Exhaust gas analyzer for tuning mixture

With the little ports at the bottom of the exhaust down-tubes, I'm sure that's how the factory did the initial mixture setting. (My FZR1000's plugs are hard to get at, so it's impractical to keep removing those to check color). I'm curious if anybody has used such a method to set mixture? It seems like a "1 and done" method in that once you get it right, it should be good for quite a while.
I'm curious to see what a "good enough" one would cost. Is O2 enough, or CO better?

 

https://en.wikipedia.org/wiki/Air–fuel_ratio

14.7 to 1

Difficult to find an analyzer with a small enough probe to fit in that tiny hole. I thought that hole was for drainage.

 

O2 sensors make a voltage that you can read with a DVM, if you can get the exhaust to flow over the sensor and the sensor is the right temperature, you've got it.
heated sensors take care of temp, now just figure out how to get pure exhaust to flow over it.
i tried to make one awhile ago and got wild readings on the bike but the car was perfect...?
it's easy to shove it up the tail pipe of a car but not so easy on the bike

 

They are for gas analyzer probes. In the old machines the sensor is onboard the anaylzer and a small pipe is used to send gas into the unit.
Many used sensors that were wired to the analyzer, but only had a short probe tube and kept the sensor outside of the exhaust pipe.
You can fine used ones pretty easily, but for some reason they cost quite a bit now. I used to see full-service Sun machines for less than $50.

 

The best way in my opinion is a wideband 02 in the collector. With a modern system, connect an rpm reference, you can datalog across the rpm band with it too.

 

Far more resolution and data than a carbureted engine needs. It would be a good tool to use when developing a starting map for an EFI conversion though. Get the fueling data for the carbs before switching the hardware to get a rough idea of what the engine really likes.

 

if it goes in the collector you get all four, mixed. that's what i tried, it don't work

 

I was toying with making something that fits into the fairly small openings in each down-tube. I figure with a little back pressure, a tube attached there would get some exhaust flow and I could have an O2 sensor fitting mounted very close. Then, just use one of these (or something slightly cheaper) https://www.summitracing.com/parts/...R-6Jng7ruvab--rGKUrqNuHkin0ApclwaArBuEALw_wcB

Of course, if I could fashion a tube with a 90" bend in a properly sized arc, I could have it function like a pitot tube on a plane and get the velocity of the exhaust to push more gas into the tube.

 

I'm convinced I have a clog on #4's pilot passages since pipe temp at low RPMs is much lower than the other 3. I was about to tear down the bike to get the carbs and plugs out, so I thought I'd check the main jet functioning. I ran the bike and then held it at 5000 RPM for 10-20 seconds, then shut it down. The mains start taking over between 3 and 4000 RPM, so I got a reading of main jet mixture by doing this.
It told me #2 is lean and #4 is rich, so that's something I can work with! Setting up a way to use an O2 sensor would be better to get them dialed in properly though!

 

i'm not on board with that. to be in the main jet circuit the slide needs to be all the way up or at least close . until then the needle jet and main are in play, but, the vacuum operated slide is controlled by load/vacuum. without a load i doubt the slide moves very much at all.
take a cable operated carb and turn the idle screw to get 5k and the slide only moves a little bit, like 1/8 inch

 

hmm interesting. So, being really rich, I could have a colder pipe from all the unburnt fuel. That would also explain the popping in the exhaust as that fuel combusts.

 

Does the popping happen when throttling up, or when throttling down?

 

When the RPMs are holding steady, I want to say. I plan to throw the plugs back in and run it on the stand after tweaking the mixture before I pull the carbs out. I'll know more after I do that.

 

These should screw into the exhaust fittings: https://www.motionpro.com/product/08-0040 I'd have to ditch that rubber o-ring for something else. I was thinking of taking up brazing so I can attach to a pipe that the O2 sensor can screw into. I'll have to see what kind of flow I get out the tube and see if a fan assist is needed. I guess I'll know once I see readings from the meter. https://www.amazon.com/gp/product/B001QFH2S6/ref=ppx_yo_dt_b_asin_title_o00_s00?ie=UTF8&psc=1

 

You will need quite a bit of exhaust flow to affect a wide-band element properly. Those are designed to be sampling in-stream, not in a stagnant pool of gas. Make sure you calibrate any sensor you use to open-air before you reference it for AFR.

