How does an A/F Wideband work?

[voidhawk]

I was wondering how an AF meter actually measures the AF ratio in the engine? From what I've read they use an "oxygen sensor" of some type, which converts the amount of oxygen in the exhaust stream to a proportional voltage. This makes sense if you are running lean or stoich, but how would it be able to measure fuel rich conditions since there won't be any oxygen left in the exhaust stream? It seems like the sensor would have no way to measure how rich you are (other than the fact that you are fuel rich).

The reason I've been thinking about this, is that when we test rocket engines at work, we measure oxidizer to fuel ratio (OF) by measuring mass-flow of the fuel and oxygen entering the engine, and taking the ratio. This works, but can be quite the PITA. If there was a way to adopt the "AF sensor" approach and measure the ratio directly, it would be a very useful improvement.

So anybody know what exactly makes a wide-band AF tick?

[alex4nder]

Here's a good summary of different types of o2 sampling strategies (i.e. PID and Innovate's 'direct digital' approach):

http://www.innovatemotorsports.com/r...s/AN19_1.2.pdf

[voidhawk]

Originally Posted by alex4nder

Here's a good summary of different types of o2 sampling strategies (i.e. PID and Innovate's 'direct digital' approach):

http://www.innovatemotorsports.com/r...s/AN19_1.2.pdf

Good find - thanks! Now I understand why oxygen sensors are so expensive, they are pretty complex internally. Still not quite sure how they measure richer than stoich though ...

[djh]

Originally Posted by voidhawk

Good find - thanks! Now I understand why oxygen sensors are so expensive, they are pretty complex internally. Still not quite sure how they measure richer than stoich though ...

A WB has an O2 pump that can drive oxygen in to or out of a measurement cell based on a supplied pump current. The complex electronics control the pump current. The measurement cell has a regular NB (Nernst) cell and some platinum catalyst.

In a rich mixture there's no free O2, so the NB cell generates a high voltage. The controller sends current through the pump, supplying O2 which reacts with the unburned HC and CO (that's why there's a catalyst). When enough O2 has been driven into the measurement cell the NB voltage drops. The controller knows how much O2 was supplied to use up the HC and CO and calculates lambda from that.

In a lean mixture with free O2, the NB generates no (or very low) voltage. The controller sends current in the other direction driving O2 out of the cell. When the NB voltage goes high the controller knows how much O2 was removed. A WB can measure lean lambda more accurately over a wider range than a NB on its own.

The thing that's different about the Innovate WB is the current control electronics, the sensor itself is a standard Bosch or NGK.

Dave

[dubbleugly01]

nice.... I've had the same question for some time. Also explains why the WB's life is shortened due to overly rich conditions.

[voidhawk]

Originally Posted by djh

A WB has an O2 pump that can drive oxygen in to or out of a measurement cell based on a supplied pump current. The complex electronics control the pump current. The measurement cell has a regular NB (Nernst) cell and some platinum catalyst.

In a rich mixture there's no free O2, so the NB cell generates a high voltage. The controller sends current through the pump, supplying O2 which reacts with the unburned HC and CO (that's why there's a catalyst). When enough O2 has been driven into the measurement cell the NB voltage drops. The controller knows how much O2 was supplied to use up the HC and CO and calculates lambda from that.

In a lean mixture with free O2, the NB generates no (or very low) voltage. The controller sends current in the other direction driving O2 out of the cell. When the NB voltage goes high the controller knows how much O2 was removed. A WB can measure lean lambda more accurately over a wider range than a NB on its own.

The thing that's different about the Innovate WB is the current control electronics, the sensor itself is a standard Bosch or NGK.

Dave

you're the man - thanks I'm really interested in trying this in a rocket application, especially with Innovate's digital approach resulting in higher sampling frequencies. I've pinged them about it too, hopefully they'll bite .

[nothere]

"trying this in a rocket application"


do tell

[voidhawk]

Originally Posted by nothere

"trying this in a rocket application"


do tell

100,000+ hp to play with

http://www.airlaunchllc.com/video.htm

[klatinn]

voidhawk,

Pls shoot me an email at Klaus at innovatemotorsports dot com. (written out because of spam bots).

- Klaus

[voidhawk]

Originally Posted by klatinn

voidhawk,

Pls shoot me an email at Klaus at innovatemotorsports dot com. (written out because of spam bots).

- Klaus

you've got mail!

[dubbleugly01]

at least keep us up to date on whether it works or not, this is interesting stuff!

[merkzu]

Originally Posted by djh

A WB has an O2 pump that can drive oxygen in to or out of a measurement cell based on a supplied pump current. The complex electronics control the pump current. The measurement cell has a regular NB (Nernst) cell and some platinum catalyst.

In a rich mixture there's no free O2, so the NB cell generates a high voltage. The controller sends current through the pump, supplying O2 which reacts with the unburned HC and CO (that's why there's a catalyst). When enough O2 has been driven into the measurement cell the NB voltage drops. The controller knows how much O2 was supplied to use up the HC and CO and calculates lambda from that.

In a lean mixture with free O2, the NB generates no (or very low) voltage. The controller sends current in the other direction driving O2 out of the cell. When the NB voltage goes high the controller knows how much O2 was removed. A WB can measure lean lambda more accurately over a wider range than a NB on its own.

The thing that's different about the Innovate WB is the current control electronics, the sensor itself is a standard Bosch or NGK.

Dave

Thanks! I had been wondering this for a long time and never found a good explanation.

[never_finished]

good read, i was about to post a thread.