Misc: Everything Else Air Fuel Gauge 02 Sensor Lambda Sensor

Cyborg

Well Known and Active Forum User
VOC Member
In my case, its not that I care about the bung being visible. I’d like the second (used) pipe because I’m starting with a new flatslide carburetor and I’m guessing on where to start with the jetting. I don’t want to turn my new chrome pipe completely blue.
When I put a bung on the new pipe I can use the TIG.

When shopping for a sensor and gauge, instead of choosing to use an O2 sensor, I chose a wide band A/F sensor, because I believe it will be better for my situation.
 
Last edited:

Magnetoman

Well Known and Active Forum User
VOC Member
The sensor is hooked up to a mini digital voltmeter that tells me 0.5 V is perfect, up to 0.7 V is a bit rich but OK, below 0.5 V is too lean.
An oxygen sensor of the type you have (now known as a "narrow-band" oxygen sensor") has a sharp transition in its output voltage at stoichiometric 14.7 AFR. For air pollution purposes on a car that's perfect because it can be incorporated in a feedback loop to keep unburned gasoline in the exhaust at a minimum. However, for our purposes it's a lot less than ideal.

Maximum power from engines happens around 12-13 AFR, very far from 14.7. While a narrow band sensor still provides a signal when off stoichiometry, that signals is nonlinear and very temperature dependent (as opposed to the sharp transition at 14.7). As a result, a voltmeter reading of 0.5 tells you the AFR is 14.7, but any voltages above or below that only tell you the mixture is richer or leaner than 14.7, not how much richer or leaner.

Having a wide-band sensor is like when you were a child playing hide-and-seek, where you were constantly given clues like "getting colder, very cold, warmer, very warm, hot" to help you find the object. In contrast, a narrow-band sensor wouldn't give you any clues until you were very near the object and then only would tell you "hot" (or "cold" if you were close but had just passed it). To stretch this a bit further, a narrow-band sensor will help you find the 14.7 prize, but a wide-band sensor will help you find the continuum of more-valuable 12-13 prizes.

Further on the subject of sensor location, following is another excerpt from the Innovate LM-1 manual:

27067
 

oexing

Well Known and Active Forum User
VOC Member
You are quite right, the voltage reading says little about exactly what AFR I got. BUT whenever I change one jet or set the needle higher or lower I can see how sensitive the reading is. With yout AFR reading you will not be able to say exactly what jet you have to set so it is trial or error just the same. You will test jets and other components for seeing changes - and so do I . I don´t see much of time saving with one setup or the other - but then what do I know , just messing around with old machinery. And I am sure, sooner or later you will just accept a good compromise else you´d keep testing forever, no end.

Vic
 

Magnetoman

Well Known and Active Forum User
VOC Member
With yout AFR reading you will not be able to say exactly what jet you have to set so it is trial or error just the same.
That's not quite accurate. If the AFR reading were, say, 11.0 at full throttle with a 300 jet, and I wanted it to be 12.0, the engine would need 11/12 91.5% of the fuel. That new amount of fuel would be supplied by a 275 jet (0.915x300=275) so I would know before my next run that by swapping for either a 270 or 280 I would get 12.2 or 11.8.

If when pulling away there was a slight hesitation that felt like it might be too rich, is the cause the cutaway or the pilot circuit? An AFR meter would help address that faster than simple trial and error. Knowing the AFR from the pilot circuit at idle would let me know if it already was fairly lean, or if I could lean it somewhat. If the latter, I could set it to have a new, lean, AFR and see if that cured the hesitation. If it didn't, then I'd know the cutaway was the culprit.

Bouncing down the road things don't necessarily work to many significant digits of precision, and not all Amals patiently wait while you adjust their settings, but the point is that having actual AFR readings certainly reduces the amount of trial and error.
 

chankly bore

Well Known and Active Forum User
Non-VOC Member
Ah, five pages of diverting information. In the "good old days" over a century ago, "Wizard" O'Donovan tuned his Brooklands bikes by noting the flame colour from a stub exhaust pipe: Mind you, in those far-off times Petrol was Petrol, Men were Men, and women kept respectfully quiet about their good fortune.
 

Cyborg

Well Known and Active Forum User
VOC Member
Magnetoman..... thought of you while playing with this. Pages worth and wireless to boot. All that info, but no EGT. There was a reading for the cat, but it doesn't actually have a sensor located in it. I assume the ECU figures that out using the voltage signals from the A/F sensor heaters fore and aft of the convertor. Got kind of a funny look when I asked where the F/P reading was. Apparently thats being added now that they are going to much higher fuel pressure. Would have made for a good troubleshooting tool. BTY the cat temp was 3 hundred and something C, much lower than I expected. Not a lot of heat being wasted these days.

