 deg to 36 deg using the advance shift feature.

Cheers

Nige
 
Even with old cars, the idle quality improves significantly with the addition of the vacuum advance. The base timing is set (with the timing light) at 8-12 degrees (depending on engine) with the vacuum hose disconnected, but that is just the baseline. As soon as you plug in the vacuum hose back in the advance shifts to 25+ degrees. This advance is accompanied by a significant increase in idle RPM and engine smoothness.

It stands to reason that the advance setting that provides the highest idle is the one you want. I'm not sure I understand the need to move the throttle, do mag checks, etc. Just advance the ignition to the highest idle speed and then back the throttle stop down to the desired RPM.

All that said, the transition from this low MP/high advance setting to takeoff power needs to be smooth, and that might be tough with the limited adjustability of the other ignitions out there.

I've never really looked at "idle quality" as a discrete data point before... I guess I need to play with my ignition settings more and see what the result is.

Nigel, drop by sometime and we can compare notes.
 
Kevin Eldredge was over at my hangar 2 weeks ago and he said his timing is at 8 to 9 deg advance at idle and the engine hangs out at 400rpms happy as a clam. We were talking about all the dyno testing he has done to determine his best set up. I will also be doing some ignition experimenting as my idle is not happy. I am looking at CPI a little closer now that I know the EFII only allows you to customize at 250 RPM increments and the CPI allows for 100 rpm.

Unfortunately Speedy and I received some bad feedback on the EFII system last weekend and now it has me thinking what else is out there.
 
Axel PMag Cable

Axel,

The PMag interface cable is on the way, posted it today. You will need to download the program from the PMag website.

Cheers

Nige
 
what EFII feedback?

AX-O,

What feedback was bad with EFII? I'm running the dual EFII ignition and it runs great on my end. Curious what you are hearing if there's something to be watching for.

Thanks,
 
EFII

All our systems are set up with 100rpm programming now.

If you will send me a PM, I would like to know what the unfavorable feedback was,

Robert Paisley
 
AX-O,

What feedback was bad with EFII? I'm running the dual EFII ignition and it runs great on my end. Curious what you are hearing if there's something to be watching for.

Thanks,

VAF is a powerful tool. People say the wrong thing and folks loose business. I am researching ignition systems. This maybe an isolated case so I will continue to look for data.
 
EFII

Don't forget, our ignitions are all upgradeable to include electronic injection.
The reason our systems idle so smoothly and down to rpms that are really too low for engine oiling health is that the electronic injection always provides balanced fuel delivery. This is true from engine cranking to idling to cruise and max power. This is one of the big advantages of electronic injection.
The second big advantage is the tunability. Programmable fuel delivery is a pretty cool thing to have and a big leap forward from the alternatives.
No one ever goes back to the old systems once they update.

Robert
 
Anyone have data to support a 'less ignition advance at idle = better idle quality' argument?

I don't think this is true unless you have too much advance and I wouldn't want my idle timing much lower than 25 In my experience mixture adjustments have a much greater impact on idle quality than timing, assuming your timing is in a reasonable range. To a point, increasing idle advance will increase RPM, but not necessarily idle quality. In the EFII that I put on my Porsche air cooled engine, I used idle advance to boost RPM when the air conditioner is engaged. This is any easy way too boost idle RPM electronically (i.e. without moving the throttle plate). But that doesn't necessarily mean it is the best or most optimum way to reach an idle speed/setting, as Ross pointed out.

To Ross' point, optimum idle is a combination of mixture, airflow and timing.

Larry
 
I have a new data point for discussion.

Flew the Rocket today and just prior to shutdown I performed a little impromptu test. I set the idle to 850 and leaned per my normal practice. In this condition the CPI was showing 30 degrees advance. I activated the LOP switch (which I have set to add another 3 degrees) and the engine immediately picked up 10-20 RPM. Switching off the switch brought the RPM back to 850. OK, my engine seems to like 33 degrees better than 30... Lets see how much it will take.

So I activated the LOP switch again, saw the expected 20 RPM rise, and then went into the LOP menu and started adding advance. With every couple of degrees the engine kept increasing RPM until it flatened out at 900 and 42 degrees advance.

Then, as a final test, I shut the LOP switch off and the engine immediately dropped back to 850 RPM. Interesting though was the perceived change in smoothness. The normal 850 idle felt plenty smooth before, but it was noticably rougher than the 42 degrees/900 RPM setting.

So I'll not argue that there are some engines that do fine at 8 degrees, but mine seems to "like" 40 the best. Pretty easy to verify when all you have to do is flip a toggle switch and experience the change for yourself.
 
Last edited:
So I'll not argue that there are some engines that do fine at 8 degrees, but mine seems to "like" 40 the best. Pretty easy to verify when all you have to do is flip a toggle switch and experience the change for yourself.

