Scanguage IGN guage, what is good? 40 either is really bad or really good
 

I am messing around with the fuel in the buick and would like to determine what the IGN guage is actually showing. It is supposed to be timing but I doubt 40 advanced or retarded is possible (thats where I seem to be at highway) premium it goes to about 38 which is really terrible or really good.

When I start the car on a cold day and shift into gear it moves to -3 while the engine is running slow.

So what is a good IGN value? As low as possible? As high as possible?
Is it REALLY showing an Advance of 40 and is that a good thing.

Just not sure how to guage what is ideal for that number.

Cheers
Ryan

...gasoline combustion takes time, and more time means "advanced" timing.

...the more advance, typically, the better the burning rate (not detonating or pinging).

...timing is commonly "advanced" for both acceleration and higher engine RPMs.

The higher the ING number without pinging the better.

I must be missing something, that seems somewhat misleading and wrong to me. The basic reading is degrees before top dead center.

For simplicity sake, lets just say 'when the piston is at the top of its stroke'. Typically, the piston is attached to others, so it will go down even if we don't ignite the fuel at all (a misfile). So what we want to do is ignite the fuel so that the resulting heat and expanding gasses will put the most downward force on the piston, not heat the cyl or go out the exhaust as heat.

The reaction takes time, so we generally ignite the fuel early. But the speed of the flame front is not fixed, it can vary with different factors, especially the air fuel ratio. At the risk of oversimplifying, lean mixes burn faster, rich mixtures burn slower. So, under significant load, a well tuned engine will generally be around 16 degrees before TDC. This isn't a hard number, since it is the 'effective' timing that matters. Different spark systems, even cyl shapes, will move this. When the timing gets yanked significantly from this, it is usually in response to something like knock detection. Basically, sacrificing 'work' for lower peak pressures.

When there are very light loads, a lot of modern engines will use very small amounts of fuel and have relatively slow piston travel rates. So, to get the most oomph, timing is often very close to TDC.

I'm sure you know all this (I'm reiterating it more for the lurkers), and you are probably thinking of some of the other factors, like exhaust pressure, that I haven't even mentioned. I'm just struggling with the 'higher number without knocking is good' as a blanket statement. Often, the higher number is the ECU reacting to knocking, and this isn't generally considered a benefit in combustion efficiency.

Regards,
-jjf

The ScanGauge IGN is not ignition timing BTDC, AFAIK. What tasdrouille said is correct. If I can keep IGN up in the high 30s and low 40s as much as possible, I can beat cruise control on a test route.

It depends on the car. On VW TDIs the reported value is the actual start of injection value in degrees BTDC, negative values are ATDC. But on all gassers I've tested it, the higher the value, the more timing advance there was. Mid 40s is about as high as it'll get in the Elantra.

Now this discussion has me curious. The timing specs for my 1NZ-FE engine in my xB are 8-12º BTDC at idle. I just started it up, and the Scangauge IGN value stabilizes right in that range. So I now agree IGN does reflect timing BTDC.

So, I recall from the old days that under load, you had to retard your timing, or go to high octane fuel, to avoid detonation and engine damage. It stands to reason that under light load, coasting on the level or slightly downhill, timing could advance significantly, and give us our best mileage.

The 1NZ-FE has variable valve timing allowing a range of 60 degrees relative to crankshaft angle. Is that what allows the ignition timing advance up to 40+º BTDC?

Now you see my confusion above. But I've been thinking about it. If you are an aggressive driver, high numbers are probably bad, because they suggest that timing has been retarded because of detonation.

But, if we start with the premise that you are trying to drive to maximize fuel economy, there is a factor I missed. The comment about the test track above made it click. A lot of times, you aren't even in 'light load', but overrun, essentially coasting with the injectors closed and O2 in the exhaust.

In another thread I posted some logs to show the trouble with OBD-II MPG calculations. If you look at the very first log, before I turned a bunch of traces off, you can see that timing advanced significantly when all the other indicators are for overrun. So, in that sense, 'high' is good. But, at the two ends of the log, when I was going up hills, higher would not have been good, but bad, so I'm still not sure if 'higher is better' always works.

-jjf

On the buick My FE was stuck around 27mpg on the highway instantaneous (terrible) I noticed my timeing wouldn't move past 25, threw in neutral, shut off motor, turned back on FE jumped after a few to around 37mpg instantaneous and timing was around 37 to think I wasted 15 miles driving at low fe when all it took was a restart.

Forgive the engineer in me, but it's not 100% clear what actually was going on under the hood. The first question would be 'where is the FE reading coming from'?

