Hey guys,

Well, I have a Daytona Twin Tec TCFI unit installed on my 96 that makes data recording of every engine parameter.

So, I was checking for how long the injector was working on high TPSs and RPMs...

The log shows 25ms @ 5500rpm, full throttle. Let's say 1ms = 1s/1000

So that's it 0,025s

But how long it's an intake stroke?
I'll have to do some simple math.

Well if an engine is reving at 6000rpm (it's almost the same for 5500rpm, just rounding) we have:

6000rpm/60seconds = 100rps (revs per second)

So, how many seconds it takes to do 2 full rounds (720 degrees or 4 strokes)?

1sec / 100rps = 0,01sec/revolution

We need half round to havejustaintake stroke or 2 rounds to have the four strokes.
So the intake stroke lasts 0,005sec and
four strokes last 0,02secs

Well, if my injector pulse withd is 0,025secs, that means 5x longer than an intake stroke!!!!!
And if you think (as I did) that the injector is working all the time, well 0,025sec is still bigger than the 0,2secs of all four strokes, so 0,025 secs will be over the next four strtokes!!!!

This makes no sense at all, I'm pretty sure I'm missing something here, or my math is totally wrong (hope not)...

Could someone give me a light on this issue?

 

Good catch.....

5500 rev/min = 91.666 rev/sec

1/91.666 = .01091 sec/rev or 10.91 ms /rev

So this gives 10.91 ms * 2 = 21.82 ms for 2 full revolutions of the crank shaft. At 25 ms you are exceding the 21.82 ms availableso the commanded pulse with is exceding the available time.

The realistic max pulse width of a 5500 RPM motor is 85% duty cycle or 21.82 ms * .85 or 18.55 ms

The part of the math that you are missing is that with most 4 stroke gas motors the injectors spray the majority of the 720 degrees of the crank rotation. With a Harley I would think the injector timing is fairly sensitive a idle and low RPM but thay don't want to talk about that stuff and they don't give you the chance to mess with it in Harley sofrware. Google "batch EFI" and "sequential EFI". This stuff is HUGE in diesel EFI.

Bottom line is somthing is wrong with your bike.Possibilities are...
Your engine or injector size calc in the program setup is incorrect.
Your injectors are too small.
Your fuel pump or filteris not keeping up and fuel pressure is dropping.
Your wide band could be messed up or reversion is sucking in air.
You could have an exhaust leak sucking in o2

Most data logging software reports Duty Cycle to warn the the tuner of this common issue. The SERT software assumesthat tunershave the math background to run the critical math on the side. Or that they have purchased other things (software or hardware) that does run the math.

Good luck and I hope this helps

AW

 

In this log you can see me downshifting thru the gears. Coasting for a little bit and then nail the throttle in first. You can see the duty cycle come up to a max of about 55% and then see the Delphi software hitt the rev limiter and jank fuel from thhe front cylinder and then later the rear cylinder. The min duty cycles in that log are running about 2.0% of 720 degrees or about 14 degrees of crank rotation.



Hope this helps

AW

 

Duty Cycle is measured in cam shaft rotation not crank shaft duration. You will need a larger injector when you see 20ms at 6000 rpm, 20ms at 3000rpm poses no problems...only at 6000rpm @20ms (which is 80% duty) will the injector tail bite itself. Most injectors you can buy are rated in lbs/perhour. We need .5lbs/per hour for every horsepower per cylinder. IE...100 hp / 2 = 50 hp 50hp x.5 = 25lbs/per hour. In the SERT program the injectors are listed in g/sec (grams /pre second) So I attached a calculator to convert every thing you need to convert to fine the correct fuel injector for your build. One other thing is the injector co. rate the lbs/per hour at 43 bs fuel pressure....Harley has a pump that puts out 55 to 62 lbs so this is a another factor needed when finding the correct injector and again here is a calculator to do that for you. Example: Harley injector 3.91 converts to 31lbs/h @ 43psi
Harley injector 4.35 converts to 34.5 lbs/h @ 43psi
Harley injector 4.89 converts to 38.8 lbs/h @ 43psi




http://www.rbracing-rsr.com/fuelinjectors.htm

 

Anytime I find a problem I ask, "what is the last thing I touched that would cause such a problem?" If the answer was, "installed a Daytona Twin Tech," that would be my first suspect. I would question the pulse width numbers,possibly actually1/2 of what it is displaying would be a ligit number.

 

Could you elaborate on that 5562 lbs number. 5562 is the weight of two small cars.

The other thing that can get into these calculations is the number of squirts per 720 degrees of crank rotation. I think Harley uses one but I have yet got around to throwing the duel trace scope on my bike to figure that on out. Also some (if not all) of theinjector manufactur's rate their injectorsat 100 DC meaning they put exactly 12 volts DCon the injector for 10 sec and report how many CC of fuel end up in a cylinder. Others machines put a 12 volt square wave at 80%duty cycleand say some frequency ofequilevent to XXXRPM and measuer that number of cc in the container. The other thing that confuses all of this is the first 1ms of PW does not deliver fuel at all as the injector charges up to open.

