More dyno discussions: Tailpipe readings vs. internal WB02 comparison
 

So there's been some discussion of the relative merits of trying to tune your car on a dyno when the only option is the tailpipe or 'butt' sniffer.

In fact I had an additional bung welded into my exhaust system because I wanted to have them use a 'real' O2 sensor instead of trying to use the one in the tailpipe but they still insisted on using the sniffer technique. I'd been told that this can be inaccurate and depending on how they attach it it can be out of the flow of the exhaust (against the side) or can fall out, giving you inconsistent readings. It's also after the catalytic converters so that will affect the readings as well.

Suffice it to say you shouldn't accept the tailpipe reading as gospel for tuning.

This week I went to the dyno and logged my WB02 readings at the same time as the dyno operator was logging the tailpipe sniffer. I then did an export from both systems and matched them up as closely as possible by RPM and here are the results:

What you'll see is that there is an average difference of .308 total AFR, but it's not across the board (all leaner or all richer). Up to about 5k RPMs, the tailpipe sniffer shows consistently leaner than the WB02, while above that it shows richer. (The last line shows the 'lift' where the WB02 goes way rich, even while the tailpipe sniffer lags behind, but this line is not included in the average calculation).

It's not even a case where the tailpipe sniffer is just slower or behind the WB02 (which is of course upstream) it just seems to be different.

Now as I mentioned, this test may not be typical because there are way too many variables, but I was just curious about the differences and was surprised by the results and thought you guys might be interested.

Mike

I can add an anecdotal "data" point. Knowing my 90GT (no O2 sensor and no cats) was running rich about a month ago I took it to get the idle mixture adjusted back to spec ( 0.5 -1.5% CO) Final result after adjusting was 0.7% CO at idle (Down from 3.5-4%).

Thursday I was at Hiltons to braze up his Sniffer pipe with wideband OS sensor. Once finished we decided to check the GT idle readings at the tailpipe - seemed to average around 14.4 AFR which would seem to be perhaps a bit leaner than the 0.7% CO reading would suggest.

Were heading to the dyno today. Will see if there is an opportunity to log similar data.

As I think about, I have a wbo2 installed in my car. I also have a "portable" Wbo2 that I can stick up the tail pipe. Both are from innovate.

Might be a fun experiment to hook them up and data log them to see what the realtime variance is.

Tail-pipe sniffers suck.

I think but do not know that wbo2 sensor has more lag but is potentially more accurate if installed after the cats. The logic is that the sensor itself has a small catalytic converter and it has to complete the burn before it can measure the oxygen demand.

.....have a "couple" of beers and get back to us!:cheers:

the widow has a techedge WBO2 and the dyno had a innovate......they both read nearly the same during dyno runs.....however the techedge would do some crazy readouts every once in a while

I do not set mixtures on a Dynojet. They seem to always read my vehicles as being lean, when they are not. Read some very interesting literature about this, which cautioned about setting/using mixtures on a rapidly accelerating dyno...let me see if I can find it.

You have to be somewhat realistic about tools that measure CO, that are not calibrated off of a known sample of CO. They are going to vary.

I have an extra bung just for the dyno guy to use. That way I have my own logged data and his too. This is also handy for when your sensor fails in the middle of tuning!
I trust him to use quality sensors and to change them regularly.

Greg, you have a point about tuning on a rapidly changing dyno. I do know that you can accelerate right through issue spots and not find them until on track in high gear where it has a chance to linger and act up!

One reason why fuel-related tuning is difficult when the engine accelerates fast is that the x-tau correction impacts the fueling. I believe the "acceleration enrichment" parameters on Sharktuner control the x-tau correction. One would have to tune both the steady-state maps and the acceleration enrichment maps the same time, which is difficult.

It makes much more sense to first tune the steady state fuel and ignition to desired settings (on a load dyno such as Dynapack) and only when that is done start tuning the transient enrichment.

That makes sense.

FWIW the idle mixture on my GT was actually set and then part load mixture checked on an eddy current dyno allowing steady state conditions. Even from a settled idle mixture then gentle acceleration up to 2500 -3500 rpms then holding (producing about 60-80 ish rwhp) the tail pipe sniffer took a second or so to settle and maybe 10 seconds to get a near as dammit unvarying reading.

I too did a back to back comparison with the tail pipe reader, and using the bung in the header. however, even that has its flaws, in that it was taken at the right side header, so we are assuming that the other side is equal. with the lag and the shape and level of the readings, it sure looked leaner when measured at the tail piple. spark plug visual checks and tail pipe soot sure confirm that Im in the way rich range with my stroker, vs the old holbert motor. (with the only change being bigger injectors and lower pressure)

anyway, when we compared the tail pipe method and the bung wide band, the tail pipe was leaner and had some lag.

as far as setting mixture i think its a great idea to follow an accelerating engine, but not for the emission stuff, or part throttle, where you have some of the variables in steady state. it would be interesting to see the literature to the contrary.

The sensor should always be as close to the engine as possible. Also there is no burning in the wideband sensor. It is simply comparing exhaust stream oxygen content to ambient external air oxygen levels.

http://wbo2.com/lsu/lsuworks.htm

"In a rich mixture the oxygen ions will combine, on the pump cell's catalytic surface, with fuel to produce water and carbon dioxide. When all the fuel is consumed there will be no free oxygen and the resulting mixture will be at stoich."

"For the controller to work at all, the sensor element must be heated to the correct operational temperature where oxygen ions can sustain the necessary catalytic reactions."

The new series of Dynojets (like the one I use) have Eddy Current Brakes.

Wouldn't the heat of the exhaust gases do this?