Kawasaki Z125 Forum banner
Cancel
Create thread New Forums

EFIE Experiences

Jump to Latest
21 - 40 of 64 Posts
#22 · (Edited)
Now thats a top notch report :wink2:
.
good to know I'm probably not running lean with a aftermarket pipe and slightly open intake :grin2:
so I'm staying as is for now ,

I thought the blubber i stated at 4000rpm was a rich blubber, but looks like its a lean starve ,
Then again My butt's lied to me before . It's hard to tell on a 8hp bike ?

looks like i wont need the expensive fuel controllers and wiring, maps, laptop programming that most dont seem to get quite right anyways.... so i know i never would.

I wont have the benefits of the compensating that come with the 02 but its running great and I dont plan on any other mod's
 
Save
Reply Quote Like Like
#23 ·
12:1 is actually too rich for max power unless you have like 15% alcohol fuel, might have to run the common 10% alcohol stuff it this bike. That said, you might have been reading rich because the sensor might have been too far into the pipe. My Innovate MT-X wants a 1 inch bung for small tubes like we have, this is what I welded to my KAYO and it worked well until the sensor failed.

I just jumped the gap by feeding more wire into it when I did the KAYO. It's ugly, but I'm a hack so I don't care, was done with a cheap fluxcore welder.
 
Discussion starter · #24 ·
Greg- I checked this out this morning on my way to breakfast. I had prepped (2) 1/8" thick washers in case I needed to shim the sensor. I rode to breakfast with 1/8" total shim, then rode home with 1/4" total shim. AFR was unchanged. So I think the bike really does run 12:1 at WOT. Maybe not best for power... but I'd wager Kawi set it up this way to keep temps cooler. Just my guess.
 
#25 ·
You should check the amount of protrusion into the exhaust, the sensor accounts for a pretty big disturbance in these little pipes. The least amount of protrusion that still works is the best way to go.

What fuel are you using? We might want to tune using different amounts of alcohol to get that max power right where we want it. Maybe a little Seafoam or Lucas cleaner added to thin it out. I may need to weld a second bung on my exhaust and do some looking. Start with 90 octane alcohol free and work backwards until I get the power where I want in open loop full throttle and let the computer worry about mid throttle. Assuming I get time for racing next season, want all this stuff worked ahead of time.
 
Discussion starter · #26 ·
Need some help boys.

Got EFIE wired up today. Sorta.

There are 4 wires to O2. (2) white which I believe are HEATER circuit, a black which is positive signal (to ECU) and a grey which I THINK is just a ground. But it is a ground that goes back to the ECU.

I intercepted the Black (positive signal wire) with EFIE and set it up so voltage that passed through it was reduce by 0.250 volts. Used multimeter to test I verified that it worked on the bench and once in the bike. I tested the positive signal wire on both sides of the device and all was good.

Fired up the bike and it threw a code. I thought maybe the modified signal was out of range. AFR was whacky. Started high almost 15 then dropped to 12-13. I adjusted the device to drop signal by 0.150 volts and went for a ride. AFRs were mixed. Giving it more throttle the AFRs seemed lower. with the throttle cracked they were 14.2-15. This behavior was similar to running without the O2. Coming off throttle idle would end up around 13 afr then creep up to 14.6ish after maybe 15 seconds. Almost like the system was fighting with itself. Maybe idle is open loop and previously learned trims were coming into play.

After the ride I pulled the code - #67 for O2 heater element. Strange. I checked resistance through the 2 white wires (running to the O2) and got 18 ohms which per manual is in spec. I checked for voltage feeding the heater circuit at the O2 harness and had battery voltage. So I'm not sure why I'm getting heater circuit code. The manual says next step is check continuity to the ECU and if that is good then get a new ecu lol. I think it must just be a fluke code because I didn't mess with ECU or harness wiring. I tapped into O2 harness only.

I was reading 1.2 - 1.35 volts on the positive O2 signal wire before modification, this was giving me AFR readings around 14.2. This is my first time playing with narrow band O2s so I don't know what kind of voltage to expect. I'm measuring positive signal to ground. This may be an important detail. The manual, when testing the O2 positive signal wire voltage, has you connect to the grey wire (the mystery maybe a ground that definitely goes back to ECU) and not just any ground. Not sure if this matters. For more understanding of the ramblings in this paragraph crack open that manual to section 3-62. It is for diagnosing O2 sensor performance. It says connect positive lead of voltmeter to BLACK (check) and negative lead to GREY (I used chassis ground). It then goes on:

(Rich) DC 0.8 V or more
(Lean) DC 0.24 V or less

I'm getting 1.3 V at 14:1... so wtf?