This is all taken into account in this spread sheet. If I remember tha sheet assumes that the injector is rated at 12V dc and it backs into the required injector that will be close.

http://www.ncs-stl.com/fuel/ReqInjectors03.xls

These numbers should get you in the ballpark for a stock Sportster


[align=left]Target HP 55[/align]
[align=left]Max RPM 6000[/align]
[align=left]# of injectors 2[/align]
[align=left]Brake Specific Fuel Consumption (see Right) 0.55[/align]
[align=left]Max Duty Cycle (< 90) 55[/align]
[align=left]Fuel Pressure 30 - 65 (43.5 normal) 55.0[/align]
[align=left]Engine Effencity (aprox) 85 street to 110 race [font=arial][size=2]90[/s

 

Well if I remember my math correctly we need to figure out how much time we have to get the fuel into the cylinder first. So to figure that out you take the following math to do it.

60,000/RPM = Ms per rotation

So an example would be

60,000/6000 = 10 Ms

So now you can only inject the fuel so long so you need to divide that number down to what you think you can get away with.

Let say you think your cam has the intake valve open for90 deg. Since the 10 Ms is 360 deg divide the number by 4

10 Ms/4 =2.5 Ms

So at 6000 RPM and 90 deg of injection time you only have2.5 Ms to get everything done! So what is really going on is the injector is spraying fuel into the intake tract for much longer than the intake valve is reallyopen. What you have to be carefulwith is known as wall wetting. Wall wetting occurs from having the injector on too long prior to the valve opening. This allows the fuel to stick to the walls and become a liquid and run into the cylinder when the valve opens.......... not good. So there is a lot more to it than just duty cycle to worry about when properly sizing an injector.

 

Steve,

So now he is getting ten times the max answer and the rest of the fuel is dripping off the intake valve and getting aride in the hot exhaust as polutionor unburned hydrocarbons?

If anyone wants to learn more about this stuff, google x-tau, but this stuff is real painful reading. Cool stuff though. This is the stuff acceleration enrichment is all about. There is a book by John B. Heywood that is really good at explaining this stuff. Internal Combustion Engine Fundimentals that is worth the price of admission but it is pricy.

A little side note This stuff can be a big deal at low RPM and high map load where fuel flow is fairly low, MAP is fairly high and the motor wants to respond like a wet mop. At high RPM and huge air flows it tends to be all about AFR and getting the VE correct. 100% DC is going to be a problem and we need to get to the sourse of the issue. When the injector transitions frome 85% thur 100% DC it is totally unpredictable as to the fuel flow. You may be able to google injector impedance saturation throwing in Duty Cycle may narowthe search down a little

AW

 

Sorry it was the Harley pump pressure.....55 psi to 62 psi I thought I did it like this 55/62 but I missed a key.

 

Well, first, thanks!

- I use the DTT TCFI for a couple of years on several different engines and to say the truth is the first time Icheck the Inj Pulse Withd at high rpms, or at least was never an issue but now I'm curious, I'll check my old logs(later). Anyway it was a 88ci with a 4.35g/s injector, so is not comparable.

- The AFRlooks ok, I use 12.3 in WOT and the (nearly stock) enginesounds well at WOT/High RPMs, the only thing I notice its really hard to go over 115mph on this 96ci (6th 5th or 4th), differently on the 88ci where I used to go over 125mph.

- The other point is, even if I use a bigger injector, let's say a SE 4.89g/s, that means:

4.89/3.91 = 25%

25ms - 25% = 18.75ms

18.75ms is still more than the 18.55ms supposed to be the 85% of the duty cycle.

- Leakings. Well, I'll have to check the exhausts, but the probabilty oftwo exhausts leaking the same way at the same time sounds quite difficult. The other point is, at 5500rpm full trhottle the volume of gas flowing out make me think it's impossible some cold airget into the suposed gap.

- Well, if the injector is during more than the duty cycle, how the ECU is making to count from 0 to 25ms if the four strokes last 20ms??!?

- I use fuel with 25% of ethanol (I know sounds strange but its quite normal here), wich means around 10% more fuel flow needed than regular fuel. Is the same if I had an 103 engine or something. I don't see people changing fuel injectors on a 103 engine, so I can't or don't want to believe a 3.91g/sec injector is undersized for a nearly stock engine. Even the 4.89 should be small with the numbers I've made. I don't believe its right.

So, I have two more probable answers:

As whittlebeast said: 1) "Your fuel pump or filteris not keeping up and fuel pressure is dropping."
Or, as mentioned by geargrinder, 2) "I would question the pulse width numbers,possibly actually1/2 of what it is displaying would be a ligit number."

1) How can I check the pump pressure? Or is there some way to increase the pressure on it manually?

2) Is it normal the ECU show wrong pulse width? Why you suggest the half value of it? Is there a way to calculate the real value, if this is the case?

I'm starting toclarify the problemnow, so thanks all for the aswers posted and to be posted,

Hernan