Most things are a matter of considered compromise. A respected friend has long held that yes, they idle nice with advanced timing, but tend to get hot given a long taxi and/or a long wait on the runup pad.

No personal experience here. Probably worth a "confirm or deny" observation. At OSH, the Thursday morning conga line for departure can get really long ;)
 
Well, fortunately it doesn't get very hot here in Mojave. Its been almost a full year since I've seen the OAT break 120.
 
I'll have to explore the idle CHT/advance relationship more, but it seems to me that ignition advance is not going to have much effect on engine power at idle so the thermal loads should be pretty light in any condition. If anything, the smaller throttle opening to produce the same RPM with advanced timing should reduce the thermal load. We shall see how my engine behaves this summer.
 
Most things are a matter of considered compromise. A respected friend has long held that yes, they idle nice with advanced timing, but tend to get hot given a long taxi and/or a long wait on the runup pad.

No personal experience here. Probably worth a "confirm or deny" observation. At OSH, the Thursday morning conga line for departure can get really long ;)

Just a data point. RV-10, IO-540 parellel valve, Bendix injection, 9.5:1 compression, 1 mag set to 23 deg, Lightspeed system with visual and datalogging interface. Runs about 37-40 deg advance in the normal idle range ~900 RPM. CHT has never been a problem even when taxiing over 20 min on a day with ambient temp over 90 deg F with the air conditioner on.
 
I have a new data point for discussion.

Flew the Rocket today and just prior to shutdown I performed a little impromptu test. I set the idle to 850 and leaned per my normal practice. In this condition the CPI was showing 30 degrees advance. I activated the LOP switch (which I have set to add another 3 degrees) and the engine immediately picked up 10-20 RPM. Switching off the switch brought the RPM back to 850. OK, my engine seems to like 33 degrees better than 30... Lets see how much it will take.

So I activated the LOP switch again, saw the expected 20 RPM rise, and then went into the LOP menu and started adding advance. With every couple of degrees the engine kept increasing RPM until it flatened out at 900 and 42 degrees advance.

Then, as a final test, I shut the LOP switch off and the engine immediately dropped back to 850 RPM. Interesting though was the perceived change in smoothness. The normal 850 idle felt plenty smooth before, but it was noticably rougher than the 42 degrees/900 RPM setting.

So I'll not argue that there are some engines that do fine at 8 degrees, but mine seems to "like" 40 the best. Pretty easy to verify when all you have to do is flip a toggle switch and experience the change for yourself.

FWIW the Subaru ECI advanced timing to 35-40 after start at idle. It was very smooth.

My EFII single ignition retards to 15 at idle and is smooth also, but the mag is still set at 25 so don't know what any of this means.

I do believe the one short fall of EI on a Lycoming is lack of inputs vrs auto systems. The Subby H6 has 3 O2 sensors, air flow and anti knock; all going into an ECU that was priority locked with some features faked out to make it work in an airplane.

I just removed the O2 sensor on the RV-8 doing the condition inspection. It proved to be unreliable, not sure it was installation or 100LL. The system is not required to make EFII work, just information. I was interested in the AFR at various power-mixute settings but it was inconsistent. One take off it would show 12:1 next time 15:1 with full rich set. I did see 15:1 most of the time running LOP at altitude, but I was most interested in take off AFR and it was all over the ball park there. I do not believe AFP AFR changes that much with mixture set at full rich.
 
...I do believe the one short fall of EI on a Lycoming is lack of inputs vrs auto systems. The Subby H6 has 3 O2 sensors, air flow and anti knock; all going into an ECU that was priority locked with some features faked out to make it work in an airplane...

Not an expert by any means, but I have been around the block a few times, so let me throw this out for general consideration:

I think that full authority spark control (anti-knock) as seen on current automotive applications is virtually impossible to implement on any air cooled engine ? particularly one as mechanically noisy as a big bore aircraft engine. That said, I?m not so sure it would be of much (if any) practical value either. We do not have rapid throttle transients in an aviation application and we are not looking for the bleeding edge of emissions compliance. We are really only looking for a few stable conditions: i.e. Takeoff, cruise, ROP/LOP, and now ?idle?. All of these are pretty easy to establish through testing, so I?m not sure having the computer sniffing around for the edge of detonation with a knock sensor is going to buy you anything.

And further, WRT this thread, I sure can?t see any possibility of detonation at idle RPM. So as long as the temps aren?t going through the roof (which they don?t, in my experience), then I?d think that the advance setting that provides the highest RPM and ?smoothest? feel at idle is the goal. The only question beyond that point is whether the ignition system has the adjustability/flexibility to come off that high advance and transition to an appropriate setting for taxi and takeoff power. CPI definitely offers that capability, I can?t speak to the others.
 
Back
Top