Maybe the negative side actually has some positive voltage that would lower my reading when measuring across the black and grey wires. In any event a 0.250 V drop considering the range of 0.8 - 0.24 may be too much.

Thoughts?

Tonight I disconnected the ECU hoping overnight all fuel trims will be forgotten. Not sure if I should zero out the EFIE and ride it until the CEL clears or just ride for a while with the 0.150 offset.
 
#27 · (Edited)
MrEvilpirate said:
Need some help boys.

Got EFIE wired up today. Sorta.

There are 4 wires to O2. (2) white which I believe are HEATER circuit, a black which is positive signal (to ECU) and a grey which I THINK is just a ground. But it is a ground that goes back to the ECU.

I intercepted the Black (positive signal wire) with EFIE and set it up so voltage that passed through it was reduce by 0.250 volts. Used multimeter to test I verified that it worked on the bench and once in the bike. I tested the positive signal wire on both sides of the device and all was good.

Fired up the bike and it threw a code. I thought maybe the modified signal was out of range. AFR was whacky. Started high almost 15 then dropped to 12-13. I adjusted the device to drop signal by 0.150 volts and went for a ride. AFRs were mixed. Giving it more throttle the AFRs seemed lower. with the throttle cracked they were 14.2-15. This behavior was similar to running without the O2. Coming off throttle idle would end up around 13 afr then creep up to 14.6ish after maybe 15 seconds. Almost like the system was fighting with itself. Maybe idle is open loop and previously learned trims were coming into play.

After the ride I pulled the code - #67 for O2 heater element. Strange. I checked resistance through the 2 white wires (running to the O2) and got 18 ohms which per manual is in spec. I checked for voltage feeding the heater circuit at the O2 harness and had battery voltage. So I'm not sure why I'm getting heater circuit code. The manual says next step is check continuity to the ECU and if that is good then get a new ecu lol. I think it must just be a fluke code because I didn't mess with ECU or harness wiring. I tapped into O2 harness only.

I was reading 1.2 - 1.35 volts on the positive O2 signal wire before modification, this was giving me AFR readings around 14.2. This is my first time playing with narrow band O2s so I don't know what kind of voltage to expect. I'm measuring positive signal to ground. This may be an important detail. The manual, when testing the O2 positive signal wire voltage, has you connect to the grey wire (the mystery maybe a ground that definitely goes back to ECU) and not just any ground. Not sure if this matters. For more understanding of the ramblings in this paragraph crack open that manual to section 3-62. It is for diagnosing O2 sensor performance. It says connect positive lead of voltmeter to BLACK (check) and negative lead to GREY (I used chassis ground). It then goes on:

(Rich) DC 0.8 V or more
(Lean) DC 0.24 V or less

I'm getting 1.3 V at 14:1... so wtf?

Maybe the negative side actually has some positive voltage that would lower my reading when measuring across the black and grey wires. In any event a 0.250 V drop considering the range of 0.8 - 0.24 may be too much.

Thoughts?

Tonight I disconnected the ECU hoping overnight all fuel trims will be forgotten. Not sure if I should zero out the EFIE and ride it until the CEL clears or just ride for a while with the 0.150 offset.
Click to expand...
.
Questions are just really trying to learn here more than I'm helping you I'm sure , But maybe something in one of my questions might trigger something for you ?
Dont answer anything that is waaaay of base .. and that answer my question ;)


