Yellow Bullet Forums banner

1 - 20 of 82 Posts

·
Registered
Joined
·
1,908 Posts
Discussion Starter · #6 ·
It is common with production vehicles to utilize feedback from various sensors to adjust timing/fuel to the engine. When the fuel system is using these sensors to make changes to the fuel, it is called running in closed loop (as you already know). Timing varies the same way in this closed-loop system. This is not common in race cars.

Anytime an engine is at WOT and is accelerating, the faster it accelerates (for a given load or weight), then the more power it is making. If we can make a slight change to the timing and (under WOT) see an increase in the acceleration rate, then the motor is making more power and therefore the car is going to go faster (as long as traction is available). Currently, the user makes all decisions on where the timing should be to make the most power, and as we all know, very few users are “experts”. Why not allow the ENGINE to determine where the timing should be to make the most power? If you had a “timing window”, and allowed a controller to adjust timing (within the user’s specifications), then the engine’s acceleration rate would tell the controller where the timing needs to be. For example: We have a nitrous car on two stages and the user has programmed 15* total timing on two stages. He makes a pass, and it’s a good, solid pass. Is there more left in the timing, or is the timing perfect? That depends on his experience/expertise. What if he adds a degree and it picks up power? Does he keep adding more and more until it no longer picks up mph on the time slip? If so, what if he goes too far and runs the motor into detonation? Will he catch it in time before hurting parts? Now, with that same scenario, let’s say a controller had a “closed loop timing” function. The user inputs the parameters: Time from launch input until off, how many degrees above the “baseline” to adjust, how many degrees below the baseline to adjust, etc. Now, with the functionality turned “on” at the release of the trans brake, the controller has the user input baseline of 15*(on two stages in this scenario), with the closed loop system allowing an increase to 15.5*, and decrease to 14.5*. So, with a 1* “window”, the controller will adjust timing up or down depending on the acceleration rate of the engine. If the controller commands 15.5* timing, and the rpms climb faster than the pre-set baseline, then the controller continues to hug the 15.5* timing curve throughout the run (like riding the dots on a slew rate), constantly getting feedback from the rpm/acceleration rate. The data acquisition would see this, and the user could increase the baseline to 15.5* for the next run, and the 1-degree “window” could be moved up to 16*-15*. Conversely, if the engine did NOT like that extra timing, then the engine gets into light detonation, loses power, the acceleration rate slows, and the controller goes the opposite direction with timing. As it hugs the 14.5* line, the engine is accelerating the fastest, and therefore continues to command 14.5*. The user can then move the baseline down to 14.5* and therefore the window would move down to 15*-14*. The user can set up how big this “window” is (1*, 1.5*, 2*, etc), and also set up how long the functionality should be active (5 secs after launch, 10 secs after launch, etc). The user can also set the baseline at 15*, with the controller adding .5* as the upper limit, and retarding 1* as the lower limit. So, the baseline does not have to be “in the middle” of the window, the user can program where the window is in relation to the baseline timing. The user can also dictate when the functionality starts (upon release of trans brake, 1 sec after release, 2 sec after release, etc). The functionality could also be used for traction control…if the acceleration rate of the engine is a certain range outside of the baseline, the user could tell the controller to retard by X degrees, then ramp back in.

Thoughts?
 

·
Registered
Joined
·
1,033 Posts
My thoughts are that would be hard to do effectively on a 2000 HP small tired car, and it would be a lot of work to get anything out of it. You just constantly fight traction on something like that, making it hard for that to work effectively/properly. I think if you had some type of in cylinder pressure measurement, you'd maybe be able to do something with enough development, but that would take a lot of effort and sharp people.

Just pretty much simpler to change the timing on a combo, and see how it responds. On the dyno, or on the track.

Neat/good thoughts, but I think it would take a lot of effort for something better suited to basic A-B-A testing on such engines.

On a low HP car, sure, you could likely benefit from something like that with enough work.

Not saying it wouldn't/couldn't work, but I think it would be difficult "in the real world" on high HP, traction limited cars.

For TC, that is done right now.
 

·
EFI/N2O JUNKIE
Joined
·
17,922 Posts
I echo what Doug said above. I saw a LOT of R&D for the US Government being performed at Ford a long time ago. Some really neat stuff going-on there.

Short story is this: They had pressure transducers inside each cylinder of a diesel engine (that technology is available to everyone today, but it isn't cheap!), and a crank sensor reading off the flywheel, there were a LOT of teeth on that flywheel, not your typical 60 tooth setup. They were able to see how quickly the crank accelerated and could see from that, plus the pressure transducers, when it went into detonation.

