Wideband in tailpipe on dyno?

I've probable seen 100+ videos of people tuning cars on a chassis dyno.

They always put a wideband in the tailpipe to log the AFR separately.

What is the point of this on a car that already has wideband sensors from the factory?

Wouldn't the reading from the tailpipe have a much higher chance for error?

It helps to confirm what the built-in O2 sensors are reading (i.e. if you have a bad O2, then you can fix it). Also, in may cases it's because the dyno software will only log it's own O2 sensor, so if you want to provide a graph showing torque/power/afr, you have to use the dyno wideband.

You learn when to trust it (i.e. works fine at high flow rates), and when to ignore it (idle conditions). Now some dyno O2 will have a sample pipe that goes much further up the exhaust and allows more useful data, or even a pump to "pull" exhaust gases to an externally mounted sensor.

I've seen some of those venturi style "pumps" in action David, there was one that hooked into the shop air compressor and drew enough air through it that at idle there was a reverse flow into the tailpipe as the exhaust mass wasn't enough to supply the flow demands.

The other thing that wideband in tail pipe helps with is correct AFR reading with regards to ignition timing change affect. When wideband sensor is located very close to exhaust port changing the ignition timing can affect on lambda reading as much as 0.5 AFR depending on advancing or retarding... Placing WBO2 sensor as far as possible eliminates this issue...

There is also the option of using a 4/5 gas exhaust analyser in the tail-pipe, rather than just the basic O₂ lambda sensor.

TBH, I'm surprised they seem to be so uncommon, because they give so much more information.

Hi Gord, 4/5 gas analysers are good in steady state operation, but due to their latency in sampling (I've seen some automotive test units that have a 15+ second sampling response time) are not usable for transient and accelerative testing. Lambda does have it's drawbacks as well, but those are less than those of the 4/5 gas analysers.

Good point, especially as most use inertia dyno's, or acc'n ramping.

How can we use HPTuners to compensate for the end of tailpipe WB latency? It seems that it has a 3-4 sec delay and this is not syncing for tuning accuracy.

Can I use the "cell hit" required, or what do you do?

You will need to build the delay into your tuning process, so make the chance, wait for the transport time, and then see the results. Otherwise, can you install the wideband closer to the exhaust ports?

For HP Tuners, I often remove the OEM 02 sensor (if it's narrowband), and installed my wideband (MoTeC PLM) in the stock location. If you have the chance, adding an additional O2 Bung into your exhaust is a good idea.

Evan,

Getting widebands in a location closer to the engine where you get more accurate data without reversion induced contamination at the exhaust tip, and don't have to deal with as much latency, is more ideal.

In response to the question about using an external wideband when a factory wideband is present, consider factory turbo cars. The factory wideband is often in the exhaust manifold, and not necessarily pressure compensated or able to read accurately during very high exhaust pressure. In this instance it may work great at partial throttle, but having an external unit in a downpipe may help get more accurate data at high engine load.

Mike, what is your view on oxygen sensor AFR difference reading related to ignition timing change when it's placed close to exhaust port? On one hand it reacts quicker comparing to exhaust tip location, on the other hand it reads AFR differently (as much as 0.5 AFR) when ignition timing is changed which isn't very percise I would say...