Fueling without cam sensors

[LakeWylieJoe_imp]

This is a somewhat down-in-the-weeds technical question that I have not seen addressed before.

I just finished my third valve clearance check and adjustment at 24,000mi on the odometer. A few valves were at the upper or lower spec limit each time I've inspected them and I always set them back to the center of the spec range. It seems foolish to me to leave any at either extreme end of the range because if they drift at all in that direction, then it's out of spec.

Anyway, while reinstalling the cams, a question occurred to me: How does the ECM know if a particular cylinder is TDC on the compression stroke or TDC on the exhaust stroke? Unlike most modern engines, the CB1100 does not have cam position sensors; it relies completely in the crank position sensor(CKP). The CKP can only give information that a cylinder is TDC, but not whether it's compression stroke or exhaust, which is 100% dependent on the cam positions (which turn at exactly 1/2 the rate of the crank).

The ignition system doesn't need to know the difference as it sparks the #1 & #4 cylinders together and the #2 & #3 cylinders together. A spark is wasted during each firing because only one of the two cylinders associated with each coil fires is needed for a given crank cycle. A wasted spark is immaterial, though.

However for the fuel injection, you can't fire a fuel charge into a the back of a closed valve then again into an open port on the next crank cycle, so how does the ECM know when to fire the injectors? The system must determine if a particular cylinder is at TDC compression or exhaust somehow.

On the single cylinder CRF dirt bike (which also has no cam sensor), the CKP can sense the slightly-slowing crank rotation during the compression stroke and the increasing rate on the power stroke to determine when TDC compression is. That wont work on an inline 4-cylinder, though, with a power stroke every 180 degrees of crank rotation.

Anybody of you smart guys have any thoughts on how Honda overcomes not having cam sensors with the CB1100 motor?

 

[pekingduck]

That's a good question. On a single, it could take a MAP sensor reading to sense intake, but on our CB1100, the MAP sensor reads the average from the airbox.

Most new Hondas don't use a cam sensor (except new GL1800)

 

[peterbaron]

CrPS is a magnetic or Hall effect sensor that generates voltage with cooperation of a tone wheel. Missing tooth from a tone wheel, ring gear or flywheel is TDC. These two tell ECM the exact position of the piston. Crank to cam = 2:1 = 720*:360 = one full cycle of four stroke engine.
ECM uses its program to ground out particular injector, while B+ supplies constant power when IGN is on.

Tone wheel with missing tooth/TDC

This missing tooth/TDC will make a different kind of pulse as seen in pic below

CrPS pattern with TDC signal

CrPS + missing tooth + ECM do the trick, this is how I see it.
Not sure if CB uses tone wheel, ring gear, flywheel or...?? as a TDC reference, but must use something.

 

[pekingduck]

CB1100 only has a crank pulse sensor on the end of the crank (with "missing teeth") that determine crank/piston position. Ignition and fuel mapping can then be determined by the ECM with input from the other sensors

But since the camshafts are spinning half of crank speed, how is it determined which if the two crank rotations the ECM supplies fuel on? To paraphrase LakeWylieJoe, there can be a wasted spark, but not a wasted squirt. So how is that determined?

 

[peterbaron]

To my best automotive knowledge, all it cares is to see TDC of cyl #1, then ECM program will take care of the rest, knowing its firing order = 1,2,4,3...
...when I am riding my CB, I do not care how it works....I'll fix her if she troubles, IF she does

 

[pekingduck]

The latest Honda CBR1000RR uses a cam pulse generator, as does the latest Honda Fit. But the NC700/750X twins, which base their engines on the Fit motor, do not have one.

 

[peterbaron]

Honda engineers always the best and surprising

 

[Ben70_imp]

If anyone has a multi channel oscilloscope available, we could hook into cyls 1 and 2 injectors and determine if they are firing at the same time. If so, it would be ported fuel injection with a waste pulse, or possibly continuous pulse like the old throttle body injection on cars.
If they don't fire at the same time, it would be sequential fuel injection.
I don't remember reading anywhere that it specified sequential fuel injection.
Interesting point to determine, in any case.