If I remember correctly detonation did not cost them power, until the engine came-apart. Pre-igniton WILL cost you power when you try to shove the piston/rod assembly back down where it was coming-from BEFORE it got past TDC.

They threw just about any type of combustionable fuels into those engines to see what did what, and I mean anything!
 

·
Registered
Joined
·
1,908 Posts
Discussion Starter · #9 ·
My thoughts are that would be hard to do effectively on a 2000 HP small tired car, and it would be a lot of work to get anything out of it. You just constantly fight traction on something like that, making it hard for that to work effectively/properly. .
I see what you are saying, especially in the early part of the run...but as the run goes on, you should start to see where the timing is adjusting itself for the most power. This is why you may delay the funtionality until later in the run in this type of car.

I think if you had some type of in cylinder pressure measurement, you'd maybe be able to do something with enough development, but that would take a lot of effort and sharp people.

.
Would that really be the best way? I think that timing advanced TOO far would increase cylinder pressure MORE than a properly timed motor, it's just that the peak cylinder pressure would not be at the best point of crank angle, and not beneficial to power. Rate of acceleration would be the best indicator to power, in my opinion.

Just pretty much simpler to change the timing on a combo, and see how it responds. On the dyno, or on the track.

.
Sure that's simpler...this would, imo, be a better way to get to the perfect timing number WITHOUT HURTING PARTS. If you do track testing and advance the timing, make a pass, and the mph slows, then you've lost power...and possibly could have hurt it already. With this concept, the controller would never allow it to hurt parts, as real time acceleration rates (or more importantly the LOSS of acceleration) would cause the timing to retard IN REAL TIME.
 

·
Registered
Joined
·
1,908 Posts
Discussion Starter · #11 · (Edited)
I think it would be too much work to rely on it. ]
How? Dial in .2* above and .2* below the timing that you would run normally...that way you are barely even using it. Slowly open the window as you feel comfortable. No work at all really...besides a couple key-strokes on a computer.
Personnally, I think it would be far more dangerous than what most people do...which is listening to what, "their buddy's timing is with X amount of nitrous" even though their buddy's combo is completely different than their own.
 

·
Registered
Joined
·
8,468 Posts
How? Dial in .2* above and .2* below the timing that you would run normally...that way you are barely even using it. Slowly open the window as you feel comfortable. No work at all really...besides a couple key-strokes on a computer.
Personnally, I think it would be far more dangerous that most people do...which is listening to what, "their buddy's timing is with X amount of nitrous" even though their buddies combo is completely different than their own.
You're not getting me. I'll give you an example....


The new Ford Coyote 5.0 has 2 knock sensors that Ford probably has millions wrapped up in. They work AWESOME. But not AWESOME enough, because it can't hear cylinder #8 good enough. So if #8 is knocking and not other cylinders are knocking, you won't catch it.

And that is with millions in Ford R&D. Knock detection is very advanced and I don't see it working out well unless people start working on kits for specific combinations, which they won't.

Knock sensors work great on OEM stuff because in general the engines are very quiet. Try getting a KS to tell the difference between knock and the other 10,000 noises coming out of a high power race engine. Add a power adder and it only gets tougher.
 

·
Registered
Joined
·
1,908 Posts
Discussion Starter · #13 · (Edited)
You're not getting me. I'll give you an example....


The new Ford Coyote 5.0 has 2 knock sensors that Ford probably has millions wrapped up in. They work AWESOME. But not AWESOME enough, because it can't hear cylinder #8 good enough. So if #8 is knocking and not other cylinders are knocking, you won't catch it.

And that is with millions in Ford R&D. Knock detection is very advanced and I don't see it working out well unless people start working on kits for specific combinations, which they won't.

Knock sensors work great on OEM stuff because in general the engines are very quiet. Try getting a KS to tell the difference between knock and the other 10,000 noises coming out of a high power race engine. Add a power adder and it only gets tougher.
What I'm talking about has nothing to do with a knock sensor. It also doesn't take the place of reading plugs or time slips. All that it does is allow a small controller-based change to the timing that you can datalog to see if increased (or decreased) timing is causing the motor to accelerate more. No knock sensors involved.

You very well could see a need for more timing, put it in, then see the need to pull timing away from hot cylinders (based on reading plugs), but this is something you would do anyhow (as you increase timing), even without this technology.
 

·
Registered
Joined
·
8,468 Posts
What I'm talking about has nothing to do with a knock sensor. It also doesn't take the place of reading plugs or time slips. All that it does is allow a small controller-based change to the timing that you can datalog to see if increased (or decreased) timing is causing the motor to accelerate more. No knock sensors involved.

Well a closed loop system requires some feedback. When talking about timing, this feedback comes from a knock sensor.

