Jump to content

starting the bike


gtecomp

Recommended Posts

talking about sequential injection or banked/group I did an experiment on my stock 2007 1250 ECU, it does not run a cam sensor.

In the factory manual it lists the possible fault codes and it has listings for each individual coil and injector. In the failsafe section it mentions that if you lose 1 coil it will cut the fuel to that affected coil and continue to run on the other 3 cylinders. 

 

IMG_9730.thumb.JPG.3bf017ef7ab580f541607f9f0c22da87.JPG

IMG_9731.thumb.JPG.d728e8c764f3c02435c07156a973750f.JPG

IMG_9732.thumb.JPG.b6cd8ad737fb32c18b968b8187d70d33.JPG

 

This tells me that the ECU knows each individual injector and coil and if its functioning.

I used a healtech to look at live data and it to showed different injector pulses for individual injectors.

 

1.thumb.jpg.1b33fcf9dbd98101c94d0c10db3b4df5.jpg

 

So finally today I worked out how to 'see' the injectors running sequentially and see it cut fuel to the affected injector when a coil fails.

 

IMG_9728.thumb.JPG.7bdca68443101053edc005f4ff59985d.JPG

 

I placed a LED light inline of injectors #1 and #4 (companion cylinders) to see the injector stop when the coil was disconnected.

 

 

This video shows a non cam sensor EFI running and detecting and disabling #1 injector when I disconnect #1 coil,  proving that each coil/injector have individual drivers.

Here is the ECU input/output diagram.

 

002.thumb.jpg.241f2f322fd45eafe1f8674cbffce1c6.jpg

I  had a thought/guess as to how the ECU knows what is #1 cylinder without an input from a cam sensor.

It knows on initial start up and first running where #1 and #4 by default by the trigger wheel so I imagine it would be at first running batch style injection to get it running until it sees a constant rpm and it could then calculate from the trigger wheel position (and now we know it has individual drivers for the injectors from the video) it could use the #1 injector driver as its position input and fire the #1 injector driver only on the next rotation timing position for #1-#4 and thus starts the sequential sequencing as it knows the firing order and is counting trigger wheel position for the other cylinders.

 

IMG_9729.thumb.JPG.442dfa987591fec8fd419aa7393619d9.JPG

Arttu my trigger wheel is timed when at #1 TDC the pickup is positioned between tooth 7 and 8 where the red mark is above, the arrow on the trigger wheel is also correct for engine rotation. So #1 TDC the sensor is between tooth 7 and 8 after the -2 gap.

Phew.

 

Edited by GSX1100dreamn
added info
Link to comment
11 hours ago, gtecomp said:

Yep...I'll write an e-mail to them and asked for more detailed explanation....but here is my new bunch of questions: 

1. cam cover - is weldable? ... any magnesium strange alloy to be expected? I still consider the option to put hall sensor over some cam lobe.

2. dead time/voltage correction table of stock say what now!? 600 injectors. Before 7 years I put some weird values and I don't remember from where I get them.

3. some ignition table and VE table that work?   

1. Look in my build thread "half thousand horses" there are pictures of my instalation on cam sensor without welding.

 

3. There are a basemap for gsxr/bandit to download on maxxecu website 

Edited by Fredrik_Steen
Link to comment
5 hours ago, GSX1100dreamn said:

Arttu my trigger wheel is timed when at #1 TDC the pickup is positioned between tooth 7 and 8 where the red mark is above, the arrow on the trigger wheel is also correct for engine rotation. So #1 TDC the sensor is between tooth 7 and 8 after the -2 gap.

Phew.

If I followed this correctly this means #1 tooth at 100° BTDC. Which makes sense since this seems to be common choice on Suzuki EFI engines.

Regarding sequential operation without cam sensor. First, having separate output channels for each injector and coil doesn't necessarily mean that the ECU runs sequentially. Coil and injector faults are usually detected by current so the ECU doesn't know if the cylinder is actually firing. However, there are a couple of ways to figure out the engine cycle without cam sensor. One is to use MAP sensor to detect intake stroke. I know this is used by some OEM bike systems and some aftermarket ECUs are supporting this too. This usually requires connecting the MAP sensor only to one cylinder which may result other problems with MAP signal stability. And more or less tweaking might be needed to make it work properly. Another method is looking crankshaft acceleration and detecting power stroke from there. This is used by many OEM car ECUs for misfire detection but I don't know if it's used for sequential synchronization anywhere.

Link to comment
12 hours ago, gtecomp said:

2. dead time/voltage correction table of stock say what now!? 600 injectors. Before 7 years I put some weird values and I don't remember from where I get them.

I'm using 0.9ms dead time at 13V and approximately 0.15ms/V voltage correction for them. These aren't 100% correctly measured values but close enough to make it work just fine.

I have several ignition tables for 1100 oil cooled turbos and maybe even some for a 750. But if there is some example table on MaxxECU site it's probably as good starting point. There is always some variation between the engines so in the end you have to tune your own in dyno. Copying the VE tables is even less useful. In addition of differences between the engines there are also numerous variables in the ECU setups, sensor calibrations etc. that affect on the VE table. So based on my experience it's always pretty much full re-tune for every engine.

