Using O2 sensors for carb tuning?

[danci1973]

Hi!

When I tinkered with a friends car (and later my own too), I discovered an interesting DIY project http://www.diy-wb.org - it is a wideband O2 sensor controller used to measure A/F ratio.

It sounds like this would be a usefull tool for tuning the carbs, however I'm not sure if it would work well on a bike - due to it's rather short exhaust.

Does anyone have any experience with that or more knowledge than me?

Thanks, D.

[jiggersplat]

i don't know of any reason why it wouldn't work.

[V8Pinto]

I built something similar using an LM3914 and it works pretty well.  I was able to shrink it all down to fit inside a 2" gauge housing.  It's cool to watch it on a computer controlled vehicle.  If you hold your foot steady for a few moments, you can see the computer start to cycle everything lean-leaner-leaner-then back to lean.

I haven't gotten around (as usual) to installing it on the GS.

[The Buddha]

Yea I did it ...back in 99 and never stopped carping about it ...
Buy a sensor for a 94 chevy cavalier and it has 1 wire and a simple nut in the pipe to hold it ... now you need to slice the nut in 1/2 ....only need ~3 threads ... then weld in your pipe ... and then mark your throttle positions in the grips and switches and ride on ... works great... and you are sure to hit the Srinath formula square on the head...  :lol: ... Unless you got a heavy modded motor...
Cool.
Srinath.

[jiggersplat]

srinath,

the problem with just grabbing the o2 sensor out of a car, is in general they are not suitable for wide-band.  they pretty much just say lean or rich.  they don't have a linear response.  the more sophisticated wide-band kits have a response that is much more linear and tell you HOW rich or lean you are.  there are a few car o2 sensors that are suitable, which are probably mentioned in that diy wideband site, but they are hard to come by.

[The Buddha]

The one I had was simply to be read with a digital millivolt meter... under .500 is lean and over .500 is rich... its analog ... wideband ... what is that ... Matter of fact I have the thing in my eli1000 right now ...
Cool.
Srinath.

[jiggersplat]

okay, so yours you basically know rich or lean conditions right, but not HOW rich or HOW lean?  a wideband will have a linear (or close to it) response, so for example for every point richer, you might get 0.1v higher reading.  stoichiometric is the a/f ratio where there is exactly the right amount of oxygen molecules to fully combust all of the fuel.  left of stoich is rich, right of stoich is lean.  the graph is the voltage you would read out of an o2 sensor for a given a/f ratio.

[The Buddha]

Woweeee dude ... You do know the damn GS is a POS carbureted bike with carbs that predate the dinosaurs and a motor that possibly uses more oil than gasoline right ... in any case technology is technology and the extra knowledge is great bench racing talk ... so ... knock yourself out... and post back here how it goes...
Cool.
Srinath.

[90GS500rehab]

The one I had was simply to be read with a digital millivolt meter... under .500 is lean and over .500 is rich... its analog ... wideband ... what is that ... Matter of fact I have the thing in my eli1000 right now ...
Cool.
Srinath.

That is a narrow band. Plus the single wire sensors cool down too much to keep accuracy. A 4 wire narrow band is not too much more and has a dedicated heater and ground circuit.

If anyone doesn't want to build a wideband kit, you can get a wideband sensor and meter from Innovative for $300. It will also datalog but we don't really have any sensors on the GS to measure except voltage and rpm.

[danci1973]

I was looking at the Innovative unit, but $300 sounds a lot for tuning one GS...

OTOH, DIY could be fun to build. Provided I can get my hands on a WB O2 sensor for less money.

D.

[Rema1000]

Little off-topic:  when people say that their O2 sensor shows an Air Fuel Ratio of (say) 14.6:1, they are referring to the mixture of the intake, right?... but they are measuring the exhaust.  If so, then it sounds like the O2 content of the exhaust is measured, and then the AFR is derived, based on an assumption of the fuel type (100% octane fuel).

The reason I ask, is that in MN, all gasoline is 10% ethanol (and the governor has called for 20%, like the gasoline in Brazil). Ethanol has an ideal ratio of 9:1.  That might mean that in MN, an ideal AFR is closer to 14.04:1.  But if the O2 sensor is just looking for left-over oxygen in the exhaust, then perhaps this difference in fuel doesn't matter.  When the AFR analyzer says "AFR is 14.6:1", then maybe what it really means is "There is the correct amount of O2 in the exhaust.", in which case the AFR is optimal regardless of fuel.

...?

[The Buddha]

The one I had was simply to be read with a digital millivolt meter... under .500 is lean and over .500 is rich... its analog ... wideband ... what is that ... Matter of fact I have the thing in my eli1000 right now ...
Cool.
Srinath.

That is a narrow band. Plus the single wire sensors cool down too much to keep accuracy. A 4 wire narrow band is not too much more and has a dedicated heater and ground circuit.

If anyone doesn't want to build a wideband kit, you can get a wideband sensor and meter from Innovative for $300. It will also datalog but we don't really have any sensors on the GS to measure except voltage and rpm.