What are you suggesting you use for feedback? Accel rate? You already have that. It's called a timeslip. :)
 

·
Registered
Joined
·
1,908 Posts
Discussion Starter · #15 · (Edited)
Well a closed loop system requires some feedback. When talking about timing, this feedback comes from a knock sensor.

What are you suggesting you use for feedback? Accel rate? You already have that. It's called a timeslip. :)
EXACTLY what I'm talking about...acceleration rate. And yes, you have the timeslip after the run is complete. You have this REAL TIME during the run. Essentially, it's allowing the MOTOR to determine where the timing should be, NOT the user.
 

·
Registered
Joined
·
8,468 Posts
EXACTLY what I'm talking about...acceleration rate. And yes, you have the timeslip after the run is complete. You have this REAL TIME during the run.
Ok, so let's use an example....

Accel rate = 145 feet per second

You system has to take action, right? So what does it do? Add timing during the run to see if it accelerates more? What if it was going to be accelerating more anyway because you are applying more power?

Let's say it adds timing, how does it know it's added too much? Accel rate slows down? How quickly do you think the computer will know the car has slowed down? enough to save the engine because it's detonating because you added too much timing?

Basically I give you credit for thinking outside the box, but I don't think accel rate would be a manageable solution to evaluate your timing requirements in real time.
 

·
Registered
Joined
·
33,118 Posts
Late 80’s Nissan turbo V6 engines sold in Japan and Europe had strain gauge washers under the spark plugs that measured cylinder pressure and an ignition computer that managed the timing for each cylinder. The system adjusted timing for the next time an individual cylinder fired, based on the pressure peak location the last time that particular cylinder fired. I think it had a target of 15-16° ATDC for peak cylinder pressure and also detected knock. Been a long time since I read about it and I never saw one. I guess some of those used engines imported from Japan had the washers and the wire harness on them, but I never saw them. Google it.
 

·
Registered
Joined
·
8,468 Posts
Late 80’s Nissan turbo V6 engines sold in Japan and Europe had strain gauge washers under the spark plugs that measured cylinder pressure and an ignition computer that managed the timing for each cylinder. The system adjusted timing for the next time an individual cylinder fired, based on the pressure peak location the last time that particular cylinder fired. I think it had a target of 15-16° ATDC for peak cylinder pressure and also detected knock. Been a long time since I read about it and I never saw one. I guess some of those used engines imported from Japan had the washers and the wire harness on them, but I never saw them. Google it.
Much more expensive versions of this are used today and would really be the best way to do some tuning. VERY PRICEY

integrated pressure sensor:

 

·
Registered
Joined
·
1,908 Posts
Discussion Starter · #19 ·
Here is an example: Between the blue and red lines, my motor is accelerating 753rpms in 1 second. If the controller feeds it .5* of timing advance (my added blue lines/dots), and the motor increases to 780 rpms in 1 second (my added green line), then it liked the timing, and you can feel more confident in adding that .5* to the baseline timing for the next run. Do you see the delta sign showing the rpm increase? Also, the slew number? Power Grid already has the capability, it just needs a tiny amount of software written in to make it happen. If the controller added the timing, and it resulted in NO increase in acceleration rate, or a DECREASE of acceleration rate, then the controller can pull that timing out, and go the other direction (retard the timing). The box would be doing this in real time, faster than we can think...so it would be capable of changing timing as needed before any detonation had a chance of hurting anything. Besides, you set that timing window as tight as you feel comfortable.
 

·
Registered
Joined
·
8,468 Posts
Here is an example: Between the blue and red lines, my motor is accelerating 753rpms in 1 second. If the controller feeds it .5* of timing advance (my added blue lines/dots), and the motor increases to 780 rpms in 1 second (my added green line), then it liked the timing, and you can feel more confident in adding that .5* to the baseline timing for the next run. Do you see the delta sign showing the rpm increase? Also, the slew number? Power Grid already has the capability, it just needs a tiny amount of software written in to make it happen. If the controller added the timing, and it resulted in NO increase in acceleration rate, or a DECREASE of acceleration rate, then the controller can pull that timing out, and go the other direction (retard the timing). The box would be doing this in real time, faster than we can think...so it would be capable of changing timing as needed before any detonation had a chance of hurting anything. Besides, you set that timing window as tight as you feel comfortable.
This would only work if you had a base acceleration map to start with. I'd like to see you give it a shot and prove me wrong, but I don't think that would work well and/or there would be better ways of doing it IMO. Even though I don't like aftermarket knock sensors, I think that would work better.

There are just too many factors to adjust the timing based on accel rate quick enough. I can see it being used for traction control (already is) but not to add power.
 
1 - 20 of 82 Posts
Top