AFR Measurement

I almost started to hijack this thread http://www.accessnorton.com/new-head-running-issues-t7760-15.html, but then thought better of it. I have an LM-1 and have been struggling with the best way to get a measurement (without welding a bung on my pipes - although if anyone has a set of junk pipes with no cross over, that might be interesting). I started by using the sniffer that Innnovate sells

but got funky measurements which I attributed to ambient air intrusion. I then attached a 18" or so section of copper tubing to the inlet to try to sample further upstream.

Better, but I still wasn't sure I was getting anything useful. I then read a nice article on the Daytona Sensors site about their sniffer recommendations (see Broken link removed) and tried to copy that.

I think the readings on this setup are better. However, since I've never used a AFR Meter before - I really don't know what I'm supposed to see. So can anyone render an opinion on what seems like the best sniffer set up? Jim?

FYI - I wanted to make this setup work for on the road use (as I don't have a dyno). I'm also logging throttle position via hacked potentiometer and RPM too. When I get this sorted out, I think it's going to be killer. I can set it up on the bike in about 20 mins, so it's easy to switch around.
.

David

Dave,

if anyone has a set of junk pipes with no cross over, that might be interesting)

Click to expand...

Yep I do in pair of stock 1 3/8" 750 headers, Oxy sensor bung and also EGT needle probe holes. Its very informative to watch the CHT and EGT needles respond to changing conditions. i never hooked up the 1 wire Oxy sensor, after I found out that these narrow band sensors basically only read well at WOT. You are welcome to try my old pipes and do surgery to fit the sniffer. My CHT and EGT gauges got trashed in a double upside down loading event with a broken axle. Only the headers survived.

I'll have to study up on wideband but think they will read well once up to temp and may not need the heater element except for cold starts like a CPU controlled car. If bike system can't keep up with the heater on lower rpm, might just strap on extra battery to power it long enough to get your tuning set on the milder conditions. This is how Jim Comstock does his EFI maps, takes a measuring run, then removes the excess devices and runs around keeping eye on CHT, Cylinder temp and EGT. Jim put a 6" long temp probe deep between the bores and has bolted/epoxied CHT probe between the intake manifold. Off to read up on gas sniffers

Highest reading on topped out WOT in summer 95' F - CHT 425' , EGT 1375' F.
EGT changes by the second, CHT by the minute. A vacuum gauge is also entertaining to watch.

I have tried several sniffer setups like it appears you have. The only one I have had success with is one with a long sniffer tube but I had to connect a vacuum pump at the sensor end to draw the gasses through the tube. It works well on the dyno but is not vary practical to mount on the bike. However I have seen commercial setups that have a pump mounted in a box with the sensor. They are usually mounted to the bumper of a car and the tube sticks into the exhaust pipe a couple foot to get a good reading. They use a 12 volt turbine type pump to draw the gasses up the tube. It could be possible to make something like this for a bike but not very simple. The venturi type systems I have tried work fine on a car motor but with only one or two cylinders the pulsing in the exhaust was too heavy to work without air messing up the readings.

I finally welded a bung into a short tube on the side of an old exhaust pipe that I install if I want to get readings on the road. It works well but it's a pain to install.

Here is a picture of the sniffer I use on the dyno. It is the only sniffer I have come up with that gives me readings that are comparable to mounting the sensor in the pipe. Jim

Napa, did you get better results with the other tail piece, I thought I was being technical with a bit of tape around the throttle

I seemed to get much better results with the second version. However since I had nothing as a reference I couldn't really decide if what I was seeing as far as AFR numbers were to be believed. The numbers seemed rich to me. I ended up jetting by feel, mileage, plug chops (although I don't really trust my ability there) and I put this project away until I could get better info. I think what I need to do now is use a welded bung on an old pipe, figure out what is reality and then go back and see if the tailpipe version is good enough. I'd rather not have to swap in pipes to check these measurements. Jim's info really makes me wonder if I'll get there. BTW here's how I did the throttle measurements
.
Its just a potentiometer with a 5v ref in and the output goes to the LM-1. Held on by the mirror nut and there's a collar around the grip. I need to redesign it however as it doesn't work as well as I'd hoped (too much slop in the gear - which I made from a Lego part...)

