O2 Sensors

We pulled the head of my bike today to fit inserts for the three studs and I was quite surprised at the black junk on the piston tops and the combustion chambers. It runs really well and I wasn't ready for the evidence richness it showed. For a little while now I've been contemplating fitting an O2 sensor and gauge to see how it is running right through the rev range. I get over 60mpg so it would be interesting to take this to its logical conclusion to see how much better it can be.

A mate of mine with a lot of experience of engine building said something very interesting once. He said the greatest benefit of fuel injection was that the engines can be run much leaner, resulting in longer engine life. Something which runs around my head from time to time.

Any experience on here on any particular make and model and results achieved?

For a non fuel injected engine a wide band Bosch O2 sensor is the one to go for.

Yes, but wahat sort of monitor/readout?

i've used an AEM successfully on several triumph twins to look at the carburation. mine came as a simple kit-- the gauge (with built-in controller), a UEGO O2 sensor with harness to hook to the gauge, and another small harness to hook the gauge to the battery. i put the gauge on the triple tree.



this gauge combines a digital readout with a light that circles the perimeter of the gauge. the readout is slow enough that you can actually read it, but fast enough to track mixture changes as you go through the throttle. the little light is so sensitive that it's mostly worthless.

https://www.rbracing-rsr.com/rsrgauge.htm
If it works on a Harley...

I use a Koso narrow band on my commando,it's helpful for setting up from nothing ,but if your bike is running well and getting 60 mpg I doubt it would help you get it any better,it "may" help you accertain where in the carbs range you are getting the richness , cheers

I thought Jim Comstock had a fuel injected/O2 sensored/computer controlled fuel system... Paging Mr. Comstock to the white courtesy phone....

they can be tricky to interpret. i first had one in a muffler, and would get readings quite a bit leaner from those i got after i moved the O2 sensor up into the head pipe.

the chemically correct hot-rod ratio is supposed to be around 12 or 12.5 to 1 (air to fuel), but my street bike was fastest at 10.5 to 11.0. i think the apparently rich numbers were a function of lousy triumph combustion chambers. to get a correct burn you have to dump in extra fuel that will be detected in the headpipe as over-rich. but it doesn't show up in the muffler, because by that time it has burned itself up without adding any power.

might be more accurate with a norton chamber, as flat as the pistons are.

I think I would err on the side of caution with a richer setting for air cooled engines. I think Jim C had some problems with pistons going south when he went thru some bad gas on his fuel injected bike. Older post. If I recall we talked about it at the Barber Motorsport a number of years back.
Cheers,
Thomas

Too rich is very common. If you lean off the carburation, you should be able to tell when you are loo lean. With the main jets, it is by doing a plug chop and reading the plug. And it is always better to stay slightly too rich. Most people don't ride the bike on full throttle for long anyway. Getting the needle and needle jet right is of critical importance. If you are too lean you get flat spots, if you are too rich, the motor will be slightly sluggish. An oxygen sensor uses empiricism in its calibration. Achieving a theoretical value for oxygen content of the exhaust gas, does not necessarily mean that it is best for performance or durability of the motor. In fact the effect will probably change with the different types of fuel you use.
I'd question whether fuel injection gives the best results. Especially if the oxygen sensor is used to set it. A dyno might show that slightly above or below the theoretical optimum might give better results. There is another thing too - car guys are usually more interested in fuel economy than most motorcyclists. The set point of the injector for the optimum oxygen content in the exhaust is probably different.
With petrol and using conventional carbs, The needle jet setting for best power is very slightly on the rich side of too lean. The difference between too rich and too lean is about half a thou of an inch in needle jet ID. The difference between a 0.106 and a 0.107 inch needle jet is one thou, and so the five notches on the needle each probably enable about 2 tenths of a thou variation. So normally raising or lowering the needles two notches should cover your whole range of jetting between too rich and too lean..

Deets55 said:

I think the problem with putting the sensor in the muffler is that when the cylinder is in compression the sensor gets contaminated with fresh air coming back into the muffler causing a false reading. Not sure about the science behind it, but it needs to be near the exhaust valve. For some reason 20" sounds about right.

Click to expand...

dunno. i had it pretty far forward in an old 2-into-1 alphabet exhaust, there behind the footpeg. later on i drilled out the right-side pipe just below the first bend and just read the right cylinder.



there are recommendations about what the exhaust gas temperature should be to get comparable readings, and i thought about getting a temp reader and some thermocouple wire to stick up the pipe to find the right spot, but in the end i got close enough to dial it in with a stop watch.

just measuring idle it would read 17 to 1 with the sensor in the muffler, and 14.8 with the sensor in the head pipe, no other changes.

Interesting exercise, but for a bike with carb's would a Dyno run not do the job well enough?
Coincidently pulled the head on mine recently, running a fullauto with FCR's to the settings recommended by Brian on this site, getting 60+MPG and that also looked a bit sooty. Am planning a dyno run when all is back together, and work time permits.

Hi Iain, what was the reason you pulled the FA head?
Regards Mike

Hi MIke,

i was seduced by the waisted head bolt set from CNW. The head had to come off to be helicoiled for the replacement studs.
I also had a slight weep from the head gasket after the Sth Island rally trip earlier in the year, once i had decided to lift the head to resolve that and then the bolts just about bought themselves.
It is with Plazz for helicoiling, should have it back in one piece this weekend.

Ok all good, my FA head is finally due back home, as I had a few things on the go , it got put aside after it was gas flowed .
Im slowly getting all the 880 cc motor bits together and 850 race motor has a new crank as the drive side finally fell off after 4 years of thrashing!
I'll be looking forward to your dyno results once you get it all back together.
Where will you dyno?
Regards Mike

gortnipper said:

http://www.accessnorton.com/measured-air-fuel-ratio-for-optimization-t18609.html

Click to expand...

I had forgotten about that thread and I did not know it had progressed so far. I think an oxygen sensor would be good, but it might need calibrating on a dyno and fudge factors built into the programming of ignition systems. I don't have a heat problem because I use methanol fuel. So Jim Comstock is probably correct - I might be sending you all on the path to destruction.
My commando motor absolutely loves methanol. I would never race a four-stroke using petrol, and I'd only use methanol in a two-stroke if I was desperate to win. I sold my methanol-fuelled T250 Suzuki to a guy because the ignition system would have cost as much as the bike was worth. He won 28 races including 4 historic championships with it. The first time I raced the Seeley, that bike was in the race. I was using a single chromed from disc and went into the death dive in front of it in one slower corner, so I had to get the lead back off him - it wasn't easy.

what i have done is ignore the published numbers on what the O2 sensors should give, and instead calibrate them by running against a stopwatch. doing that gave me numbers for what the motor liked, which i could then look for at other throttle positions. i've used them to tune mikuni VMs and also keihin FCRs, with very good results.

but the numbers i look for aren't the ones the books say are optimum.

If the numbers are based on stoichiometry they probably change with the type of fuel you are using. When I tune my bike, I know how long it approximately takes to get down one particular stretch of race circuit. If I have trouble getting around the next corner, I know I have an improvement. If I had an oxygen sensor, I would set it against that.