 

I have been working on building one also. No way was I going to spend 130$ on a air/fuel ratio meter. I found an A/F meter on Wish.com for 15$ and a 4 wire O2 sensor for 20$. Still have to assemble it and make and adaptor for my exhaust pipe. Hope it works like I think it will

 

I have some parts left over from the turbo genesis project including an Innovate SGC-1 and wide-band. I think there are some 45° O2 adapters around. I have the exhaust off my 550, are the ports the same thread pitch and size?

 

Im going to install an adaptor in my 4 to 1 headers when Im ready, I already got one.
I believe all Yamaha exhaust test ports are the same size. At least they are between my XJ11,Xs11 and Venture. I assume they are the same on other XJs.

 

https://www.partzilla.com/product/yamaha/97013-06012-00?ref=a81c73f7d1684c7db3e9a34764b723a8327582a2

 

Well, it depends on what you're trying to do, if you're trying to diagnose a problem with one carb, it won't. But if everything is running correctly cylinder to cylinder, it will allow you to determine what exactly your jetting is doing across the powerband. Every fuel injected motor, car, pickup, etc since about 1985 has used closed loop o2 feedback, usually 1 sensor per bank of cylinders. I would be very curious to datalog the fueling on my maxim x, the way it feels doggy at around 5 or 6k, then comes alive at 7k-10+. I'm not saying I want to go buy a wideband just for that, but I still am curious. I know I'll never use another one on another gas car, for me it's all diesels from here on out until electric cars become affordable.

 

I've seen some of these older analyzers on ebay, as well as some newer handheld models for $100-$200 bucks online. Anyone have a specific model and use it? My bike runs well, and i've verified the valves are within spec, run propane around all joints to verify no leaks, and synced my carbs with gauges, but I have never been able to properly get a Colortune to work. I've had two early 80s Yamaha 4-cylinders, and I think i'm on my third Colortune, and at no point has any Colortune worked properly on either bike in the 15 years i've been tinkering with them. The bike seems to struggle idling with it in use, and if the Colortunes aren't sparking and causing the bike to miss, they simply show blue no matter if I turn them in or out. I'm a firm believer that they're a garbage product.

 

I've used a wide band in the tailpipe. It's OK for tuning idle mixture. Not going to tune the rest of the range with it, but that shouldn't be needed.

The thing is, you need to get the sensor into the exhaust stream somewhere before it's likely to have any ambient air mixing in. What I did was to buy a length of galvanized pipe from a hardware store, about two feet long, and then I drilled it and welded in an O2 bung so that it would be just at the end of the muffler, with the rest of the pipe sticking further back. This extra length of pipe keeps air from getting to the sensor, but also changes the resonance of the exhaust. At idle that shouldn't be a problem, though. The pipe certainly doesn't add any noticeable back-pressure.

You still need basically no gap where the pipe goes into the muffler, and the pipe should stick into the muffler a bit to further prevent induction of air into the pipe. I've used this successfully on a Kawasaki GPz 750 with a 4:1, and in stock exhaust, but with baffles removed, in a Ninja 250.

If you're going to be tuning a particular bike from time-to-time, it would be better to weld an O2 bung into the collector. They can be purchased with a plug so you can just seal the test port when it's not in use.

 

My header pipes are original, so they’ve got the little ports at the bottom where an EGA probe can be screwed in. Not sure how specific these ports are, and if you need to fabricate some sort of adapter when using anything other than the original Yamaha EGA.

 

I know this feed is a little old but I'm trying to find alternate ways to tune the carbs because I just can't seem to use a colortune very well. I'm curious about these O2 sensors. To get an accurate reading could you just remove the exhaust EXCEPT the headers and place the sensor on the bottom end of one of the headers (where it connects to the collector) and turn on the bike and start taking readings? I know it would be a little loud but who cares.

 

If your bike is a "YICS"" you should use the blanking tool to isolate each cylinder from each other.

 

if you have a digital tach, you can tune for max rpm without too much trouble.
these engines have valve overlap so a ega measures exhaust and scavenged a/f mixture

 

not needed.

