Fitting an air fuel ratio meter

[Rustinmotion]

Just in the process of fitting a simple narrow band gauge
Made a cardboard template to make sure it fits and it’s possible to remove it ( made that mistake before)
Welding won’t win any prizes but it’s gas tight and solid
I have the signal cables pulled into the cabin so I’m just waiting on the post to deliver the gauge

[Rustinmotion]

Originally Posted by Rustinmotion

Just in the process of fitting a simple narrow band gauge
Made a cardboard template to make sure it fits and it’s possible to remove it ( made that mistake before)
Welding won’t win any prizes but it’s gas tight and solid
I have the signal cables pulled into the cabin so I’m just waiting on the post to deliver the gauge

Just a couple mor photos

[142 Guy]

Narrow band or do you mean wideband? Narrow bands can only do three things:
- sort of tell you you are at 14.7
- tell you that you are above 14.7
- tell you that you are below 14.7

The last two items are what they do best. They cannot tell you how far above or below 14.7 you are. They will tell you whether you are consistently above or below 14.7 and on a carb equipped B18/B20 it will probably show that you are consistently below 14.7 except when you have a trailing throttle condition.

I know that there are narrow band gauges out there. If you have one coming I suggest you return it if possible and spend the extra $ for a wideband. If your objective is to use it for tuning I think you will find the tuning information provided by the narrow band unsatisfactory. If you want it just because, then OK.

I have attached a data log from the B20E in my 142E that I did last August. It is the very tail end of the data log with the engine idling. At the location of the cursor, the engine RPM is 895 (top graph white line). The bottom line has three graphs:

• Yellow is what I am targeting for the AFR - 14.8 at the cursor in this case
• Red is the measured AFR - 14.4 at the cursor
• Purple is the amount of fuel correction being applied to try and get the measured AFR to match the target.

What is germane is that just a moment before the O2 sensor measured 14.4 it was measuring 15.4 and that is fairly common to see that the actual AFR tends to bounce around a bit. I have a D camshaft which is not a particularly nice idling camshaft. At highway speeds with the engine under more load the AFRs are not quite so bouncy, but, it is still there.

I have attached an output curve for a typical narrow band sensor. For this sensor my 14.4 AFR would give a voltage of about 0.65 volts and 15.4 would give about 0.15 volts. Since the sensor's full range of operation is only about 0.05 to 0.85 volts, my 14.4 to 15.4 AFR bounce is going to have your gauge swinging from almost full rich to full lean. That works for digital fuel injection where the sole purpose is to keep the AFR at 14.7. Not so useful for tuning information. As a side note, at the very left of the graph you can see the max operating AFR range during the log session. My lowest AFR was down at 12.1 during the cold start and my highest AFR hit 19.8 - probably when I was engine braking coming off the highway. A narrow band sensor will not be able to distinguish those types of events.

[Rustinmotion]

Originally Posted by 142 Guy

Narrow band or do you mean wideband? Narrow bands can only do three things:
- sort of tell you you are at 14.7
- tell you that you are above 14.7
- tell you that you are below 14.7

The last two items are what they do best. They cannot tell you how far above or below 14.7 you are. They will tell you whether you are consistently above or below 14.7 and on a carb equipped B18/B20 it will probably show that you are consistently below 14.7 except when you have a trailing throttle condition.

I know that there are narrow band gauges out there. If you have one coming I suggest you return it if possible and spend the extra $ for a wideband. If your objective is to use it for tuning I think you will find the tuning information provided by the narrow band unsatisfactory. If you want it just because, then OK.

I have attached a data log from the B20E in my 142E that I did last August. It is the very tail end of the data log with the engine idling. At the location of the cursor, the engine RPM is 895 (top graph white line). The bottom line has three graphs:

• Yellow is what I am targeting for the AFR - 14.8 at the cursor in this case
• Red is the measured AFR - 14.4 at the cursor
• Purple is the amount of fuel correction being applied to try and get the measured AFR to match the target.

What is germane is that just a moment before the O2 sensor measured 14.4 it was measuring 15.4 and that is fairly common to see that the actual AFR tends to bounce around a bit. I have a D camshaft which is not a particularly nice idling camshaft. At highway speeds with the engine under more load the AFRs are not quite so bouncy, but, it is still there.

I have attached an output curve for a typical narrow band sensor. For this sensor my 14.4 AFR would give a voltage of about 0.65 volts and 15.4 would give about 0.15 volts. Since the sensor's full range of operation is only about 0.05 to 0.85 volts, my 14.4 to 15.4 AFR bounce is going to have your gauge swinging from almost full rich to full lean. That works for digital fuel injection where the sole purpose is to keep the AFR at 14.7. Not so useful for tuning information. As a side note, at the very left of the graph you can see the max operating AFR range during the log session. My lowest AFR was down at 12.1 during the cold start and my highest AFR hit 19.8 - probably when I was engine braking coming off the highway. A narrow band sensor will not be able to distinguish those types of events.