If it is from OBD-II, it could be a chicken egg problem. From MAP/RPM/IAT or MAF you can calculate fuel used, but only if you know actual AFR. Modern cars run at stoich a lot, but not always. It is quite possible that a particular MPG gauge uses other clues, like sparkadv, to try to guess when other cases occur.

This raises the question, was the timing really indicitive of poor fuel usage, or suitable for the conditions at hand?

If you are getting MPG from something like injector pulse widths, then we can assume that the injectors never were going into overrun or coast and fuel was being used 'less efficiently' than your normal driving.

But this raises an alternate question, was it really a matter of 'resetting' the ECU, or was the ECU responding correctly to conditions that were altered by shutdown? This is where I find logging a wideband combined with OBD-II data pretty useful. Instead of 'it went away', I can look at the log and get clues for exhaust pressure problems, a clogged injector, bad shift points, etc.

I actually think that even a crappy wideband is pretty useful, because, like EGT, you are looking at the combustion results, what the ECU accomplished. But, unlike EGT, you aren't as easily fooled by abnormal combustion (though a slow wideband can still be fooled, because you are looking at a running average of measurements fed into a PID loop). A fast wideband is nice because you can see, say, individual misfires as spikes in the log. This stuff is impossible to see on a live gauge but when you look at a log you can spot things like a lean cyl or occassional spikes.

None of this says you are wrong, I'm just saying that without knowing how the FE reading was originally derived, and without any engine data to look back at, it is hard to know, for certain, what was really occurring.

-jjf

I was at steady state highway travel about 58mph in 4th gear.

The readings were off the scamguage.

The motor was warm and in closed loop

Previous experience has shown that in open loop my FE on the highway is about 27mpg (don't ask how I know)

It appears that when my timing at highway speeds (again previous experience) is in the 20's my fe is the same as open loop which also has timing in the 20's

I have had experiences across the board on different vehicles where fuel economy would drop for no apparent reason and stay there while lightly feathering throttle going down a hill to maintain speed. In those cases however I could let off the gas and wait for the GPH meter to bottom at idle rates then my FE would go back up.

I have sometimes heard of this called detroit fever or a long term correction.

From what I have learned the motor tries to operate at stoich but very rarely does except on the highway at steady state and even then few motors actually operate at true stoich but usually a few degrees below, exceptions being very modern motors. Even light acceleration usually drops you out of stoich.

Also note is was quite cold, many motors go a bit excessive on timeing & A/F corrections when it is cold thus why many suffer MASSIVE FE reductions even on long trips, couple that with "winter gas" and you have a craptaskic combination.

Cheers
Ryan

OK, now we have some clues. But the first thing to bear in mind is an OBD-II MPG gauge is an educated guess, not a precise technical measurement. The reason is that there is no general PID for 'fuel flow' or 'injector pulse width'. Instead, there are basically two paths to a measurement.

MAF -> Guess at operating AFR -> combine with VSS -> display 'MPG'

Combine MAP, RPM, and IAT -> Guess at VE -> Guess at operating AFR -> combine with VSS -> display 'MPG'

Sometimes, the guesses are wrong, that's why you generally have to 'calibrate'. It isn't just tweaking for precision problems in the sensors, but driving style and even common routes.

Now, this could start a whole bunch of seperate arguments, so I want to reiterate, I'm not saying that the measurements are useless. They are not, they have a strong statistical correllation to actual fuel consumption. I'm just saying that when you need a 'fudge factor' you have to take the precision of individual measurements with a grain of salt.

This is the problem with distinguishing correlation from direct cause and effect. The optimum peak pressure point in the engine is fixed by geometry, it is built into the design. You are looking at timing in DEGREES, but to understand timing, we have to think in terms of TIME.

At it's simplest, this is easy to visualize. When we are at idle, timing generally drops to near zero, this is because the piston is moving slowly. As RPM goes up, we have to pull the timing back in TIME to have the burn hit the same peak pressure point.

But there is another dimension, the burn of the mix is not a fixed interval in time. It is effected a great deal by turbulence, or swirling of the mix, AFR, and by temperature. Temperature may be the biggest factor.

The reason that there is a statistical correlation between economy and high timing numbers is that the style of driving is lower RPM's and lots of time off the throttle. When RPM is low, you have to move back in degrees, because the slower speed of travel on the piston. When you are coasting and letting off the gas a lot, things have to pull back because your effective mixes are lean and the 50% burn is slow. 'Lean' doesn't have to mean air, it can also be exhaust gas.