 

the s tock exhaust has a port on each header pipe

 

I have all the parts of the system now. This photo shows the Innovate O2 sensor/meter, air pump and exhaust fitting (from Motion Pro). I just need some rubber tubing and an enclosure for the O2 meter with input and output air fittings. (and a box to put it all in). I'll probably try to reduce some of that wire length where possible. This innovate setup includes a control box that has serial output to connect to a laptop, which I don't need (at this point).

 

Then all you'll need is someone to steer your bike while you sit on the pillion watching the display, and looking over their shoulder at the throttle and tacho....

 

na, I'm going to thread the small brass tube into the down-pipe bung and test on the center stand.

 

When you are ready, let me know. Maybe I can lend a hand or at least watch!

 

Thanks, I plan to get on CAD later and design up a little bracket to mount the air pump that I can 3d print. Then, design the "test chamber" that the air lines and O2 sensor connect to. Wishing I had a lathe and a mill! I'll 3d print for a prototype, but I assume the heat of the O2 sensor may be too much for the plastic.

 

Well you'll only learn about the idle and progression circuits that way, even buzzing it the needle and main jet will only just be coming on.

 

Hmmm.....I'm pretty sure the engine can run up to redline while on the center stand. It won't be under load, but he'll get a good ballpark refrence.

 

That raises an interesting point. Since this Innovate gauge can do data logging, can I also capture RPM and then correlate engine speed and AF ratio... I'll have to look into that. First, get the thing to work, then worry about a case, then worry about logging and other inputs.

 

Sure it will but the needle will barely be lifted, due to the little air needed for this at no load - and just as importantly, this is a useless part of the speed/load "map", only seen if you buzz the motor on the stand. Hardly any point to it once past 4000 rpm or so.
Datalogging - now this IS useful - but you need an rpm and ideally throttle position to make it useful.

 

Yes you can, but you will likely need a pair of diodes connected to the coil negatives, and it will read double the rpm.
Can you also rig up a potentiometer onto the throttle spindle? OR, a lot easier, a map sensor teed into the petcock feed (or any of the other vacuum ports). You can find one of these at a breaker, 0 to 5V output usually.

 

looks good. It's a little more sophisticated then what I'm trying to do but I'll be interested in what you come up with and compare results.

 

you'll be half way to FI when this is over

 

Hey, I could at least add a servo to adjust the mixture screw for me. (just joking, I don't trust something like that to not bugger up the screw, or a software error drive it hard into the seat)

 

Why would he go to Finland?

 

Nowhere near it, but he will have the means to callibrate it!

 

3D printed a support for the air pump. It purrs nicely when energized and creates suction at a nice level for this application.

The lambda temp could get to 600 degrees (per the spec sheet), so the test chamber must be metal. Plastics I could print melt under 500 degrees. I'm working on a design in CAD. I plan to braze inlet and outlet pipes, and may as well braze the cap. I'm planning a part that can be built on a lathe that the sensor screws into and has a fitting to draw gasses in near the end of the sensor and an outlet (to the air pump) near the base of the sensor, hopefully to give the gasses the best chance of being measured since the sensor has openings at the end. I need to think about condensation and drainage. I'll probably let the air outlet be low and act as a drain as well. The air pump can expel any drops of water along with the air. Keeping the test setup powered on after shutting off the engine would allow water to clear (more or less).

 

GOOD ONE!!

 

i found it. the copper went up the pipe, the rubber hose went about 2 feet into a metal coffee can so any suck-back was still exhaust. worked on my VW, yamaha showed crazy wild swings
and then i quit and went for a ride

 

Is that silicon sealer? If so don't use it, I read somewher it damages lambds sensors (or was that catalytic converters?)

 

i know what you mean, i've heard that too. they must have solved that problem.
this is what is on there
https://www.amazon.com/Permatex-81878-Maximum-Temperature-Silicone/dp/B0002UEOPA

 

I just wired and powered up my lambda controller and gauge. I did the calibration in free-air and the gauge read 22.4. Not sure what's expected there. I also used a Fluke temperature meter to measure temperature at the threads of the sensor, which came out to 150F. That's low enough for PLA which melts around 400F. I may just 3D print a test chamber and see how that holds up.

 

air is about 21% O so that's close. i think if you rev it at all, your going to melt. print two