Fully understand the difference and want to try with narrow band as my aim is to set up the carbs for a good idle and know I’m not running lean under load and cruising speed
Once I’m happy with that the gauge is redundant so not keen to spend £200-£300 on a full wideband , narrow band is good enough to trim fuel maps on a modern car
This is what I’m waiting for in the photo

[Rustinmotion]

https://youtu.be/iU0TkgErMyA

[Ron Kwas]

RiM;

I haven't yet looked at the linked vid, but 142Guy's info is spot-on...and he's working with an injected system...you're talking about a carbed system where mixture control is even less tightly controlled...I expect you'll find the indication with a narrow band sensor "bouncing between the guardrails" so much so, that it may not be usable to any great extent! (Sorry to be the bearer of bad news!) ...but we will certainly be following your results with interest!

Cheers

[Rustinmotion]

Originally Posted by Ron Kwas

RiM;

I haven't yet looked at the linked vid, but 142Guy's info is spot-on...and he's working with an injected system...you're talking about a carbed system where mixture control is even less tightly controlled...I expect you'll find the indication with a narrow band sensor "bouncing between the guardrails" so much so, that it may not be usable to any great extent! (Sorry to be the bearer of bad news!) ...but we will certainly be following your results with interest!

Cheers

Watch the video, I’m only guessing but I think there is significant damping of the signal to give stable information, true it’s never going to be as good as full wideband but I think it will do the job, if I didn’t already have a narrow band sensor I may have gone for the Zietronix wideband if you buy it without a gauge and just use a laptop it’s only a little more expensive than the one I have if I include the cost of the O2 sensor
I’m struggling to get an even idle and the right plug colour at the moment I can chose one, I’m hoping this will at least give me confidence I’m not doing any damage

[142 Guy]

Fair enough. You seem to be fully aware of the issues. My comments were in part based upon the fact that the price differential between a narrow band and a wideband system is not what it used to be.

https://www.summitracing.com/int/sea...tOrder=Default

If you are looking for a temporary installation for testing, consider the 14point7 SLC Free 2.
https://www.14point7.com/products/si...troller-free-2

It is only $95 US$ including a 4.9 sensor which is about £80 + shipping (and duties?). I use the previous version of this controller as back up whenever I want to do a sanity check on my AFR readings or determine whether I have a sensor / controller problem. You do need to make your own cables; but, I find that an advantage because I always extend the cables anyway because I do not want the electronics in the engine compartment.

As a related observation, you sensor positioning is not ideal (appears to be horizontal). However, from my 142 experience getting an ideal mounting position (sensor mounted 10:00 - 02:00 o'clock) can be difficult / impossible. I assume that your narrow band sensor is heated? If so, make sure that the heater is not turned on until after the engine is running. If you heat up the sensor prior to starting the engine you risk getting a blast of condensation hitting a hot sensor element which can fracture the element. Narrow bands do not run as hot as wide bands so the damage may not materialize as quickly. A blast of moisture can kill a pre heated wide band tout de suite. Been there, done that, got the Tee shirt.

[Rustinmotion]

Originally Posted by 142 Guy

Fair enough. You seem to be fully aware of the issues. My comments were in part based upon the fact that the price differential between a narrow band and a wideband system is not what it used to be.

https://www.summitracing.com/int/sea...tOrder=Default

If you are looking for a temporary installation for testing, consider the 14point7 SLC Free 2.
https://www.14point7.com/products/si...troller-free-2

It is only $95 US$ including a 4.9 sensor which is about £80 + shipping (and duties?). I use the previous version of this controller as back up whenever I want to do a sanity check on my AFR readings or determine whether I have a sensor / controller problem. You do need to make your own cables; but, I find that an advantage because I always extend the cables anyway because I do not want the electronics in the engine compartment.

As a related observation, you sensor positioning is not ideal (appears to be horizontal). However, from my 142 experience getting an ideal mounting position (sensor mounted 10:00 - 02:00 o'clock) can be difficult / impossible. I assume that your narrow band sensor is heated? If so, make sure that the heater is not turned on until after the engine is running. If you heat up the sensor prior to starting the engine you risk getting a blast of condensation hitting a hot sensor element which can fracture the element. Narrow bands do not run as hot as wide bands so the damage may not materialize as quickly. A blast of moisture can kill a pre heated wide band tout de suite. Been there, done that, got the Tee shirt.

The photo is not representative of reality it’s around 5 degrees from horizontal perhaps a little more it’s to promote condensation to drip off the end of the probe
I’m not running the heater il just let the exhaust warm it up, not too bothered about monitoring cold start conditions, I used to run a narrow band and analog gauge years back on another car back when O2 sensors had one wire, it just took a while to warm up
I like the look of the SLC but cautious on the “required assembly” me and pc board soldering don’t get on well

[142 Guy]

If you are not using the heater then the clocking of the sensor and fracturing of the sensor from moisture impact is probably not an issue .

If at some future point you do want to consider the SLC Free, on my early version of the SLC Free the ICs are socketed. The soldering was limited to through hole devices such as resistors, capacitors and IC sockets so fairly easy as long as you use a good soldering iron. Since the designer has switched to surface mount components on the new board which are only user friendly for keeners, I expect that the board is largely assembled except for soldering on the connector plugs and fabricating the cables.