This is the part of the puzzle I missed initially. Under load, high numbers are bad, but you are striving to drive under low loads.

We could tell for sure with a wideband in your exhaust, but it seems much more likely to me that this is simply a case where the FE reading is just plain wrong. You seem to be describing overrun conditions. In that case, the injectors are often closed and the tailpipe is full of free air. Since the MPG gauge is not looking at injectors, but the MAF or MAP side and then presuming fuel use, it is reporting poor FE when you are not using any at all.

Try setting your scan tool up to show one of the O2 sensors. They are easier to look at as a bar graph than numbers, but when it is oscillating, you are chasing stoich, when the voltage is at the higher end, you are richer than stoich, when it is pegged low, you are lean, and frequently in overrun.

Actually, they run at stoich a lot. As much as possible. Again, you can tell by looking at the O2 sensor. The narrow band sensor detects a very narrow range around stoich. It is very accurate at finding stoich because the chemistry actually reverses as it is crossed. If the O2 sensor is oscillating, the ECU is 'chasing it', and you are there. If it is reading high, you are enrichened, usually under fairly aggressive accelleration (when you kick the ECU into "Open Loop"). When it is low, you are lean, typically in coasting and recirc situatations.

It HAS to hit stoich a lot for the CAT to function at all. Cat efficiency plummets as you move off stoich. This is because the high EGT is needed for the cat processes to occur.

Emissions is the ONLY reason engines are built to target stoich. Best Power is richer, Best Economy is leaner.

Think about this for a moment. Small airplanes LOVE cold weather. Because they are underpowered and often run full throttle. When the air is cold, it is denser. More air molecules means you can mix in more fuel, which releases more energy, which translates into faster climb rates.

But when you are in your car, you don't say 'OH BOY, It's cold, I can go *faster*!'. You generally drive at the same speeds. But what happens to the air charge at the inlet?

Just like the airplane, it is denser. So, to go the same speed, you throttle back more. The more you throttle, the more vacuum is created behind the throttle and the lower your efficiency. The cold air gives the engine more performance, but we have no use for it, so we operate the engine at lower effiency to meet our demand on it.

To better understand why, take a look at why high compression engines are more fuel efficient that low compression ones. It is the the same fuel and air, but the top of the V/P graph is taller and sharper, so we get a greater effective change.

But, that is generally not why the timing is "aggressive". Remember, burn rates are highly influenced by temp. It would be counter productive for the ECU to run at the same timing at low IATs that it does at high IATs!

As far as the 'low FE', there are two possibilities, one is that it is actually reporting the inherent loss in efficiency in throttling back more because of the higher air charge. The other is that it is just more wrong than usual about guessing at fuel consumption. Remember, it is primarily looking at the air side, not the fuel side.

Going just on timing is probably a bad idea. For example, when you shut down you car in the cold weather, cooling happened quickly. When you started up again, there was a shift in timing. But there is no reason to believe that the ECU suddenly decided to be more efficient. It is more likely that a temp change in CHT changed burn and swirl. Rather or not this resulted in better or lower true fuel efficiency is hard to say. In general, the cooler engine throttled back would be less efficient, but there are lots of exceptions.

-jjf

Jfitzpat,

Please tell us which company you work for and the products this company sell. I'd like to know and so do other members. I think it could be of interest to some people here.

More and more recent cars have slow wideband O2 sensors. Can't you get a better estimate of fuel usage by simply combining MAF and AFR. I'd think that would be pretty close to reality. With my Elantra I have set up X-Gauges for average trip MAF and Lambda. So far it's been pretty accurate. I just wish I could have an X-Gauge displaying a result based on 2 PIDs...

There's maximum brake torque timing for every operation point. Your goal is to be at that point all the time. Stock timing maps are often slightly retarded from MBT timing, so that's why we say more advance is usually better, but up to a point obviously.

On my TDI I advanced timing quite a lot since stock it was retarded (more advance means more pressure and more stress on the components), and I run higher than stock EGR rates, which slows the burn rate, and the ECU was not compensating for it.

On recent gassers with knock sensors timing is often kept at the most advance possible by the ECU, but then again I'd like to have a switched resistor on the IAT sensor for low load cruising to see if I could gain a degree or two. There are no huge gains to be had, but every bit counts.

If you just look at the ING value you can almost think of it as a vacuum gauge, or for a given operating point it can serve as a diagnosis test. Generally speaking, all else being equal, it changes proportionally to RPM, and inversely proportional to load.

If you are looking at a wideband, MAF, and your odometer, you should be very close to reality. But there are some inherent errors.