Ben

 

[the Ferret]

Like reading Kevin Cameron lol

 

[pekingduck]

If anyone has a multi channel oscilloscope available, we could hook into cyls 1 and 2 injectors and determine if they are firing at the same time. If so, it would be ported fuel injection with a waste pulse, or possibly continuous pulse like the old throttle body injection on cars.
If they don't fire at the same time, it would be sequential fuel injection.
I don't remember reading anywhere that it specified sequential fuel injection.
Interesting point to determine, in any case.

Ben

Throttle body injection worked best in multi-cylinder applications like V-8 and V-6 where there are multiple intake ports constantly drawing the mixture in. Most bikes and cars now have sequential/multipoint/whatever injection (except some Triumphs and Guzzis, NC700X.750X twins, and Gold Wings), meaning one injector per cylinder.

With that, there would be a waste of fuel half the time. It wouldn't be a 50% waste of the fuel, but still noticeable. Bikes with the cam pulse signal generator do not get markedly better mileage.

If there was a waste "squirt", you could swap the injector connectors for #2 & #3 cylinders, and it should run the same.

 

[GoldOxide_imp]

Fueling without cam sensors can also be achieved with engines without camshafts, or cam-less engines.


Pretty advanced technology, but should not really be a surprise in the year of our time 2019.

 

[pekingduck]

its not that hard to program. the cam is a direct relation to the crank. The code will just count the pulses and interpolate the position. its not a single pulse. fyi, theres NO map sensor on the grom. you can do it simply with the crank signal, engine temp, some sort of an 02 (unless you are running a static map), and tps.

Yes, you're right, no MAP on the Grom. But again, how can crank signal know which of the two revolutions of the crank in relation to the cams?

Crank and cam are not a direct 1:1 relation, as it takes two revolutions of a crank to turn the camshaft once. On each revolution of the crank, the crank pulse sensor reads the exact same signal. Yes, the ECM can be programmed to fire and fuel on alternate crank revolutions, but how does it establish which of the two revolutions?

If it is all in the ECM, then if the ECM is unplugged, the crank turned once, the ECM plugged back in, the engine will always start normally.

That suggests that all bikes would not need cam position sensor, yet some do.

 

[max]

LakeWylieJoe I like this puzzle and enjoy all the answers we get, you are right, how does the bike decide if it starts with the left foot or the right foot:huh:

It has never been a problem with carbs and the waisted spark, i got that but this is a new trick, i have three options and they are all different, in the mean time i have deleted the last one, our bike does not fire the injectors at the same time for cylinders 1 and 4 ( like the spark ) at half the dose when it is running but it has worked out which piston is on the intake stroke and doses each injector accordingly every time it starts.

but how? maybe the ecu has electronic dice?:rofl:

Never given it a thought but it keeps me thinking and enjoy some more answers, which is a good thing:celebrate:

 

[GoldOxide_imp]

Not suggesting the immediate demise of the [woefully antiquated] internal combustion engine, but the discussion of cam synchronization, maximum efficiency tweaks and so on - I wonder if analog discussions were going on during the final years of the stream-driven locomotive? We are likely living the era of the best known methods to squeeze every joule of energy out of every drop of dino. This could go on for awhile, but sadly it will eventually be dust in the wind and displays in museums and retro restaurants.

 

[pekingduck]

Not suggesting the immediate demise of the [woefully antiquated] internal combustion engine, but the discussion of cam synchronization, maximum efficiency tweaks and so on - I wonder if analog discussions were going on during the final years of the stream-driven locomotive? We are likely living the era of the best known methods to squeeze every joule of energy out of every drop of dino. This could go on for awhile, but sadly it will eventually be dust in the wind and displays in museums and retro restaurants.

We will be the dust in the wind, and school kids will be tele-ported to the i-phone museum, to see what communications was like before microchips were embedded in their brains......

 

[GoldOxide_imp]

Not suggesting the immediate demise of the [woefully antiquated] internal combustion engine, but the discussion of cam synchronization, maximum efficiency tweaks and so on - I wonder if analog discussions were going on during the final years of the stream-driven locomotive? We are likely living the era of the best known methods to squeeze every joule of energy out of every drop of dino. This could go on for awhile, but sadly it will eventually be dust in the wind and displays in museums and retro restaurants.

We will be the dust in the wind, and school kids will be tele-ported to the i-phone museum, to see what communications was like before microchips were embedded in their brains......

We will be the dust in the wind, and school kids will be tele-ported to the i-phone museum, to see what communications was like before microchips were embedded in their brains......
:rofl: Yeah, I somehow don't doubt.

