Topic: Using O2 sensors

[eldar]

So I have been toying with the idea of using an o2 sensor for tuning. I mean it IS more accurate than a plug chop is or even using a shop tach.

Now this does not seem as complicated as even doing a carb sync.
You get an o2 sensor, preferably a heated style with 4 leads. A good high impedance digital volt meter, powered ones seem best.  Hook the sensor leads to the volt meter and the heater leads to a 12V source. Now with a warmed up engine, set the volt meter to 1 volt and watch it. The stoichiometric ideal is 14.7-1 which reads at about .45 volts on the meter. Probably going at 14-1 maybe 13.5-1 would be about .5 volts or so and would probably be ideal.

Here is the problem, an O2 sensor does not activate till it hits 600F. Pretty hot. So now using a 4 pipe exhaust on a hot engine, do you think the temp at the end of the pipe will be hot enough? Remember this is a heated O2 so it warms up faster and stays warmer longer.

Or would it be better to tap holes underneath the pipes at the joint between the header and muffler?

Remember, I have slip-ons so my exhaust is not all stock.

Any one done something like this?

[paulages]

funny- i just welded an O2 bung into the collector on the header going on the 718 yesterday. i'll need the help finding the mixture on that one for sure, and don't think i want to pay dyno prices until i get close at least. mine mounts just in front of the center stand (when on it), and just behind the oil pan.

[eldar]

Thing with it is, is that unless you do each cylinder separately, you only get the sum of all 4. While not bad, you cant tune each one on its own. For me, with all 4 pipes, I wll probably drill and tap the holes. That way, when I am done, I can put a bolt into the hole to seal it.

[paulages]

there's not really much room in each downpipe for a sensor. at least not on the 650 pipes. i realize that the collective mixture will not be as accurate as individual readings, but it will give me ballpark mixtures to work with. a dyno doesn't read individually either. welded at the inside of the collector, the bung isn't visible with the sensor or a plug. i'd be hesitant to add bulky, visible fittings in the downpipes personally. i say go for it though, if that's what you want to do.

[eldar]

ballpark is better than nothing isn't it! Yeah I am really looking into it. The holes would be there but sealed. There isnt a lot of room in the 750 pipes either. I am just wondering what would be the best location for the holes. I would like it if the exhaust was hot enough at the end but I just do not think it will be.

[eldar]

Do you currently have an o2 sensor now? How big are they around and in probe length?

[hymodyne]

would someone explain what the sensor is supposed to do, and how it replaces (ideally) the plug chop and Dyna measurements? What does the sensor connect to? a gauge, or some sort of electronic meter?

hym

[paulages]

the sensor, when hooked up to an autometer--AKA air/fuel gauge-- will tell you if you are running rich, lean, or stoichiometric. a wideband sensor will tell you how rich or lean, whereas a narrowband will only tell you one of the three. do a google search on "autometer", and you'll come up with all kinds of info.

eldar- the sensor i have came off of a ducati sport classic. it's probably 4 inches in total length, and the probe is probably about an inch long.

[hymodyne]

Many thanks, I'll do a search on autometer to get more info.

hym

[eldar]

Wow an inch will barely fit! how big around is it? Just want to know how big the hole is going to be!

[KeithB]

I have seen O2 sensors that are used at the tail pipe of a car so it should work that way for the bike.
With a 4-4 exhaust it's a winner.
1 carb per pipe.
4-2 would mean working with 2 carbs to get a balance and a 4-1 would be the hardest.
With some patience, it should work.

[eldar]

That would be ideal keith, do you know what brand of sensor or what I need to look for? I would like to avoid holes if possible!

[bradweingartner]

For tuning I wouldn't suggest anybody even bother with a narrow-band O2 sensor.The only place where the NBO2 works is in a closed loop feedback EFI system. It allows the system to basically bounce back and forth accross the threshold. It gives you decent ballpark stoich for cruise but lacks the resolution necessary for high-load wide open throttle tuning.

