Camel's K8

[REVIT13]

good work camel. thanks for the info as well Petespitstop.

still watching everyone else at the moment for now.

[GRUNTMAX]

Thanks Camel, this will help me a lot with the Stroker Motor Build when I finally get to that part.

[Camel]

Thanks mate, I've just written up the the text for fitting the LSU 4.2 wideband oxy sensor but can't upload photos from the location I'm currently at. I'll email the lot through to the administrators (Clients always make sure you've got email access!).

Cheers,

Camel

[Camel]

Who said there was a global economic crisis? I've never been busier! As a result I've had bugger all time to do anything but spend hours stuck in aircraft (all sizes) between various inhospitable portals to hell.

But I did manage to get my data logger hooked up and my wideband sensor in place. So it hasn't been a total loss.

But since when did I need to be told to lean when cornering?



Camel

[bigfoot]

haha lol!

[NEMESIS]

An i quote camel

"between various inhospitable portals to hell"


Way too funny!



Hope your doing well brother

[Camel]

Yeah, thanks guys, I'm surviving. In Karumba today. Of all places.....

While I'm waiting for the client to show up I'll post some more on my ECU Editor exercise. So today: Data logging, fitting a Wide Band sensor

Iâ€ve fitted a data logger to the K8 to gain insight into how the Engine Control Unit (ECU) operates. For benchmarking purposes Iâ€ve reverted the bike to stock so Iâ€ve got something to compare modifications against. I even reflashed my ECU with the stock K8 bin file (map), because itâ€d be bloody pointless benchmarking without it! The only mod the bike now has is plugging of the airbox-to-PAIR line. When next home Iâ€ll be looking for a reputable dyno in Perth to get a more scientific assessment of base-level performance.

Once Iâ€ve got my benchmark figures Iâ€ll start making modifications and documenting my findings. First mod will be flashing the ECU with a new base map that has the top speed limiter and gearing restrictions removed. This basic change (using a pre-release version of ECU Editor 2(EE2)) essentially applies the fifth gear map to all gears. Iâ€ll also deactivate the standard narrowband HO2S using EE2 (this feature currently being programmed, not available in latest download). Then Iâ€ll do the usual basic mods, air filter and pipes, and make adjustments to the ECUâ€s ignition and fuel maps using EE2 to maximise the modifications potential. Then back on the dyno to finetune and get a reading.

Anyway, back to the data logger. I purchased Tech Edgeâ€s WBo2 version 3B1 data logger, which was supplied with a Bosch LSU 4.2 wide band oxygen sensor.



The Bosch LSU 4.2 wideband oxygen sensor is still the weapon of choice for data logging. Bosch has since released the 4.9 and a mini 4.9 (28 grams as opposed to the 4.2 and 4.9â€s 120 grams) but Iâ€m still undecided on the usefulness of the advantages in accuracy the 4.9 has over the 4.2 in the lean and rich areas. Besides, with the 4.2, pricing is better, with Tech Edge selling them for AUD 97.00.



Donâ€t get the wide band sensor confused with the stock Busa narrowband jobbie. Their capability and job descriptions vary like chalk and cheese with the LSU 4.2 making the narrowband look like a real dumb bastard. However, theyâ€re both based on the same principle, using yttrium flavoured zirconia dioxide ceramics and thin porous platinum electrodes to measure the voltage produced by oxygen ions transferring through the ceramic when it gets above 350 degrees Celsius. The ions move when there is a difference between oxygen levels on the exhaust gas side to the reference (outside) air side. The difference between the narrow and wideband sensors is that whereas the narrowbandâ€s signal is merely a direct voltage reading, the wideband uses a ‘pump cell†to maintain a consistent stoichiometric air/fuel ratio in the sensorâ€s monitoring chamber and the signal sent to the datalogger is actually a figure (between 0.7 to 4 volts) determined by monitoring and measuring the current used to maintain stoich (14.7:1) in the chamber.

If youâ€re interested in reading up on the LSU, Tech Edge has a comprehensive library of information on exhaust oxygen sensors available on their web site at http://wbo2.com/lsu/sensors.htm

Hereâ€s the standard Busa narrow band sensor on the left, and opposite; the wide band sensor. Obviously the wide band oxygen sensor (WBO2) wonâ€t fit in the bung hole for the narrowband sensor.



Now, I could flick the standard narrowband sensor completely and use the narrowband emulator from the 3B1 data logger, but apparently the LSU wideband sensor has a much slower response time than the narrowband and this might persuade the ECU to flick me a fault code. Also, I want to keep the bike as stock as possible for initial data logging so apart from the presence of the widebandâ€s probe in the gas stream everything should be stock. Another reason is that I donâ€t want the wideband in there all the time; I want to use it on a couple of other projects.

Hereâ€s a blow-by-blow account of my wideband sensor fitment to the standard Busa exhaust.…..