Link to comment
9 hours ago, Fredrik_Steen said:

1. Look in my build thread "half thousand horses" there are pictures of my instalation on cam sensor without welding.

 

3. There are a basemap for gsxr/bandit to download on maxxecu website 

Thanks for advice. I've checked your thread twice, but....can you show which page was installation shown?

Link to comment
1 hour ago, gtecomp said:

Thanks for advice. I've checked your thread twice, but....can you show which page was installation shown?

Found a really old picture in my Phone :)

 

I use what we call "Tredobricka" in Swedish. A Steel/rubber washer from the hydraulic world on both sides on the cover.

Grind some fin's of the cover to get a nice sealing surface.

A standard cherry red sensor, 1mm from the exhaust lobe. Don't forget to account for the sealing between the cover and head when the cover are torqued down. 

received_533999823955583.jpeg

  • Like 1
Link to comment

hm

...there is not much settings in MAP as sync function. I connect just 1# cyl to MAP sensor and is true, needle jumping like crazy at idle. I put the settings and ECU even don't read rpms in this mode. I write down an e-mail to MaxxECU support with question , what is the logic behind this function. For now is dead end.

max ecu.jpg

Link to comment

yep I log the MAP readings on cranking and 65kpa is lowest reading. I put 70kpa , but no rpm readings.

 

I suspect I should connect every coil to own ignition channel (which is problem, cos I have 3chanel VAG coil driver and I'm 1 channel short :) )

Edited by gtecomp
Link to comment
  • 2 months later...
14 hours ago, gtecomp said:

unstable idle. Strange that I receive 1ms effective injection time on idle. I'm with stock gsxr600 Keihin injectors which should be around 240cc at 3 bar??? It act like i have 600cc injectors :)

You have an idle injection pulse of 1ms? That is pretty short - i'd doubt most injectors can't open and close reliably at 1ms - maybe? How about dialling back the fuel pressure to 2.5 or even 2 bar and see what affect that has? Reading through the thread again I see you are using a 044 pump - nothing wrong with them as regards pressure but their amp draw probably equals what is drawn by everything else - maybe the two are connected?

Link to comment
21 hours ago, gtecomp said:

hm...new problem

 

unstable idle. Strange that I receive 1ms effective injection time on idle. I'm with stock gsxr600 Keihin injectors which should be around 240cc at 3 bar??? It act like i have 600cc injectors :)

 

any idea what I'm doing wrong?

Is that effective injection time (without dead time) or total? For total that's way too short, like @Gixer1460said the injectors wouldn't open almost at all. If it's effective injection time it's closer to the reality but probably still on lean side with 240-280cc injectors.

Basic settings that affect on injection time are injector size and dead time. Ideally you should set these at correct values and not touch them after that. Third relating setting is injection interval, 360° or 720°. Smaller interval drops effective injection length to a half but if the dead time is set correctly it shouldn't affect on mixture. However, there is a risk that resulting injection time gets too small for accurate fuel metering. But with those injectors this shouldn't be a concern.

Once these basic settings are correct it's just a matter of tuning the fuel table correctly. You tune the table to get correct AFR and resulting injection time is what it happens to be.

Link to comment

yep there is a problem.

 

Curently I run stock GSXR600,750,1000,1300 I GEN injectors (no solid info of their flow)..group 720deg ...1 time per 2 rev.

I visit my thrusted tuner and he said me the same. We argue very and he even start his drag/street car to show me how 2liter engine with 3000cc injectors and 3ms injectors time at idle works.

So i in panic and do some changes (without success)

- switch to hall sensor at crank

- switch to 3 bar FRP (the old one was 4 bar)

-do 4ms dwel of coils at cranking speeds..

 

I i thnik I killed AGAIN all spark plugs.

 

Next step...remove injectors for cleaning and flow meter them.

Link to comment

Sorry if I'm stating obvious but injection pulse length is result of what you enter to fuel settings in the ECU. Fuel pressure, crank sensor or other "external" factors have no direct effect on it. During tuning process you adjust injection pulse length by the fuel table and other variables until you get proper AFR as result. So strictly speaking it doesn't matter what the pulse length is if the engine is running properly. But clearly "wrong" pulse length can be indication of some fundamental problem with your setup and especially if you have difficulties to get the engine running properly it's a good idea to figure out what is going on there.

1ms effective injection time isn't that bad, it should be still within operating window of the injectors. But indeed it sounds shorter than expected for those injectors. For comparison my GSX idles around 2ms effective on the same injectors. If the AFR is showing sensible values I think there might be something wrong with the injectors. Maybe some of them are leaking so the engine is getting more fuel than what is commanded. That could also result poor idling since some of the cylinders are getting too much fuel and others too little while the lambda sensor is showing ok mixture.

Link to comment

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

×
×
  • Create New...