Fuel injected fj12 looooooooong term project

[fj1289]

Just a thought!
Some of you have fitted injection, has anyone perfected a way of getting a good signal from a cam sensor?

The XJR uses a backbone frame, thus a cam sensor can fit....not so on our perimeter frame. The sensor housing runs afoul with the frame rail.
RPM used batch injection on their bike, as I think Chris also uses on his race bike.

Move the frame rail

.......and the fairing?

I imagine the frame rail could be notched? 

Pat, do you have a picture of the cam sensor on the XJR?   I didn't realize they used one.   I've thought about trying to mount a hall sensor to the camshaft cover to pickup a cam lobe.  Yes, I have been running batch fire with no issues.  Sequential would be nice to run large injectors on a street bike and still be able to idle well and handle a large dry nitrous shot or a lot of boost.  You can play with injector end angles and pic up a little smoothness and economy at lower RPMs - but it is a real fine-tuning exercise as I understand it from reading - haven't messed with it myself!

[Pat Conlon]

No Chris, sorry, no XJR pictures...

[fj1289]

Awww, come on Pat, you had me excited that there was a Yammi cam sensor.

Nope - at least not in the first year for EFI on the XJR - 2007.

Yamaha is using batch fire, or some variation of it.  The mega/microsquirt ECUs will do semi-sequential - you use 2 squirts per cycle - so you essentially get half your fuel "out of cycle" and half "in cycle".  It keeps all cylinders operating the same vice a single squirt batch system where one cylinder will get its fuel "in cycle" and its pair gets it " out of cycle".  As you get to higher engine speeds and longer injector pulse widths, it really doesn't matter since the injector is adding fuel for most of the 4 stroke cycle.  The Holley ECU will also do a novel version called "untimed sequential".  Using just a crank signal it runs in sequential mode - it's just 50/50 whether #1 is at TDC on the compression stroke or exhaust stroke.  This is what I am using now on the race bike.  Why?  Because I'm running 1,000cc injectors on a 1.3 liter engine - that keeps the injectors only injecting one time per cycle - important at idle because the injector pulse width can only get so short.  Running a semi-sequential routine injects twice per cylinder - limiting how big the injectors can be and still achieve acceptable idle quality. 



I did learn something new (and somewhat important) - Yamaha uses pretty high fuel pressure in their systems - 56-57 psi vice a "typical" 43 psi.  I had just assumed my "high" fuel pressure was with a sensor that needed calibrating or a fuel pressure regulator that had been crushed for higher pressure.   

Also, not really much info on the EFI system from Yamaha - - Suzuki goes into a lot more detail of how they run their system in their repair manuals (almost like the WANTED people to "hack" their ECUs...)

[ribbert]

Just a thought!
Some of you have fitted injection, has anyone perfected a way of getting a good signal from a cam sensor?

The XJR uses a backbone frame, thus a cam sensor can fit....not so on our perimeter frame. The sensor housing runs afoul with the frame rail.
RPM used batch injection on their bike, as I think Chris also uses on his race bike.

Move the frame rail

.......and the fairing?

I didn't think that much needed to be removed, but if it does, get creative, make a feature of it. If you're fitting FI to an FJ, might as well draw attention to it.

[Pat Conlon]

Thanks Chris, does the Holly ECU have a auto tune feature? That's a must in my book.
Between the two, micro squirt and Holley, which ecu would you suggest for the street?
Noel, if there is no cam angle sensor hanging off the side of the head, there is no frame problem.
No wonder I couldn't find any pictures....

However, I recall there was a picture of a cam angle sensor on a FJ head but it was an aftermarket install.

[fj1289]

Thanks Chris, does the Holly ECU have a auto tune feature? That's a must in my book.
Between the two, micro squirt and Holley, which ecu would you suggest for the street?
Noel, if there is no cam angle sensor hanging off the side of the head, there is no frame problem.
No wonder I couldn't find any pictures....

However, I recall there was a picture of a cam angle sensor on a FJ head but it was an aftermarket install.

Pat -most important in choosing an ECU is what you have a support base for - what your tuner is familiar with, friends are experienced with, or good online support.   Most stand alone ECUs are probably at least 80% common in functionality. 