A little off topic, so I'll delete this and the photos at some point in the future. Apologies for the crappy photos. 0702 shows the lambda A/F 0697 shows cat temp.
 

Attachments

  • IMG_0697.JPG
    IMG_0697.JPG
    58.9 KB · Views: 7
  • IMG_0699.JPG
    IMG_0699.JPG
    54.3 KB · Views: 5
  • IMG_0700.JPG
    IMG_0700.JPG
    57.6 KB · Views: 4
  • IMG_0702.JPG
    IMG_0702.JPG
    54.5 KB · Views: 5
  • IMG_0703.JPG
    IMG_0703.JPG
    58 KB · Views: 6
Last edited:

Magnetoman

Well Known and Active Forum User
VOC Member
All that info, but no EGT.
For some reason this reminds me of a theorist who proposed collaborating on an experiment some years ago because I was in a position to make samples that would be useful for a theory he had developed. I made a few samples and their properties were fit by his theory, so he asked for a few more samples to fill in between the other data points. I happened to be at a conference about that time and mentioned my collaboration to a friend who knew my collaborator well. My friend laughed and said the experiment would never be finished because no matter how many data points I gave him, he always would want more to fill in between those. My friend turned out to be right.

You just posted 5 pages filled with data. But, are you thrilled with all the data you do have? Noooo. You're upset by the one piece of data you don't have.
 

Magnetoman

Well Known and Active Forum User
VOC Member
For what it's worth, the photograph shows the Gold Star's current "dashboard." The white plastic taped to the face of the LM-1 is there to keep the nylon strap from activating the buttons, and the wire ties are enough to keep the wires from flapping in the breeze. All of the cables are longer than they need to be (hence the excess wrapped around the tach and steering damper), but I hesitate to shorten them because it's a lot easier to deal with cables that are too long than ones that are too short.

The analog gauge taped to the tach should go from 1-2 V for AFR 10-20 but as I had previously determined with a power supply and voltmeter it is a bit off at end of the range. The bike wasn't running when I took the photograph so with the AFR>20 the meter should be pegged at 2.0 V rather than 1.9 V, but this imprecision doesn't matter since ~10-14 V is the only range of interest. Anyway, whether having this analog display is useful remains to be seen once I take it on the road. I've been under the weather for a few days but hope to be fully recovered in another day or two, at which time the jetting shall resume.

Competition_dashboard.jpg
 

Magnetoman

Well Known and Active Forum User
VOC Member
By accident I discovered an undocumented "feature" of my Innovate LM-1. On the last run a week ago the display had gone blank for a few seconds but it continued recording. I checked the battery and it still had 30% capacity but I replaced it anyway with one that tested 100%. Along with that undocumented feature of display blanking with a low battery, today a 16-segment bar graph appeared across the bottom of the screen which is a documented, but forgotten-by-me, feature. Apparently, it needs full battery power to be activated.

The bar graph provides an analog display of the mixture, with the middle at stoichometric λ=1 (AFR=14.7 for gasoline). Although λ is a "better" way of expressing mixture strength, especially for mixed fuels, I prefer to think in terms of AFR. In this case the display goes from lean >19.4 at the left to rich <10 at the right, so the desired ~12-13 for maximum power will have ~3/4 of the bars displayed. This means the 0-2 V gauge I added is superfluous.

After the work I had done to it, today the bike was a bit rich at full throttle, but otherwise behaved way better than previously. Actually, other than at full throttle, if not for the damn AFR gauge I might think the jetting was nearly perfect.

It ran well as I left the neighborhood with no hesitation on acceleration so the present #3.33 cutaway seems good. The first graph shows the AFR was ~13 (purple curve) at a little over 1/8 throttle (the red curve is the throttle opening, with full throttle 1.5 V).

The second graph shows that at ~1/3 throttle the AFR was ~14.5 so perhaps the needle could stand to be raised one notch (it's currently on the top notch). However, I won't alter it until I get the main jet worked out.

Once I reached the larger road there's a ~1/4-mile uphill stretch where if I time the traffic right I can give a bike full throttle for a few seconds for a crude check of the main jet. The bike stumbled from a too-rich mixture and as the third graph shows during that interval where I held full throttle for 3.5 sec. the AFR dropped to a very rich 9.2.

The bike idles well, but so far I've been too chicken to rely on that so I blip the throttle regularly when at the one stoplight on my jetting route. Also, I've become so reliant on the DocZ that I'm not sure I would even remember how to kick start it should it die at that light.

AFR1_11May19_3.jpgAFR1_11May19_2.jpgAFR1_11May19_1.jpg
 
Top