Narrow and Broad band O2 sensors and open vs closed loop systems

[lmcchesney]

Sorry guys, but I've been doing some reading again. Dave, you and I have had some of this discussion before and I would respect your continued thoughts on this setup. The crux of the matter requires understanding of the circuit utilized to turn the lamba analog data into useful information. Well, here goes.
O2 sensors in a distal position provides indirect input for the estimation of the A/F ratio seen by the engine. This information is useful in defining a lean or rich mixture seen by the sensor for the most recent peroid of time. Broad band O2 sensors are capable of producing a wide range of output voltages which can be inturpeted for measuring the ratio. In fact, the output from broad band sensors can be feed into the Electronic Control Unit several times a second. This allows for real time adjustment of the A/F mixture. Thus output from the O2 sensor is recieved as input by the ECU which inturn alters injector time/volume to preprogramed mixture. This feedback/adjustment defines a closed loop system.
Narrow band O2 sensors also read A/F ratios. However, A/F mixtures ranging from 13.7:1 to 14.4:1 results in a change in voltage output from lowest to highest (0.1v to 1.0v).
If you use a analog voltmeter or equalivernt A/F gague, the response rates are such that your readings bounce requiring seat of the pants estimate as to where the mean is.
However, with the use of circuitry involving a IC unit, the voltage output can be lengthed to a series of LED's.
My question, can not a narrow band O2 sensor be utilized to monitor A/F ratios throughout the power band. Noting the results, provide information for enrichment/leaning in the adjustable ranges of D-jet. ie, idle, partial throttle and WOT.
Cannot in dash monitoring with such an A/F circuit be helpful to identify serious alterations in the engine and prevent damage from excessively lean running?
Cannot the useful information from such an A/F monitor be inproved by silmiltaneous monitoring of CHT/oil temp?
I still do not understand the advantages of a broad band O2 sensor when used in a open loop system.
I cannot also see the significant advantage of a heated O2 sensor for monitoring the A/F ratios during that first 3 minutes.
Thanks,
L. McChesney

[John Rogers]

I might be able to add some insight? I built one of the LED meters as per the Pelican article several years ago and use a Bosch a/f sensor and it is very sensitive and has worked great in our not stock 1987 930. I also use one from CB Performance in my race car and it uses the same sensor and is very sensitive. Nice to have when you are foot to the floor around the CA Speedway and it takes the worry out of going lean.

[Dave at Pelican Parts]

The thing is that the NBO2 sensor will not tell you just how rich or just how lean your engine is running. It will tell you "rich", it will tell you "lean", but the distance away from stoich is as much a factor of the exact temperature of the sensor at that moment as it is the O2 content of the exhaust. So your gauge can be reading a nice safe 13:1, when in fact your mixture is 14.6:1 (close to stoich) or 11:1 (quite rich).

Our engines like to run rich. If they ran near stoich, the NBO2 would be a decent choice, as we could fiddle with things until it read rich/lean/rich/lean the way they do in modern cars. But we want it on the rich end of things, which means we need to know what the mixture really is. The NBO2 can give us an instant check on "holy crap, it's way too lean"--but it won't tell us if it's too rich, or is only a little too lean. It is built to operate right around stoich, which is leaner than our engines like.

I hope this makes sense. That margarita was a lot stronger than I thought!! :hic:

--DD

[lmcchesney]

John and Dave, Thanks.
Dave, I recall your statement from before but I still do not understand it in light of other postings.
http://www.buckeyetriumphs.org/technical/Carbs/AFMonitor/AFMonitor.htm
The mV A/F ration is expotential, granted, and analog volt meters cannot accurately display the result, but from my readings the ratio is a fixed ratio.
LED Voltage Range Lambda Air/Fuel Ratio Output Torque Observations

>945mV <0.80 <12 <98% Too rich, power loss
900-945 mV ~0.85 ~12.5 100% Maximum power
855-900 mV ~0.90 ~13.2 99% Good operation
810-855 mV ~0.93 13.8 98% Good operation
540-810 mV ~0.97 14.3 98% Good operation
225-540 mV ~0.98 14.4 97% Good operation
180-225 mV ~1.0 14.7 97% Good operation
135-180 mV ~1.02 ~15 96% Good operation
90-135 mV ~1.04 ~15.3 95% Maximum economy
45-90 mV 1.05-1.25 15.4-18 74-94% Pretty lean
All out <45 mV >1.25 >18 ? Too lean, misfires

I also agree in constructing the monitoring circuit, you have to calibrate to the sensor to define what stoichiometric value is represented by midline of the meter/led.