When I made a unit like the Daytona sensors unit, it gave good readings part of the time. Idle readings were good but as the speed went up I would get no readings at certain rpms or I would get sensor timing errors and the LM1 would lock up. The readings I got also lagged behind the readings I got from the sensor in the pipe and seemed to be more of an average reading. If I would hold a steady speed and throttle setting for 10 or 15 seconds then the reading I got would be correct except at certain speeds where I would get no reading at all.

I have been looking for a small 12 volt turbine style pump so I could mount the unit on a bike. I tried a turbine fuel pump and it worked for a short time and then the motor died with no cooling as expected. Any ideas on the source for a pump? Jim

If you are seeing rich readings they are probably correct. It's usually the lean readings that are suspect. Most Nortons I have tested are very rich especially with Amal carbs. That is why people get away without using the enrichener slides. They seem to run pretty good that way but it sure is tough on cylinder walls and spark plugs. More power will be found by leaning them out a bit but it is easy to create overly lean areas that can cause damage. Jim

Jim, what do you think the flow rate/volume needs to be with a pump? There are probably thousands of pumps, but I have no idea what the criteria should be. Is the one in the pic you posted like this one? http://cgi.ebay.com/Vacuum-Pump-Thomas-2107VD20-/150508633552?pt=BI_Pumps&hash=item230b037dd0? It seems to be rated at Vacuum Flow Rate .7 cfm 20.4 l/min

napanorton said:

Jim, what do you think the flow rate/volume needs to be with a pump? There are probably thousands of pumps, but I have no idea what the criteria should be. Is the one in the pic you posted like this one? http://cgi.ebay.com/Vacuum-Pump-Thomas-2107VD20-/150508633552?pt=BI_Pumps&hash=item230b037dd0? It seems to be rated at Vacuum Flow Rate .7 cfm 20.4 l/min

Click to expand...

That looks like the pump I am using. Mine is from an old Sun Machine gas analyzer. The flow seems about right. Now it just needs to be smaller and run on 12 volts. The small turbine vane pumps I have found all have plastic impellers that don't like the heat. Jim

I try and follow some of these technical threads once in awhile if for nothing to pick up a few tid bits of info, but could someone tell me what the goal is for all the work trying to set up an analyzer on a street bike? I mean I can see if you are using it to get the most out of a track bike or if you're just a tech nerd, but why the extreme for a street bike and street riding conditions? I set mine up per the 40 year old manual and have not had any significant engine issues.

For me its definitely "tech nerd". There really isn't a justification, however being a mechanical engineer you tend to see the world through data and experimentation. Plus it's fun.

I agree with napanorton and now it has progressed to fuel injection systems where setup requires exact mixture readings. And obviously I am not the only guy out there with a desire to make a Norton into something it never was as I have been overwhelmed with inquiries on the fuel injection systems. Now it has turned into existence. Jim

bb220, I just want to out run my fear reflexes in cornering and watch gauges doing stuff - before going in or after the fear catches back up in the opens.

A decade ago had a V8 optimalized for propane, WOT dyno test revealed my weird set up the builder recoiled from, made more power than the vaporizor was rated for and would of done even better but the narrow band showed too lean after 4600 rpm, [which some sanding on the fuel needles could solve].
W/o the voltage reading would of been hard to know what dropped the power rise as prior it was cam degreeing that bumped power over what equations predicted propane fuel can deliver. 15% less than gas, but V8 only lost 6% over hi octane gasoline testing for break in runs with a carb 2.5x's mixture capacity.

> Typical automotive sensors ('narrow-band' type) are based on a single
> zirconium dioxide sensor, and are somewhat temperature sensitive. I
> found accuracy figures of 2% or 0.2 fuel/air ratio near the
> stoichiometric point quoted on various data sheets. This should be
> plenty accurate for tuning a carburetted engine like a Norton. Some
> typical data sheets and other useful info can be found at:
> http://www.huntleyracing.com/o2sensordatasheet.htm
> http://www.scuderiaciriani.com/rx7/O2_sensor.html
>
> Wide-band sensors used in 'lambda' systems use dual yttrium stabilized
> zirconia cells, and are far more linear and also rather expensive.
> Linearity means that the accuracy is over a wider range of measurement,
> but near the stoich point, the single cell inexpensive type is almost as
> good. See the following for details:
> http://www.nernst.de/lambda-sensor.htm
> http://www.diy-wb.org
>
> Alan G.