OK mine cost me $20... and it works fine 5 mins after you start up the bike cos the exhaust heats up the sensor just fine ... narrow band ... you do realise the GS is old POS that has carbs dating back to 1921 right...
Cool.
Srinath.

[goat]

Well, I just bought a 1 wire o2 sensor off ebay for $15 shipped. I realize that it isn't wideband and isn't heated, but I'm just looking for an indicator. My question is this: Do I have to weld it to my header, or can I just make something that will hold it in the end of the can?

[The Buddha]

I used a nut that I welded in the exhaust and screwed it in, When  done, I swapped that exhaust for a exhaust wihtout ... though a bolt in that nut would have worked too ...
Cool.
Srinath.

[V8Pinto]

If you can't find a nut, go to your autoparts store and buy "non-foulers".  Cut the right amount off and weld it to your pipe (note-weld the bung on with the spark plug threaded in.  Heat distorted the last set I did and I couldn't get a plug in afterwards).

I made up my whole setup for maybe five bucks.

[tkm433]

Hi!

When I tinkered with a friends car (and later my own too), I discovered an interesting DIY project http://www.diy-wb.org - it is a wideband O2 sensor controller used to measure A/F ratio.

It sounds like this would be a useful tool for tuning the carbs, however I'm not sure if it would work well on a bike - due to it's rather short exhaust.

Does anyone have any experience with that or more knowledge than me?

Thanks, D.

To get the best results you would want to mount a sensor mount in each down pipe so that you could test each carb.  You could do this one at a time for the test.  By doing this you will be able to read each carb.  If you mount the sensor in the manner that it is done in cars you would end up with a reading for both carbs where one could be running rich and the other lean and the end results would average the two.

I know that when they dyno a bike they only place a sensor in the end pipe so that they can get an overall picture of how the motor is running and also they do not have to weld mounts in each pipe which would take forever and cost a good deal.

[werase643]

Tod,

the carbs are same
the cylinders are same
the exhaust...almost same length

and A/F mix is desired to be about 14:1

have you ever heard of a GS having different size jets from side to side

a V-twin  yes due to the front cylinder getting more air cooling but not a parallel twin

why would ya need 2 O2 sensors

a car doesn't have 4-6-8 sensors... one for each cylinder

[tkm433]

Tod,

the carbs are same
the cylinders are same
the exhaust...almost same length

and A/F mix is desired to be about 14:1

have you ever heard of a GS having different size jets from side to side

a V-twin  yes due to the front cylinder getting more air cooling but not a parellel twin

why would ya need 2 O2 sensors

a car doesn't have 4-6-8 sensors... one for each cylinder

I did not say you needed two sensors but stated "for best results" that would be the case.

As for your view on v-twins you are correct but the same is true on a parallel twins and all motors with more than one cylinder that not all cylinders will run the same temperature due to air flow, if water cooled then flow patterns of the water through the motor, The clutch side of the motor might generate or displace heat more than the generator side of the motor. Maybe not an issue on the GS500 but again I did state "to get the best results" you would want to check each cylinder.

As for your car theory I guess you are correct in that they do not have a sensor per cylinder but why do some cars have one for each cylinder bank such as on a General Motors V8?? I know why but the application still states "Two required " Check out link below.
http://www.oxygensensorwholesale.com/chevrolet_2003_corvette.html
This is a V8 and they run a sensor on the right and left side exhaust.

Now off of the issue of the O2 sensor which is more or less telling you how well the fuel air mixture is burning would you use only one exhaust temperature probe -- a tool to tell you how the mixture is burning -- to fine tune a RZ350 motor which is a parallel twin and trust the readings off  just one cylinder to know that the other is running the same if the carbs are set the same?

Now back to the GS500, let's say that one is truly interested in getting the carbs set correctly by using one sensor.  Where the two pipes meet would work but if one did have a problem such as an air leak in one carb system the O2 sensor would not know which side of the intake was bad.  By having or using two sensors you would be able to tell how each cylinder was burning and not just the pair.  If both carb settings were the same but one cylinder sensor was showing a leaner burn then it might alert you to an air leak in that side of the system from a bad carb mount to a faulty o-ring that could lead to trouble down the road.

Again I said for "best results".

[goat]

Hmm, I don't think I asked my question well enough. I'll try again.

I realize that this isn't going to be the most accurate thing in the world. It's a 1 wire O2 sensor. I'm not looking for perfection. If I wanted higher quality, I would have sprung for the WB sensor and built a control unit. I'm just looking for something to help me get my carbs set correctly so I stop fouling up plugs and get better gas mileage than my car.

I want to mount the O2 sensor temporarily here (pic stolen from the gallery and altered):


Is this going to make the reading from the O2 worthless? Do I need to drill a hole and weld to my header in order for the sensor to work?

[cernunos]

The sensor should probably be mounted further upstream. I can't recall off the top of my head but I think the one-wire sensor has to reach a temp in the neighborhood of 500-600 F before it starts working properly. Don't quote me on that but I do believe that the end of the exhaust wouldn't be hot enough. Fiddle around with it and see what happens...run it as a class experiment. Love the little G and this forum.

C.......