Your statement regarding the varibility of ratio depending on operating temp, again I understand the variation util reaching operating temp of 300°F, and I must say, I cannot recall the temp seen in the exhaust and as to wether it varies much. However, I understand that the temp vairation could be resolved by a heated O2 sensor $17 vs $25.

Could you expond a little more on WBO2 sensors. The cost of new is comming down but still out of the NBO2 sensors. However, I believe several newer models of auto utilize the WBO2 sensors and could be retrived from the salvage yard.
Thanks,
L. McChesney

[Dave at Pelican Parts]

I have seen a graph of the response of an NBO2 sensor similar to the one you posted above. The difference was that it had several curves plotted, showing the response at different temperatures. I don't remember them as being drastically different from each other, either. But the response (voltage per lambda) was very different. One looked like your curve, one was much closer to flat, and on the other extreme the graph was almost a step-input graph (flat, straight up, then flat again).

Heating the sensors gets them into the range where they work, but their response keeps changing after that. I don't think any of the heaters is set up to be able to keep the sensor at a temperature higher than the hottest exhaust it would see--because that is what it would have to do in order to keep the sensor at the exact same temperature all the time.

So I feel that the NBO2 is a decent "sanity check", but is not going to be much of a tuning aid for us.

WBO2 sensors are very much more complicated. There is "excitation circuitry" and an "ion pump" of some sort, but those are just terms that I've heard flung around in their discussion. I don't really understand what makes them work... You may get more joy on that by doing a web search, because obviously some people do or there wouldn't be any DIY WBO2 kits. (Which are neat, but use older more expensive sensors that wind up costing most of what the new WBO2 gauge setups cost...)

--DD

[lmcchesney]

Thanks again Dave.
Somewhere in the back of my head, I recall someone saying the NK WBO2 sensor in the Honda'a was usable for our intrest Is that what you remember? I wonder if the is a listing of WBO2 sensors from which we can find salvage sources.
L. McChesney

[Dave at Pelican Parts]

The ones from the Civic VX are probably what you're thinking about, and they're $$$ even from the junkyard usually. Plus they need a bunch of circuitry to run them.

Google on "diy wbo2" and I think you'll find at least some of the info. But as I said, it turns out to be only a little less $$ than buying the setup that was pointed out to you on the other board.

--DD

[mike_the_man]

I've been following this thread with interest. I'm interested in checking my air/fuel mixture too. Right now, all I'm interested in is figuring out if my car is running too rich or too lean. So, can I just grab a NBO2 sensor from the junkyard, hook it up to my digital multimeter, stick it in the exhaust pipe, and get some idea of whether or not I'm running rich or lean?

I noticed that there was talk of an analog multimeter. Would a digital work? There are 3 wires from a NBO2 sensor. The 2 white ones are hooked to power and ground to heat it, and the black one would be hooked to the multimeter. Do I just ground out the other multimeter lead? (I think that would make sense)

Thanks,

[pbanders]

It's more complex than simple measurements with O2 sensors. The narrowband sensor was designed for one thing - keeping the car at the stoichometric point. If you want to know with some degree of accuracy what your idle CO is, you need a shop quality analyzer. Most people would be better off simply driving to a nearby garage and getting them to measure it for you. If you want loaded measurements, you should be going to a dyno shop with a 4-gas analyzer.

[Dave at Pelican Parts]

Yes, Mike--that's what you do. I won't say that it is completely useless, as I think it can provide a "sanity check" of sorts, but the slow response time and the extreme unreliability as you get away from stoichiometric (~14.7:1) turns it into more of an "idiot light" than a real gauge.

E.g., when the oil light comes on you know you are already in trouble--the question is "how much"? And when the gauge goes lean, you know you are already in trouble... (Assuming it is working correctly.)

--DD

[Jim Smolka]

Also, from what others have stated, 911 motors generate the most HP around 13-13.5 to 1 AFR.

[jkeyzer]

Don't get a wideband kit that uses the NTK sensor. There is another family of kits that use the Bosch LSU sensor, which is readily available from VW parts sources for as little as $25 NEW.

The Tech Edge board is one. www.wbo2.com

[mike_the_man]

I'm not after an exact number, and I'm not going to try and use it to tune my car, but I think I'm running really rich, and I'm not 100% sure, so hopefully this will give me some idea. Once I get things tuned a little better, then I will be taking it to the local Porsche shop to get the idle mixture set, etc.

Thanks,

[Dave at Pelican Parts]

The NBO2 will not tell you if you're running "really rich". Or rather, it will--but it will tell you the exact same thing if you're only running "a little rich"--which is where our cars like to run. At least some time.

It is not reliable enough to use for anything other than telling you that you're at 14.7:1, which is leaner than a 914 should be run IMHO.