First, the gauge might be reporting AFR, but the sensor is actually measuring 'equivalency ratio'. That is, it is measuring how much O2 must be added or removed to reach stoich in relationship to partial pressure.

This number is inverted and reported as lambda, with 1.0 being stoich. To report 'AFR', lambda is multiplied by a fixed factor for the fuel, say 14.7 for gasoline, but that is really just an approximation, the exact ratio will vary on blend, and even fuel temp.

Eq. Ratio is actually better for engine control. You don't care about AFR, you care about combustion result. When the engine needs stoich for max EGT, you get stoich. Ideally, the ECU could run closed loop on the wideband, then when you hit .84 lambda, you'd know you are at best power.

But the way the sensors are generally measured is slow. The chamber is pumped to stoich through a PID loop. Also, for durability the sensors themselves are often slow. So, typically, the wideband is used to help the ECU make decisions, but not tightly coupled, in a stock auto.

There are a few other cases of error. For example, the sensors are typically calibrated at sea level, standard temp. But the PP of O2 changes as we deviate from this, which translates into error. Also, your AFR reading probably pegs at a value when you are super lean (injectors closed).

But, all in all, you've hit on a pretty accurate method. MAF is a great way to monitor air, and lambda gets you very close to true AFR when running pretty much everywhere but very lean (when only small amounts of fuel are involved anyway).

In general, I think that measuring IPW pulse by pulse is probably the tightest monitor only measurement, but you still have variables like injector on/off behavior.

As far as who I work for, the simple answer is I report directly to the president here:

EmiSense Technologies, LLC: Smart Sensors - Clean Emissions

But it is slightly more complicated. But I'll put an introduction in the proper area later today or tonight.

-jjf

Yup, you are right. My 0-5v stock sensor pegs at 23. That's how I make sure I hit the deceleration cutoff.

I'm having trouble picturing that, but it isn't my area at all. I generally think of EGR only in relationship to NOx emissions. I tried searching, but appear to be search challenged. Have you posted some other threads on this?

-jjf

FWIW, most sensors bottom out rich pretty easily, I typically use butane, but they can go quite a bit leaner. However, 0-5V outputs are normally limited so that they give more resolution to the analog input on the receiving end in the range that is of most use.

For example, our 0-5V wideband output is .5 lambda to 1.523 lambda, which gives a 10 bit ADC .001 lambda resolution across typical gasoline combustion ranges. But when we are outputting digitally we report readings waaaay leaner, until we reach the pump limit of the sensor, then we just report the partial pressure of O2.

Thanks for the info!

-jjf

jfitzpat -

That has been my observation. Between the closed-loop parameter on the ScanGauge and watching the (narrow band) 02 sensor on my digital AFR, I see stoich almost all the time. In fact, the way I drive, I almost never see open-loop, even when I accelerate. I think the ECU "cheats lean" under steady-state conditions. When the ECU is chasing stoich, it will go rich by up to -0.2 below 14.7. On the other hand, I will sometimes see it momentarily go lean at +0.9 over 14.7 without me doing anything. There is a part of me that wonders if I trained the ECU to "drive lean" as a function of my driving habits.

CarloSW2

Doubtful. Too much O2 in the exhaust is bad for the emission control system, so an ECU will generally avoid it.

I suspect that it has more to do with what, exactly, you mean by "digital AFR". One possibility is that the reading only has meaning over a narrow range and these 'lean spikes' are under reported. That is, overrun conditions, which happen a lot when there is a lot of coasting, are not being reported for what they are, free air in the exhaust and injectors closed.

You see a similar misleading reading quite a lot in performance automotive applications. Folks will add a mod on the inlet side for more air, then the points will be out of temp range at low loads. Spark plugs, etc. have a surprisingly narrow operating temp range in their designs. Once temps are too low, occasional misfires occur. Because many 'AFR' gauges are basically putting pump current back into a PID loop, these small parcels of uncombusted mix are not reported as spikes, but instead shift the overall readings lean for a bit.

I've lost count of how many times we've been accused of reading 'too rich', which is actually very difficult to do with our measurement principle. When we actually go out and look at a log instead of just a gauge, we are clearly correct and the abnormal combustion events are clearly visible.

I don't mean to come across as badmouthing other gear, I'm just stressing that you need to know the limits of any instrument. A conventional current based wideband is perfectly suitable for many applications, you just need to know when it will lead you astray. Like misfires shifting the reading lean. If you enrichened based on that false shift, you would probably create fouling and make the misfires worse.

-jjf