 

[zirconxi_imp]

hmmm... interesting... could it be the computer always tried twice, only during engine start, to know if its power stroke or intake stoke? The mark gear will already tell us which pair of cylinder is at the bottom and which at top. when we crank, computer will inject into a pre-pick cylinder at bottom. if its alive (means its the correct cylinder), continue with the sequence. if no go, will inject again when that same cylinder go bottom again. This way, maximum only 1 injection is lost during startup. guess once and go if correct, reset sequence if wrong. But maybe Honda engineer have a more canny way...

 

[zirconxi_imp]

Stuck in jam for an hour driving, i think of another way to do it. . Just do it like those single cylinder engine. By trying to light up only 1 cylinder. Once u got 1 right, the sequence for the rest are known.

 

[LakeWylieJoe_imp]

pekingduck is making the correct point (in a lot fewer words than I did).

Also, I'm certain the system is not programmed to inject fuel into a closed port--even a carburetor works smarter than that.

Whatever Honda figured out, it's very clever. I'd love to get to the bottom of this. The mechanics at the local Honda shop are also scratching their heads after I mentioned the issue yesterday while buying some oil. Maybe I can get one of them to make a call to a corporate tech guy.
For the sake of being complete, here's what got me thinking about the issue.

When the cams come out, you have to pull at least one of the cam sprockets off as there is not enough slack in the cam chain, even with the tensioner backed all the way out, to remove the cams with the sprockets on. To get a sprocket off, you have to pull one of the two sprocket bolts out (you start with the cam timing marks at #1 TDC on the compression stroke to access the first bolt) then turn the crank by hand one full cycle to expose the second bolt on the opposite half of the sprocket. You are now at #1 TDC exhaust stroke. Once both sprocket bolts are off, you can work the sprocket loose to one side and pull the cam all the way out through the sprocket hole.

Handy tip for first-timers: You only need to remove the sprocket from the intake cam--there is plenty of slack in the chain to get the exhaust cam out with its sprocket still on once the intake cam is out. The intake one has to come off because it's on the slack side of the chain.

The first two times I adjusted valve clearances, I dutifully turned the crank over one full cycle with the cams out (back to what was #1 TDC compression stroke) before reinstalling them with their timing marks in the correct position. You then have to turn the crank one more cycle to get the second sprocket bolt back in.

This time, I got to thinking the ECM can't be counting crank cycles to figure the cam timing--the battery isn't even hooked up! I put the cams in 180 degrees out of phase from how they came out (due to the one full crank turn to get the second sprocket bolt out). I figured as long as #1 was TDC on the crank and the timing marks on the cams were correct, it should work. Sure enough, it fired right up once reassembled.

Again, how the does the ECM determine cam position solely from the CKP? For every turn of the crank, the cams only turn 1/2 cycle. In other words, for every possible crank position, there are two possible cam positions, and the fuel injector timing depends on the cam position.

 

[max]

For the benefit of mr Norris et all ( not mr K ) here is a video clip that shows in slow motion what the injectors are doing when the engine fires up from cold.

Two led's are connected across injectors 1 and 4, they light up when the injectors are operated, there is one double firing at the start button push and then another four double shots of fuel into both ports.
After what seems like a small pause the no1 cylinder fires and then no4 cylinder and so on.

It appears that the double dose just gets the ( cold ) engine going and the decision ( electronic dice ) is then made by firing one cylinder at a time and waiting for the crank sensor to report a higher frequency at which time success is registered and the engine runs in the " uneven " sequence.
There is a lot more to this of course but now we know a bit more about what actually happens.

You all may have noticed that it takes the cb1100 to always take a bit of time to start up even if it is warm, the next question is; what if the engine is bump started, does it remember in what sequence it was before it stopped?

The clip is slowed down to 7% speed but the pitch of the sound remains the same hence the eery soundtrack, the unusual firing of the led's once the engine is running is the result of the frame rate and rev's mismatch and appears out of sync, in reality they fire at the correct time and sequence, i had a hard time getting this part as best i could on video;

And Joe, would you be able to tell us what the clearances and shim sizes were before and after so we can complete this thread here;
http://cb1100forum.com/forum/showthread.php?tid=8882&pid=139671#pid139671