Fitting the wideband sensor was relatively painless. Basically remove the lower fairings, radiator, oil cooler and horn so the headers could be taken off.

Hereâ€s where Iâ€m placing the wideband sensor (blue crosshairs), basically mirroring the position and angle of the narrowband sensor (in green).



The sensorâ€s readings are affected by all manner of things such as heat, pressure, proximity to outside air and operation of the PAIR valve. Having the sensor upstream and close to the catalytic converters (shown in red) helps maintain stability in pressure and temperature and certainly uninfluenced by the air near the exhaust outlet.

The image below shows the position of the narrowband bung (red arrow) and where the wideband bung will go (green arrow). Between the two is your standard Suttonâ€s holesaw. I could have used a HSS annular cutter but didnâ€t want to ruin it as the cat body is double-layered ‘heat treated†stainless steel. I was prepared to sacrifice a relatively inexpensive holesaw.



And sacrifice it I did, thatâ€s a tooth from the holesaw (arrow, below). I picked it out, applied heat to the metal and brought the two layers together before deburring. Then I ground off the heat treatment paint from around the hole. Note that the catalytic converter can clearly be seen through the portal.





Camel

---

continuing right along.....

Now for the bung. I needed an 18x1.5mm nut for the LSU 4.2 wideband sensorâ€s thread. I also needed to trim it back so the probe was exposed to the gas flow but not so far that the thread was exposed. 13mm does the job. Did I buy stainless? No, itâ€s in there for a good time, not a long time. An inexpensive grade 8.8 steel jobby.



I bevelled the perimeter to get good weld penetration and screwed it down onto the bolt which will act as a heatsink while welding.



The bolt also helps me position the nut at the required angle. The LSU should be mounted at a minimum of 10 degrees from horizontal to prevent condensation or fuel pooling inside the probeâ€s outer shell. Liquids will quickly kill the ceramic sensor tip.




]

Once positioned I assailed the nut with the MIG using mild steel filler wire. Yes Dorothy you can MIG weld stainless using mild steel! Do you think the exhaust shops use stainless wire to join your SS exhaust?



Now Iâ€m not keeping the LSU in all the time. If you leave it in the exhaust unheated it will quickly be ruined. So I cut down the bolt and made a copper washer for it. I made the copper washer from some old 2 inch copper waterpipe I had laying around. Annealed it with the oxy, cut it out with a holesaw, trimmed it and then gave it a trial fit. Iâ€ll be cutting the bolt head right down to save weight.



To be continued....

Camel

---

The story continues...

Don't you just love annealing? That luverly cherry red glow....



It's round where it counts...


I'll be trimming that head right back to save weight.



After trial fitting the bolt I removed it and then carefully trial fit the LSU, (carefully because theyâ€re fragile).



Here is the LSU in position on the bike. Carefully out of the way of the fairing and other bits. I need to ensure it doesnâ€t get coated with chain lube, particularly on the wire sheath as this is where the reference air is sourced from.



Then I reassembled the bike sans lower fairings ‘cause I had a need to go out and punish the bitch....



Next, Mountin†and wirinâ€

Camel

[Camel]

Mountin†and wirinâ€

Time was fast getting away on me so I decided Iâ€d mount the Tech Edge 3B1 unit (which houses the data logger and heater control) and the Tech Edge LDO2 LED display unit so I could at least see if the system was working. Rather than go with rubber dampened bolted mounts I took the quick and more effective self-adhesive-backed Velcro approach. This way the data logger fits neatly in the Busaâ€s cavernous underseat storage compartment and may be quickly repositioned or removed to gain access to wires, thermocouples, etc.



The display fits on the top triple clamp, out of the way but still visible. Its cushioned against vibration and easily removable thanks to the modern miracle of Velcro.

Everything fits together well, the wideband plugs in and is secured by a knurled nut (the silvery knurled thingy at the top of the image below). Power plug clips in with a satisfying ‘snick†sound (the white plastic thingy below the silvery knurled jobby) and the display unit plugs in on the front plate of the datalogger via a RJ45 plug.



Once wired up I calibrated the datalogger by first removing the wideband sensor from the exhaust and exposing it to ‘free air†i.e. uncontaminated breathable air with an oxygen concentration of 21%. After powering up the datalogger I wait a couple of minutes while the sensor reaches operating temperature (heater signal and power supplied via the datalogger) all the while making sure I donâ€t let my rancid CO2 and garlic contaminated breath foul the sensorâ€s readings. The sensor canâ€t be moved while calibrating so it needs to be rested on something clean. Preferably something that wonâ€t be melted by the heat.

Once heated I hold the data-logging button in for at least four seconds and then watch the red LED (see image above) as it merrily flashes away for about 15 seconds before sending out a frantic burst of six flashes and returning to steady state. Turn the unit off, let the sensor cool, screw it back into the exhaust then plug ‘n†play.