The Wideband O2 on the other hand is exactly what the doctor ordered. Problem is, until recently they were out of the pricerange of the average tinkerer. Now you can get into one for complete for under $250 with the Innovate LC1 setup. Which I guess is still a lot unless you're considering the cost of rebuilding a blown up motor that you put money into.

Now for the folks handy with electronics, the "Just Another Wideband" controller is a DIY kit (or you can buy it preassembled). Total cost for that setup if you assemble it yourself and get a good price on the O2 sensor should run you under $150.

While this may still be considerably more expensive than the standard 3/4 wire heated O2 sensor and an autometer A/F gauge, the different between the two technologies is NIGHT AND DAY. I've tuned custom turbo cars with both setups and I will never, ever, ever again tune a high performance car on a Narrow Band O2. Just not enough information to make me comfortable!

[eldar]

Well we are not tuning a race car. Even so, I am not going to use a gauge, I am going to use a voltmeter which will respond faster with a digital number. If I can get it to .45 - .55, I will be happy and for a 30 year old engine, that will be more than adequate.

[bradweingartner]

That's great in theory, but it's not exactly how it works.

It's not just resolution. In that, the Wideband puts out 0-5v signal, and a narrowband is 0-1v. But rather the linearity of the signal. The wideband is linear. The narrowband is not, in fact, a narrow band is like an on-off switch.

Check the picture:

So while it may seem like you can hit that perfect stoichiometric voltage, your tuning options are way too coarse to actually do it. It's hard to do with EFI where I can manipulate the pulsewidth by the tenth of a millisecond. But the REAL reason a narrow band is no good?

Optimum power and optimum efficiency don't happen at 14.7:1, which is the only place where a narrowband gives you any sort of information. While you're right that these aren't exactly race engines, they still have a pretty good specific output. That and lots of people here have thousands into their rebuilt engine, $150 is cheap insurance!

[paulages]

here you go eldar:

one with the plug (never mind my shoddy welds):


one with the sensor installed:


one showing total length of the unit:


one showing the length of the sensor portion:

[MRieck]

That's great in theory, but it's not exactly how it works.

It's not just resolution. In that, the Wideband puts out 0-5v signal, and a narrowband is 0-1v. But rather the linearity of the signal. The wideband is linear. The narrowband is not, in fact, a narrow band is like an on-off switch.

Check the picture:

So while it may seem like you can hit that perfect stoichiometric voltage, your tuning options are way too coarse to actually do it. It's hard to do with EFI where I can manipulate the pulsewidth by the tenth of a millisecond. But the REAL reason a narrow band is no good?

Optimum power and optimum efficiency don't happen at 14.7:1, which is the only place where a narrowband gives you any sort of information. While you're right that these aren't exactly race engines, they still have a pretty good specific output. That and lots of people here have thousands into their rebuilt engine, $150 is cheap insurance!

I agree with your statements. There are quite a few wideband meters out there with the ability to hold a lot of data. I welded sensor bungs into my CB and FJ pipes years ago.

[paulages]

That's great in theory, but it's not exactly how it works.

It's not just resolution. In that, the Wideband puts out 0-5v signal, and a narrowband is 0-1v. But rather the linearity of the signal. The wideband is linear. The narrowband is not, in fact, a narrow band is like an on-off switch.

Check the picture:

So while it may seem like you can hit that perfect stoichiometric voltage, your tuning options are way too coarse to actually do it. It's hard to do with EFI where I can manipulate the pulsewidth by the tenth of a millisecond. But the REAL reason a narrow band is no good?

Optimum power and optimum efficiency don't happen at 14.7:1, which is the only place where a narrowband gives you any sort of information. While you're right that these aren't exactly race engines, they still have a pretty good specific output. That and lots of people here have thousands into their rebuilt engine, $150 is cheap insurance!