--DD

[lmcchesney]

Update information.
First, thanks to all for the information and contribution.
Use of a narrow band O2 sensor can be used over a specific A/F range, which is O2 sensor dependent. Some even will measure A/F from 12.5 to 15.3. Use of an IC you can adjust the exponential voltage out put to a more linear output. If the voltage is read through an analog meter, despite the conversion of the exponential curve, it still is difficult to read. From the information I have gathered, the A/F reading via analog gauge results in a wide fluctuation (rich, lean, and rich, lean) of which you have to estimate the average of the fluctuation to correlate to an A/F ratio. If read via a LED system, the led's bounce from rich, lean as well requiring the estimation of the average. Thus, if used for precise tuning, the estimated average is not precise enough for fine tuning of the A/F, unless you will accept a less precise quality of the tuning.
Wide band O2 sensors are capable of measuring an increased range of A/F over the same voltage output, but more importantly, when connected to a metering/gauge system, provides a smooth display and thus allowing for a more accurate average of the fluctuation and thus more accurate A/F.
The crux was obtaining a wide band O2 sensor (WBO2), which has been about $200 in cost. Bosch produces a LSU4 WBO2 which is heated (allowing for a more stable environment to measure the O2 gas), which is used in the 2000 VW Gulf GTI. Doing some shopping I have found two sources for these O2 sensors for cost of $30-36. I have had a very difficult time confirming via Bosch that these units are the LSU4 units. I was able to confirm that they are Bosch USA # 17014 which does cross reference with Bosch LSU4 unit. I was worried in that depending on how it is listed, other O2 sensors for that same model of Golf was about $200.
Now the meter. PP has a tech article to make a LED meter. Cost for Radio Shack parts about $15. I was able to pick up a Sunpro A/F meter, which is a LED meter from the FLAPS for $25. Thus, a complete WBO2 A/F unit could be put together for $60 and require the wiring from the WBO2 sensor into the dash.
I am still trying to fully understand the WBO2 circuit. It seems there is a thermocoupler to the heater of the o2 sensor to maintain constant temp at varied EGT’s. There is also the ion pump to adjust for the quantity of HC and O2. Additionally, other circuits have RPM inputS.
I found this article: http://www.buckeyetriumphs.org/technical/Carbs/AFMonitor/AFMonitor.htm
Seems like an EE student using a NBO2 sensor and meter to tune his four carbs on his TR7. I have emailed him for an update.

[ken180]

The WB sensor will not work with the same meter the NB sensor uses. The WB sensor needs contoller and interperter circuitry.

Ken

[klatinn]

Hi,

Here's some helpful info, bottom of the newsletter.

WB/NB O2 info

Regards

[ratwell]

From the information I have gathered, the A/F reading via analog gauge results in a wide fluctuation (rich, lean, and rich, lean) of which you have to estimate the average of the fluctuation to correlate to an A/F ratio. If read via a LED system, the led's bounce from rich, lean as well requiring the estimation of the average.

Like Dave said, the hysteresis curve that the narrow band sensors produce makes it basically useless except for adjusting the mixture slightly around stoich AFTER it's been set properly using a wide band type sensor. Since the optimum setting for a VW engine usually isn't stoich the usefullness of this for tuning is severely limited.

I have had a very difficult time confirming via Bosch that these units are the LSU4 units.

The tech edge website has all the part numbers. http://www.techedge.com.au/

Thus, a complete WBO2 A/F unit could be put together for $60 and require the wiring from the WBO2 sensor into the dash.

I wish that were true. The circuitry is quite complicated and you either have to build a kit or buy a prebuilt kit.

[pbanders]

RE: Ratwell's comments...

By the time you're done futzing with building a WBO2 unit, you could have bought a calibrated 2-gas IR dispersive tester from "drsmogtest" on Ebay. Not only would you have a unit that you were sure of its range, calibration, and accuracy, but you'd be able to measure HC in addition to CO. By being able to monitor HC independently, you can make certain you can pass smog HC levels (e.g. know exactly how far you have to loosen the valves, or retard the timing to pass standards, especially important if you have a non-OEM cam).

[lapuwali]

This all assumes the whole point is to just pass a smog test. Unless you're buying a wheel dyno, too, there's no way you can check the mixture under load with such a analyser, which makes genuine tuning very difficult.

The WBO2 stuff is pretty interesting, and the price is only going to get better. Four years ago, the sensors alone were $400, and a full sensor/meter package was in the $900 range. Now you can get an off-the-shelf unit with datalogging from Innovate for $350, including a $25 sensor. Once the Megasquirt stuff is shaken out (6-12 months, I'd say), you'll be able to build one yourself for about $50-100, including sensor.

No better load testing than driving the thing.