Setup on a single cylinder kawasaki kfx450r

[Guest |829|]

We are having problems setting up the v44 on our formula sae car. Its a single cylinder motor from the kawasaki kfx 450 r atv. When we run single point group injection we are not getting the injector firing but when we run it on other types it does fire. The injector is connected to the inj 1 pinout. why would this be happening? did we do something wrong?

Where would we want to start with making a base fuel map as we can not find any fuel maps for any single cylinder motor

All in all we cant get the engine to start but we do have good spark and we can get fuel.

[rayhall]

How many teeth on the crank trigger.

How many on the cam.

Also send me the map so I can look at the settings. [email protected]

Ray.

[Guest |829|]

emailed

[Guest |829|]

the crank position sensor is a reluctant 23-1

[rayhall]

Mike,

You have 22 teeth selected. The crank trigger has 24-2 (twenty four minus two), it does not have 22-2 (twenty two minus two)

Aux Ignition outputs 4,3 and 2 are set to Ignition. These must be set to off.

You have the maximum advance set to 30 deg and have over 30 in the ignition table. Change the maximum to at least 40 deg.

You are using a volt input with external pullup resistor for the ECT sensor, when there are spare Temp inputs, which already have a pullup. What value pullup are you using. I would recommend you just use a Temp input.

IAT (Intake Air Temp) correction is turned off and no temp sensor used. Very important you use an air temp sensor and have this table enabled.

Open Loop AFR table is turned off. I would recommend you turn it on as this will make mapping faster. This table has nothing to do with 02 sensor feedback. It is used by the ECU to make background corrections.

Injection mode should be sequential not multi group.

After you get the timing set accurately, you must set the ignition delay. This will make the timing stable as you rev the engine. Very important...

ECU has very old firmware. Very important you update to version 4.7.1.

Do the update to 4.7.1 and then load in the map I will email you soon. This map I will updated and made the changes listed above. Do not put this map in the ECU before doing the update.

Ray.

[Guest |829|]

thanks ray

We got the trigger figured out with trial and error it is actually 23-1 and runs with zero trigger fails. With that when cranking the engine would backfire so we know it needs less spark advance but are now wondering how the ECU is knowing when to send spark without the cam position sensor?

We will update the firmware and make any changes. Are there any other suggestions they will be greatly appreciated

Thank you very much

[rayhall]

Cannot except the crank trigger is 23-1. It must be 24-1.

The ECU uses the missing tooth position plus the trigger offset value to control the ignition timing.

The items I listed are the ones that need addressing. Will send the map soon.

[rayhall]

While you may get no trigger errors the RPM will never be stable with the wrong tooth count.

Also they just do not make crank trigger with odd number of teeth.

Ray.

[rayhall]

Using missing tooth trigger wheel on a one cylinder engine is not a good idea. The problem is finding the missing tooth while cranking. The engine accelerates and decelerates very wildly and it is very hard for the ECU to measure the time distances between teeth so it can locate the missing one.

We recommend you use 24 teeth with none missing and a cam signal.

Ray.

[Guest |829|]

This is the reluctor wheel on the stock motor that kawasaki uses in both freestyle bikes and quads. We are updating the firmware right now and will redo the trigger calibration with the new features.

Do you know how the stock kawasaki ECU operates around the lack of cam position trigger?

Changing the crank sensor and adding the camshaft trigger will be alot of machining but if it is needed we can get it done.

edit: we're also having some trouble with the firmware update, we're running the Vipec through a COM (serial) cable and the update fails, it says to close the program and reload the monitor. After restarting the program the Vipec is in boot mode and running the firmware update again results in the same situation. Is the monitor something that needs to be reinstalled? I'll start checking the rest of the forum.

Thanks again

[rayhall]

In the USA there is a trend to use the wrong description for missing tooth trigger wheels. Good example is engine with 60-2 trigger. People call then 58 tooth trigger wheels. Not only is this wrong it just causes problems like you are having.

The world wide (except the USA) method for naming these patterns is the number of teeth including the missing teeth, minus the missing teeth.

Your engine will have 23 physical teeth plus one missing. This means there is 24 teeth NOT 23, and this is what you put in the software (24). Then add the missing tooth count which is 1.

To do the firmware update the PC must assign the Vipec USB driver to less then COM9. You can manually change this in the device manager.

[Guest |829|]

When we dug out the vipec from the backroom left from older teams we could not find the USB cable so we have been using the com port on an older computer. I cannot find the vipec in Device manager however it is using COM1 on serial to do the connecting, i saw in other posts that it is recommended to used usb for updates, but can we make it work without? Is there a driver that i need to install that i have not? and will a Serial to USB adapter accomplish the same thing?

Thanks for all your help so far, it is very much appreciated/

[rayhall]

Serial port is okay.

If using a serial cable, then there will be no Vipec USB driver listed in the COM Ports. You will just have a serial port visible. The number of this serial port needs to be less then COM9.

[Guest |829|]

The com port is regesterring as 2 but we still have this error. It says the update failed and the ECU is in boot mode.

I will email you the error message

[Guest |829|]

Status Update:

this is Jon, i'm working with mike on this

Firmware was properly loaded using an x64 computer with a Serial to USB adapter (in serial mode) no idea why that worked, but it did work, first try.

Hooked up the oscilloscope to the Crank Sensor, we got a nice clean reading, peaks from 5-7 volts depending on cranking speed, very distinct low voltage at missing teeth.

Hooked up the Crank Sensor to the Vi-PEC and inputed values of 2-4 volts as the triggering threshold, Raising the values resulted in a bit better RPM signal, but trigger error start becoming prolific. Several times we had to cycle the Vi-PEC to clear errors (max 255.) We also tried all filtering modes with no effect.