Most ECUs will either autotune on the fly, or will "autocorrect" via datalogs.  BUT, autotune CANNOT be relied on blindly.  You still need to go in and smooth the tables, look for anomalies, etc.  Yes, the Holley does autotune on the fly - without laptop connected.

For a street bike I'd go with Microsquirt.  Plenty capable - it's biggest downfall is it's adaptability to do almost anything  - because it means there are so many options and settings to deal with to get going.   This is where having help is critical.  Also, the Microsquirt is small and well sealed.  It can run open or closed loop - or a combination depending on how you set it up.   Maintaining a reliable O2 sensor is very critical for tuning and closed loop.  That caused me a ton of headaches and road side "break downs" until I figured out what was going on and how to better deal with it.   

For the race bike, I really like some of the features of the Holley.   It is very well setup for nitrous - multiple stages, multiple point progressive curves, adjustable dry fuel curves, timing retard curves, etc.   AND it will run nitrous in closed loop with user specified AFRs for each stage.  The Holley also has its own built in wideband O2 - I really like this feature as I assume it means the O2 data is much better integrated into the ECU than an external O2 sending a signal to the ECU.  I assume things like the O2 delay are better accounted fo vice me putting in some random delay n a table.   Have had the bike up and running on the Holley - just not much tuning time before I tore down the engine to freshen it up which turned into a much bigger project....

The Holley has a pretty large footprint for a motorcycle - and I'm using the "small" HP.  The Dominator is almost twice the size!  (But would be nice for all the I/O!)  Another reason I went with the Holley is the work MPS Racing did with it for dry nitrous and the advanced tables they came up with to run the air shifter system.   Finally - I went through MPS for it so I could reach out for help if I was having issues with something down the road. 

When I get the street bike back on the road it will have either the Microsquirt or the Megasquirt MS3X I ran previously on the race bike (pretty sure I have figured out what was causing my reset errors at high RPMs - swapping for a hall-effect pickup for the crank sensor should clear that up).

[Pat Conlon]

Thanks Chris, I agree there is no substitute for dyno tuning. The auto tune is just to get you to the rollers.
I remember now what the hang up was on using the XRJ fuel injection on the FJ....it was the TPS on the XJR's throttle body that interfered with the FJ frame rail.
I think I recall that was the reason RPM adapted the narrower FJR throttle bodies for their FI system.
Chris, are you still using the FJR throttle bodies?

[fj1289]

Autotune really isn't necessary - carbs have never autotuned!

With a new EFI install first task is to get it to start and stay running long enough to warm up.  Then you adjust the idle mixture for best idle - just like with carbs.  Then you start free rev tuning above idle - I work up to about 3,000 rpm.  Once it will free rev a bit you can start easy, low throttle rides to the end of the parking lot/block/etc.  look at the logs, see what the AFR is doing, and adjust the part of the table where you were riding.   As you do this you can adjust the rest of the table linearly based on what you are doing in the lower parts of the table.   Then you just keep expanding the envelope. 

You do need to make sure you smooth the table as you go - fuel should increase smoothly with both RPM and load (throttle position or intake pressure depending on your tuning scheme).  When you spend more time in a particular area of the table, but little times in others - it can make the fuel table "lumpy" due to lack of consistent data and how the ECU interpolates between cells as it moves from one part of the table to another.  Auto tuning algorithms tend to exaggerate this tendency if the tuner doesn't smooth the table between learning periods. 

[derbybrit1]

[quote author=fj1289   

When I get the street bike back on the road it will have either the Microsquirt or the Megasquirt MS3X I ran previously on the race bike (pretty sure I have figured out what was causing my reset errors at high RPMs - swapping for a hall-effect pickup for the crank sensor should clear that up).

[/quote]

Yes, a hall sensor will get rid of noise induced resets.  But most crank sensors are VR.  I have a MS solution which fixed my Aprilia VR sensor noise problems.  A small signal conditioner board sold for the MS.  Completely solved the problem.  Let me know if you need that info.

[fj1289]

Marc - or it the JB Perf BR board?  Either way I would be interested in it and your experience with it.