[[and a grey which I THINK is just a ground. But it is a ground that goes back to the ECU.]
[[ (the mystery maybe a ground that definitely goes back to ECU) and not just any ground.]
[[has you connect to the grey wire (the mystery maybe a ground that definitely goes back to ECU) and not just any ground. Not sure if this matters. For more understanding of the ramblings in this paragraph crack open that manual to section 3-62. It is for diagnosing O2 sensor performance. It says connect positive lead of voltmeter to BLACK (check) and negative lead to GREY (I used chassis ground). It then goes on:]]
.
The gray does go to the ECU, (3-63)(page 135)
.
.
[[with EFIE and set it up so voltage that passed through it was reduce by 0.250 volts. Used multimeter to test]]
.
Reduced? you mean you set it to 250 mv like the grom guys did?
.
.
[[I thought maybe the modified signal was out of range]]
.
from grom guys ... If we look at a narrow band o2 sensor voltage chart you can see a ~250mv offset will make the switching poing around an afr of ~ 13.3-13.5:1. You can not go much above 350mv decrease as you get to the point of the ecu always seeing 0 volts and this will throw a CEL and not be able to properly lean out and tune to how it wants.
.
.
[[Fired up the bike and it threw a code.]] Above statement (from grom guys)
.
Is it because you removed the 02 to test on the bench ?
have we ever conformed when CEL comes on after disturbing the 02' that it is for 02 #33 code or for the #67 heater code
.
.
[[Coming off throttle idle would end up around 13 afr then creep up to 14.6ish after maybe 15 seconds. Almost like the system was fighting with itself. Maybe idle is open loop and previously learned trims were coming into play.]]
.
Like you were thinking WOT would do .. but why would it creep
Did you get a WOT reading.
.
.
[[I'm getting 1.3 V at 14:1... so wtf?]]
.
because the Gray wire is not going to the ECU
.
.
did you take the EFIE gray wire to the 02
and the EFIE green wire to the ECU
Like the grom guys did ?
.
YOU GOT THIS :wink2:
 
Save
Reply Quote Like Like
Discussion starter · #28 ·
Kaw'ee- All comments advice welcome. Thanks for the reply. I took pics of the install I'll throw em up later, but for your questions:

Yes, Grey to ECU, for sure 100%.

Yes, Reduced by 250mv, just like Grom guys, that's why I started with 250.

Thing about the code is I figured it would throw the bad O2 code, not the bad heater circuit. Unless maybe it threw both if I had key to ON position with no O2 present. I grounded the diag port and read the flashes. Is there a procedure to see if there is more than one stored code? I held it grounded and the flashing kept cycling through the 1 code, #67 bad O2 heater circuit.

I did get WOT. It was 12, just like before. Mostly. Saw 11.8, 11.9 as well for the first time. I believe creep may be to stored long term trims???

Your text below:
------------------------------------------------
[[I'm getting 1.3 V at 14:1... so wtf?]]
.
because the Gray wire is not going to the ECU
-------------------------------------------------------
Not sure what this means Grey and Black both go to ECU. Some other forums say sometimes talk about a bias voltage to the O2. This is basically arbitrary voltage the ECU runs through O2 signal circuit. O2 only generates voltage when it gets warm enough. ECU monitors the voltage whenever it is on. So in the beginning O2 is cold, and ECU only sees the bias voltage it is sending. Once O2 warms up the voltage it generates gets added to the bias voltage... the ECU sees the change and uses the difference between bias and read voltage to extrapolate the 0.1 - 0.9 V narrow band swing. This is I think why I am getting voltages over 1.2V on the narrow band signal wire. I will check voltage at the O2 harness connection to main harness without O2 sensor installed to try to confirm there is bias voltage coming into play. This could be why the manual wants you to test signal voltage across the gray and black wires, not just ground and black, so you automatically cancel out the bias voltage. Also I need to do full continuity tests from O2 harness through ECU connection just to make sure... I've had loose wires on this bike before I had to chase down.

Yes, Cut the BLACK signal wire to O2...White EFIE wire to O2 side, Green EFIE wire to ECU side.

Appreciate the vote of confidence. Anyone else have ideas? Sucks I need to go to work and cannot continue to tinker with this sucker.
 
#29 ·
Keep in mind that narrow band o2 is not really an analog signal. They basically switch back and forth between rich and lean, the system never really finds a place and holds. The switching tells the system when things are good, and the system is constantly making adjustments back and forth to get that swing for a check.

I'm wondering if you might want to use the analog output from your wideband to simulate the effect you want, night at least get some understanding of what Kawasaki is doing in the ECU, the EFIE may not work because of this. Kind of an expensive work around to suggest that everyone runs a wideband meter and simulates a narrow band on the output, but it might help us gain some understanding of the stock ECU.

Take a look at this article https://www.a-i-e.com.au/tech_stuff/view/6974/narrow_band_oxygen_sensors
 
#30 ·
Nice write up.

You had me on the edge of my seat thinking you were going to bite the bullet and weld-er-up! I'm diggin that print in your garage directly in front of your Z in the first shot.