I agree with your statements. There are quite a few wideband meters out there with the ability to hold a lot of data. I welded sensor bungs into my CB and FJ pipes years ago.

the way i read it described elsewhere is that essentially the NB only tells you RICH, STOICH, or LEAN, but now to what degree, whereas the WB actually gives you gradations within these parameters. is that not correct?

also, any guesses as to whether the ducati lamba sensor is WB or NB? couldn't find any data on it anywhere.

[troppo]

I`ve been going through my favourites for two days now but cant find the link to a site that showed how to set up lambda sensors on old bikes but cant find it anywhere.
So a couple of googles and i found a few sites that may help

http://www.techedge.com.au/vehicle/wbo2/wbfmd.htm

http://www.redline.lt/magazine/spec-features/article/article/17/1/

http://www.pelicanparts.com/techarticles/mult_air_fuel_monitor/mult_air_fuel_monitor.htm

https://www.msu.edu/~levesqu6/airfuel.htm

http://www.autospeed.com/A_0217/cms/article.html

Hope this helps

[KeithB]

I have had a few turbo cars on the dyno for tuning and some used sensors that replaced the one in the collector pipe and others that used the tail pipe style.
I believe they were both wide band.

[mystic_1]

Heres a link I had floating around in my bookmarks which describes building a readout to use with an automotive O2 sensors on motorcycles:

How to build an air/fuel guage for tuning carbs

Also have this link, describing the same sort of project but with off-the-shelf components:

Air/Fuel Ratio Monitor


hth

mystic_1

[bistromath]

Just came in here to say that the only sensors that will work in the tailpipe of a car are wideband sensors or heated narrowband sensors. Most narrowband sensors don't have an accessory heater and will not work if they don't hit 600F. You'll want to tap the headers or the collector.

[eldar]

From what I have been reading, a narrow band will work fine. They are virtually ALL 3-4 wire now. So are heated. And DO produce a reading based on o2 levels that is readable. The reading range is .2-1 Volt and the A/F range is 12 - 17. So they will allow a person to tune according to voltage. The widebands main advantage seems to be a wider voltage range to allow finer tuning and wider A/F range. As my engine is stock, EXACT tuning is certainly not required. Stochi tuning does not produce best power, this I know, it is the point where all fuel and air is completely burned.

AS I said, everything I have read points to narrow band NOT being on/off.

http://www.datsuns.com/Tech/oxygen_sensors.htm
http://www.scirocco.org/tech/misc/afgauge/af.html
http://www.3barracing.com/product_3.htm
http://www.chevyhiperformance.com/techarticles/53159_nordskog_budget_air_fuel_meter/index.html
So it appears that wile maybe not accurate as a wideband,a narrow should be fine.

Paul, those welds look just fine. Seems pretty nice and easily accessible.

[bistromath]

Narrowband O2 sensors CANNOT be trusted to give a reading EXCEPT to say you are stoich, rich, or lean. Past that, there is just too much uncertainty to be able to determine A/F with any reasonable accuracy. In fact, richening the mixture can cause the device to read a LOWER voltage if you are on the rich side of stoich already! The numbers you see on the datsuns.com link might be good for one sensor, at one temperature, but they cannot be extrapolated to other O2 sensors at all.

While there is a wide voltage range coming out of the narrowband O2 sensor, it is NOT linear and NOT deterministic except at exactly 14.7:1.

That said, if you're just looking for an easy way to jet your bike, I say just run a narrowband sensor and get it just rich of stoich and be done with it. It's not like we're running turbo or nitrous cars where running a hair lean will destroy the thing. The Chevy LT1 guys often tune their cars so they produce 880mV on the O2 at WOT; this isn't exact but it's usually close enough to max power and since you're running rich anyway it's not like you're in danger of getting too lean.

[scondon]

 Eldar, will your set-up be for reading while riding on the road? I did read through the posts just didn't catch this info.