Now letâ€s ride! Before starting the bike I turn the key to ‘on†and let the LSU wideband sensor heater warm up the tip. The display unit gives me the good word (heat) and then indicates ‘lean†when ready.



Start the engine. Let it warm up. Then while idling the display returns an air/fuel ratio (AFR) of around 12.6:1.



This ties in well with PetriKâ€s findings, i.e. the Gen II HO2S has a window of operation as follows:

Throttle Position 1-9%
Coolant Temperature greater than 80C
Intake Air Pressure 6-60 units (as per EE2)
4000-6000 RPM

Then I rode the bike, keeping in mind that I still have the PAIR valve operating just to see what it would do. Sure enough, it works almost all the time; I observed readings between 18:1 and 29:1. So I turned around (I live in excellent scratching territory ) hammered it home, removed the airbox and shoved a marble into the hose between the airbox and PAIR valve (a green marble) to stop it giving me false readings. Note that the PAIR valve needs to be kept in place or a resistor placed inline otherwise you get a fault code. In time Iâ€ll use EE2 to program a shift kit via the PAIR valve connections.



Now with the PAIR disabled and the HO2S still connected I went for a lovely little jaunt through the hills. Hmmm. That HO2S is bloody effective. Other than when cranking the throttle the AFR clings stubbornly to 14.7:1 (stoichiometric rate). Sure it varies, but not much more either side of 14.82:1 or 14.6:1.



So, after pulling up in a sun-dappled forest glade peppered with a splendid display of spring flowers (you can almost see Bambi frolicking there canâ€t you?) I disconnected the Busaâ€s standard HO2S. Turned ignition on, saw fault code, then turned off and back on. No fault code, (but we all know itâ€s there, donâ€t we?) Wait for sensor to heat, then kick her in the guts (metaphorically speaking) and tootle off.

At idle, no change to AFR. Accelerate, and depending on throttle position, revs, etc the AFR can jump as high as 11.3:1! At cruise, where previously the bike ran in the HO2S window, (and therefore returned the stoichiometric rate of 14.7:1) the AFR is now in the mid 12â€s. Hmm, thatâ€s rich. I donâ€t reckon thatâ€s good for a bike that spends most of its time cruising. Thatâ€ll eventually root the LSU sensor and the catalytic converter. Not a drama now, but could be when/if they make downstream oxy sensors mandatory on bikes. These are used by the ECU to determine if the catalytic converters have passed their use-by date, or have been tampered with. If yes, the ECU returns a fault code and the bike goes into limp-home mode. Well if it happens its in the future and we can deal with it then.

So what have I learned to date?

Well keep in mind that I havenâ€t actually logged data yet. Iâ€ve been sneaking glimpses of the display unit while riding. While accurate I still want a data log where I can relate AFR to a host of other variables.

What I have learned though is that the PAIR valve and HO2S do a fantastic job in reducing hydrocarbon emissions when working in conjunction with the catalytic converters. Also, I suspect that without the HO2S fitted your fuel usage would jump significantly (if doing normal riding cruising). Iâ€ll be logging this to gain some real-world information that will be useful to the 98% of riders who donâ€t drag race their Busas.

For performance tuning I believe the HO2S shouldnâ€t be scuttled entirely. Sure it will fight your map changes, but only in the HO2S window, i.e. when cruising at legal speeds. I think there is a lot of advantage to be gained by lowering the top end of the HO2S window from 6000 rpm back to the Gen I cap of 4800 rpm or even lower. I donâ€t think youâ€d disable the RPM range altogether as it would create havoc when you were running at high RPM at part throttle positions. If you could alter the rpm range of the HO2S window the rider would know when the richer ‘performance†map slips in and can keep the revs above that area to maintain max power without having the stoich-to-rich transition lag occurring. The skilled tuner could also set the top HO2S limit and then tailor the fuel map to smooth the transition further. That way I reckon you could get the best of both worlds, better performance while still retaining good fuel economy when cruising in the HO2S window.

Camel


In the next episode..... tapping into what's going on, analogue, digital and thermocouples...

[motoplast]

A bloody very interesting read Camel, I for one, am looking forward to the next installment. Well done mate.

Cheers,
Tex & Bundy

[Camel]

Thanks Tex!

[fasterfaster]

Camel .............. good indeed, love the descriptive literature
If building a complete full custom fuel map would bin the o2 sensor but, as long as you set up your map for the A/F readings you desire at cruise throttle economy is still retained.
Now I'm off to buy a pic of Bambi in the flowers to hang on the dyno wall

[Camel]

Now I'm off to buy a pic of Bambi in the flowers to hang on the dyno wall

Thank God HR PC hasn't affected small business!

[Camel]