Click to expand...

There's a lot of details in here to be aware of in O2 sensing.
http://www.performancetrends.com/wide_band_uego.htm

Jim, you think a diaphragm type pump will work or does it need to be a rotary vane type? If a diaphragm type, what material for the diaphragm - teflon? - or maybe what temps will it see - something like 350 degrees?

napanorton said:

Jim, you think a diaphragm type pump will work or does it need to be a rotary vane type? If a diaphragm type, what material for the diaphragm - teflon? - or maybe what temps will it see - something like 350 degrees?

Click to expand...

I was just thinking rotary pump for it's small size. I am sure anything that would keep the air moving in the right direction would work.
To use a pump with a 350 degree limit I would need to use a cooler in the hose. To mount the pump directly to the sensor fitting it would probably see 800 or 1000 degree F.
The pump I am using is a piston pump with Teflon rings. It has not complained yet but it has about 2 ft of ribbed metal tubing and a couple foot of rubber hose between it and the sensor. Jim

There are electric smog pumps I'll try to find a part # or link.

Interesting read on exhaust gasses.

http://www.power-tripp.com/AFR_and_gas_analyzers.html

Ludwig, the sniffer Innovate sells (like the first pic) in fact uses the Bernoulli principle, the problem is that the way it's implemented it doesn't work very well on single cylinder application (which is in effect what it's seeing) - probably primarily due to reversion effects. At least that's my understanding. I did think about using a blower to induce a more constant state vacuum, but initially thought some kind of a 12v vacuum pump would be easier. Turns out the selection for pumps that will handle the exhaust gas temps and flow the require amount sorta don't exist. At least I haven't found anything yet. So maybe a blower would be much easier - it won't need to deal the high exhaust gas temps and blowers seem easy to come by. I'll have to think about that.

-- D

ludwig said:

What about a tube parallel with the exhaust , using the Bernouilli effect , something along these lines :
(evt with a blower in front )

Click to expand...

Well I played around with your idea a little last night. It works with a large enough air source but it's looking like making enough air flow is going to require more power than a Norton alternator has available. Jim

Eductors are pumps without moving parts, driven by pressurized fluid flow.

from a Bob Patton 1996 forum discussion.
So, when you cut into your exhaust system to tie in your eductor port, try to do so in a straight run, favoring a location closer to a downstream bend and farther away from an upstream bend. Avoid locations where the cross section increases, these locations trade velocity for pressure and are inimical to eductors. If you have to locate your port in a bend, locate it in the throat and avoid the heel. Avoid locating the tube so that it protrudes into the exaust flow. And finally, cut the tube that you are using for your port tie-in square, and locate it perpendicular to the long axis of your exhaust pipe and at the side or top of the pipe, not the bottom.

Click to expand...

http://www.teglerizer.com/triumphstuff/ ... enging.htm

http://www.muleshoe-eng.com/sitebuilder ... ductor.pdf

For an eductor with gas as the power fluid:
– Exhaust pressure should be less than ½ power-gas pressure (in
absolute terms)
– Exhaust pressure should be less than twice suction pressure
– Mass flow rate of power gas will be about twice suction mass flow
rate.

Click to expand...

http://en.wikipedia.org/wiki/Eductor-jet_pump
An eductor-jet pump or jet pump is a type of pump where the energy from one ... into the exhaust hose along with the fluid from the eductor jet nozzles

http://www.foxvalve.com/iindex2.aspx
Solids Conveying Eductors · ejector, ejectors, eductor, eductors. Air ejectors use compressed air or other gasses to create vacuum that can exhaust, vent

http://cdfdata.fire.ca.gov/pub/fireplan ... epdf29.pdf
5. EXHAUST SYSTEMS 5.1 Normally Aspirated Diesel Engines
File Format: PDF/Adobe Acrobat - Quick View
maintenance cycle, and in some cases a 13 to 15-day cycle. Photograph 5-13. Dirty Eductor Tube. Photograph 5-14. Looking Down Exhaust Stack with Eductor



These show the shapes, size is ~4"x7" with pulley, 3 wire is better for idling Norton and everything on drawing on its 60 amp capacity.
http://compare.ebay.com/search/mini+alt ... 542_304572