Bummer that our open/closed ECU communication nonsense isn't as straightforward as the Grom guys. I read through that thread you linked and almost bought an EFIE before I remembered "this isn't our bike... this isn't actually relevant."

What would it take to augment your wideband to log AFRs like the grom thread with the bazzaz? That seems like it would be a very useful tool for this process. Start a go-fund-me to buy the crap and I'll support you. XD
 
Discussion starter · #31 ·
OK gents minor update:

Greg- Thanks for link. I actually read that same article last night. Lots of good info in there.

Ando - Some type of standalone hardware to log to honestly I'm not sure what it would take... and I would only have AFRs, not throttle position or RPM so not sure how useful it would be.

So far today before work: I confirmed continuity from all (4) O2 sensor wires back to end of harness right before it plugs into ECU. I also confirmed that the ECU is sending voltage to O2 sensor (bias voltage) which is why I had signal on the voltmeter over 0.9 volts and why Kawi tells you to pull voltage across the pos and neg (black and grey) wires, and not just black to universal ground. When in doubt follow the directions I guess.

Today after work: I had ECU unplugged all day. I left EFIE connected but zeroed the offset (Hopefully this is actually zeroing and not the "I washed my bike today and it feels like I gained 15% HP). Fired the bike up and went for a ride. AFRs appear to be normal, not the crazy swings I had first time out with EFIE. I believe this process DID clear out the fuel trims. AFRs at idle, WOT, and above 8kish adjusted during the ride. Going to let it cool off then go out again. I really want that CEL to turn off, then I will clear fuel trims, set EFIE up for a conservative 0.100 V, and take a ride to check AFRs.

At this point I kind of hope that open loop is not effected by closed loop trims (although my guess from the first ride here is that they are). The bike actually runs pig rich at WOT. Beginning of my ride today I was seeing 11.5! It was getting closer to 12.0 as the short term trims were getting learned. Fear is that if the map gets shifted with the EFIE the already rich open loop numbers will get even richer.

The plus side, however, is that if the injectors are capable of pumping enough fuel to generate an 11.5:1 AFR then it's looking good that the stock ECU can handle intake and exhaust... but don't quote me on that yet, that'll be Phase III.

Also, the O2 heater circuit is energized when the key is on... If ECU is monitoring the heater circuit when key is on but bike is off, that could be what caused it to throw the code.
 
#33 ·
The only thing I can think of as to why this isn't working might be where they talk about the Titanium sensors giving a resistance. Is it possible this is the type of sensor that we have in the Z?

I will still probably weld up a bung for my AFR meter, but not likely this season since the season is basically over. 50 degrees and rain all day, barely got above 60 the last 2 days... I'm afraid Winter is here and will stay through to about May. And the AFR meter might depend on what I do with my KAYO, if there is room, that gauge will be going in there, it's carbed so it would make jetting much nicer than simply relying on temperature estimates for jet sizes. Need to make a new display area once I get my shift light working, shift light has a tacho in it so I can remove the big in-accurate tach and put in some aluminum. Have to see how things go. Need to make room for a battery in the KAYO as well. Also need some switches to turn on the power from battery, and turn on the cameras for recording.

For those that have no idea what I'm talking about, here are the two posts on the shift light:
Shift light and tachometer - part1
Shift light part 2
 
#34 · (Edited)
MrEvilpirate said:
...and I would only have AFRs, not throttle position or RPM so not sure how useful it would be.
Click to expand...
Oh right, derp.

Sounds like you are making good progress. The ultra rich WOT is kinda odd, I like Honda's system of extrapolating the closed trims into the open territory.

I wish the tuning world weren't so hush hush, it would be cool to hear from some of those guys on what they found in the Kawi ECU that surprised them, what they wanted to fix, etc.

Because, while we appreciate the work you are doing, it's been done before in parallel by multiple companies. So, we could keep the actual tuning secretive but share knowledge of the OEM configuration. In theory that would lower everyone's development cost, but who's going to be motivated to put in the initial work? Open source stuff is so nice... in theory. Just like communism.
 