Depending on the setup the revs jump from 1400-7000-12000 all over the place, obviously the engine cannot be run in this condition due to the timing and fuel changes that would be inappropriate (backfires, exhaust pops)

So, the question is what are we doing wrong that is causing the Vi-PEC to either get a poor signal, or misinterpret the signal. As far as i can tell the signal is quite good and should be more than enough for the vipec to work with, the replacement peak in the KAWI manual is around 2v.

So:

Is the trigger shielding supposed to be grounded? nope

Are the sensor grounds supposed to go to the chassis? Engine? Somewhere else entirely? nope

Should we try using trigger channel 2 instead?

Can we turn of the trigger that is unused, I can't seem to do it in Vi-PEC, but i see no reason for it to be on.

Does the VI-PEC used the intake pulse reading on the MAP sensor in order to determine which half of the stroke it is on? This would make sense, but i'm not sure if that's how it works, i imagine that's how Kawasaki does it.

Once the Trigger issues are sorted i think we have a good shot, but I don't know where we're going wrong on the trigger. One minute you think it's good, then it randomly jumps to 7000

Thanks!

Jon

edit: this will work won't it? The vipec works on other configurations with 24-2 (GSX-r) and the kawi works with it, i just want to know if we're beating our heads against a wall...

Edit: could we use an auxiliary MAP sensor as the cam pulse signal? I don't know if the pulse would be enough to sync up the system, and it wouldn't likely work at high revs, but if it could establish position at idle then it could keep track of the cam position in order to fire the plug and injector properly.

edit: i setup a voltage divider for the crank sensor, i'm now getting a 3.9v peak in insead of ~10v, which the ViPEC seems much happier about, it's still real crappy and even messing with thresholds and triggering extensively we can't get it to lock on to the RPM at all. I don't know what else to change, or if i should pull the case to confirm the trigger tooth count

[rayhall]

Jon,

Before you start doubting the Vipec you need to take a long look at what you are and are not doing. I spent most of yesterday answering posts on this forum and many emails. I was still sending emails from home last night. All the problems so far are things you have done wrong. Map had many incorrect settings, crank sensor wired around the wrong way, and more.

I had many questions asking how does the engine know where to fire the ignition. After a while it became obvious that you were not reading the manual in the software on this and other subjects. So before going to bed last night I made you a PDF file of the page in the manual on setting the offset to get the timing correct.

If you are willing to be more patient, then send me the map as it is now. I am also going to need a scope trace showing the signals as you start to crank and then one while cranking.

Ray.

[Guest |829|]

Ray,

Sorry if we seem impatient, we're learning as fast as we can and there is much to learn.

I will make new traces for you later today, but my main issue is the trigger sensor. I cannot get it to read nicely while cranking (fuel off), it jumps all over and with or without a good map, offset, etc it isn't going to be able to run.

We haven't tried to adjust the fuel and ignition maps yet, and the sensor to the crank wasn't hooked up wrong, but the scope was hooked up backwards when we took the picture.

Our primary concern now is that we can't get a solid, or even close to solid reading for RPM, if this is due to the trigger offset not being setup then i apologize, i was under the impression that it was for setting the spark. We've been scoping and adjusting the crank sensor trigger threshold accordingly, but still it is not coming up with consistent results, if I am missing something important here that would be a big issue. I've been through the trigger setup guide and we've been through the first run setup a couple times, but to no avail.

Based on the scope readings it cranks just over 500rpm. If i will never get a good reading at this rpm then that would be a good thing to know so i can stop trying and move on to other things.

[Guest |829|]

Here are some screenshots from the Oscilloscope we are using to analyse the triggers output

Shot from the very beginning of cranking off the starter motor

Another shot at the very beginning of cranking (included this one to further show the increase of peak size)

Shot during cranking after the initial shot

[rayhall]

Use these settings.

Trigger 1 Level = Level 1 (not Level 2 that you have now)

Thresholds

1.0

2.0

3.0

4.0

5.0

5.0

5.0

[Guest |829|]

On this motor there is an inspection port in the crankcase. It is there to line up the piston to TDC going off a flywheel mark. We were trying to use that port to check the timing in accordance to TDC for the base timing. The problem has been it is open to the oil sump meaning we cant see the mark due to oil spraying in our face

So we decided to try use the Oscilliscope to superimpose the ignition against the crank triggers output.

When we did this once we saw that the spark lined up with the missing tooth.

Then we did an offset of 15 degrees to make sure it changed

But then we did a change to 180 and we seemed to get the same output as 0 offset

Ray first of all do you know why we are getting outputs like that?

Secondly is this an accurate enough of a method to test for timing?

Thank you

[rayhall]

When you used 15 deg you can see the timing move. Changing from 180 to -180 would show the timing mark in the same place, but would be on the wrong engine cycle

Can you put a peice of Plexiglas over the hole so you do not get covered in oil. Timing light is the best method.

If the missing teeth are at TDC compression then you could set the timing that way. To do this put the scope pickup on the ignition output wire so you get a square wave signal. The ignition signal will rise then fall. Line up the falling edge with the missing tooth area.

[Guest |829|]

With the spark plug out we now superimposed from the output of the coil. We got the square wave and with an offset at 0 sitting right in the middle. We assume this means we are good to go with a spark advance at 8 degrees. Ray is this good? Can you take another look at our maps we made changes to the fuel map set our ignition to 8 degrees all over.

Edit: We setup the IAT into the air intake.

[rayhall]

This is ONLY correct if the missing teeth are at TDC. Is this the case ?..if not then it is no good and you need to use a timing light.

Ray.

[Guest |829|]

This is the case.

So try fire it up adjusting the fuel main each time till it actually starts?

We have space in the valvetrain assembly to attach a CAM sensor and does not seem like it will be hard to accomplish

[rayhall]

Yes.