Discussion starter · #35 ·
Greg - I thought about the different type of plug. No, we don't have titanium... Kawi gives the testing spec in voltage not ohms. Also, after just coming home from 2nd ride, I stripped some of the gray wire and measured voltage across grey to black with the bike running... saw oscillation between 0.1 and 0.9 V, so that one's put to bed... bias voltage confirmed! Bias is about 0.5 V, so the 0.250V correction with the EFIE shouldn't cause an out of range code. And yeah, wideband makes jetting way easier with carbs... especially the idle mixture screws. That's actually why I have this thing!

Shift light project looks really interesting!

Ando - LOL at communism being nice... in theory.

2nd run after "reset" in effort to clear CEL. AFRs are getting much closer now to my first measurement. The power off either clears the trims for sure or this is an insane case of placebo effect.

I'm getting a much better idea of how the ECU adjusts AFR based on throttle position and RPM combo... honestly it really isn't too bad. My only major gripe, after determining that there is basically zero risk of running lean (stock) at WOT is the lean surges you get occasionally around 5-6k RPM with throttle just cracked. More to come.
 
Discussion starter · #37 ·
OK boys we have proof of concept.

CEL was out today so I dialed up EFIE to a 0.125 V offset and went for a ride. This was without clearing any trims.

I noticed that cracked throttle cruising AFR dropped and I was seeing 13.7 and 13.8 for the first time. The AFRs swing around, stock is like 14.4 +/- 0.4. With this offset I got to maybe a 14.2 +/- 0.5. There are couple things coming into play here.

First, trims were not reset and the ECU most likely needs more time to re-learn. I only went for maybe a 6 mile ride.

Second, as narrow band O2s get farther away from an AFR of 14.7 accuracy drops of QUICKLY and a smaller change in voltage can correspond to a larger change in AFR. This doesn't worry me too much... AFR on carbs swings much wider than what I was seeing on my test ride today. The average AFR drops, but the "swing" could potentially grow with an aggressive offset.

So I went back to the garage, and with the bike still running, turned up the voltage offset while I watched the idle AFR drop. I got it down to around 13.3 and went for a ride. I needed an offset of 0.360 V to get there. Promptly threw code 33: bad O2 sensor. So I took it a bit too far. The offset must have sent voltage from the O2 out of range.

Rode home dialed the offset back to 0.250 and disconnected the battery. I'll take it out again tomorrow with no trims stored in ECU and see how it runs.

I will say this... even with just the 0.125 V offset... I was hunting around for lean surge on cracked throttle between 5-7 krpm and couldn't find it, so I'm optimistic. Still need to see how open loop is affected.

I have a zoom loop and a K&N filter for DIY intake coming, hoping to have both here by end of day Friday, and weather is looking good this weekend. So if the parts arrive on time the goal by Sunday night is to post up confirmation that stock ECU can handle exhaust and intake, and maybe even some tuning info on the EFIE.

One thing I think that may come into play, is that as we tune EFIE and get away from 14.7 AFR, because of the loss of accuracy, what may work on one bike may not work 100% on another, even with identical mods, due to accuracy tolerances of narrow band O2s. That's pure speculation on my part, but from what I read, and since we're pretty much asking the narrow band to shift lamda and almost act as a wideband, I think some variability between different O2 sensors can be expected.

Side note- apparently having the bike on (not running) with O2 disconnected will throw the O2 heater circuit code. Heater circuit is energized when key is turned on, and it is grounded through the ECU. I read that in some vehicles the ECU can "pulse" the ground to turn on or off the heater circuit. The code gets thrown if resistance (ohms) is out of spec, or if there is a short in the circuit, or if the circuit is open, which is the case when the O2 is disconnected.
 
#40 ·
Appreciate your write-ups. You must be an engineer when you aren't hooning around the coast on mini bikes. This is sounding more and more promising.

OT: If it weren't a heavy/awkward PITA to ship I'd like to try your Kaw'ee mod'd exhaust. I've been trying to think of ways to make a non-permanent hole (not technically my bike) in the left side of chamber #1.
 
21 - 40 of 64 Posts
About this Discussion
63 Replies
10 Participants
Last post:
MrEvilpirate
Kawasaki Z125 Forum
posts
32K
members
5.6K
Since
2015
A forum community dedicated to Kawasaki Z125 motorcycle owners and enthusiasts. Come join the discussion about performance, modifications, classifieds, riding gear, troubleshooting, maintenance, and more!
Full Forum Listing
Explore Our Forums
Kawasaki Z125 General Discussions Engines Discussions Appearance & Modifications Electronics Z125 Parts & Accessories