DIY Wideband A/FR meters: Impressions? SteveW??

[RickM]

I recall mention of a DIY Kit (Do-It-Yourself) Air Fuel Ratio Meter that was going to be built by SteveW.

I'm curious as to how accurate and usable these units really are. From what I've read the placement of the L1H1 sensor is very critical. How does it fair in the 911 exhaust configuration? Also, would it be useable in a Turbo application? Seems they mention the heat and possibly other factors could influence longevity and accuracy.

First hand experience and comments appreciated as I am interested in purchasing a unit.

BTW, I see that there are now less expensive options (Bosch) regarding the 5 wire sensors. Are these alternatives as good or better?

[ChrisBennet]

Check out this the one from Innovate Motorsports. Once they support RPM data logging (slated to be available by October 15th) it will be a pretty good setup. It has memory for up to 44 minutes of data.
-Chris

[RickM]

Thanks Chris!. For $349 that seems to be an excellent deal. I'll have to add that to the list for consideration.

I'm still curious about the kit models and how they fair in 911 and/or turbo applications. They specifically mention the the sensor should be situated before the Cat and not subject to gas temps over 850 degrees.

Also, it seems that this would be a semi-permanent install. A new bung welded on and sensor screwed in. What are the options for using on multiple vehicles?

Thanks again.

[rs911t]

I bought a wideband O2 sensor from http://www.lambdaboy.com. It also has analog outputs, so I was able to plug it into my DL-90 to generate AFR/RPM graphs.

[Steve W]

Rick, did you mean me, Steve W? I did build a wideband exhaust air/fuel monitor a couple of years back when there really was nothing else on the market other than the $4000+ Horiba and Motec analyzers. It was built from a system put together by an Australian company and uses the same NTK sensor used by Horiba and Motec. It's very fast and accurate, and measures afrs from below 10:1 and past 20:1. Alternate systems are now also using the Bosch wideband sensor that has a different voltage curve than the NTK unit. Their sensors should be just as good, however the display device needs to be calibrated for the correct sensor. Most of them use a programmable PIC chip to do the voltage to display interpolation.

Ever since the pursuit by members of the DIY_EFI boards about 5-6 years ago to reverse engineer the Horiba system, reengineer an affordable wideband analyzer, and finally putting together a kit about 3 years ago, there seems to be a spur of smaller companies developing affordable compact systems. The one Chris mentioned is certainly the best one I've seen and if I had to do it over again, would most likely be the one I would go with due to it's data logging and variable input sensor capability. I did see a unit at the Ventura GermanAutoFest by Paradox Engineering that was very interesting. It also uses a Bosch wideband sensor but the digital display unit was encased in a brushed machined aluminum case that was no bigger than a pack of Big Reds chewing gum. It's about $150, sensor not included, and has the capability to replace your existing narrowband sensor, and simutaneously output a second voltage curve which would emulate a standard narrowband sensor so you could feed it to your DME. In addition, there was also a LED row based display so racers can monitor their afrs with peripheral vision. All this stuff was developed by an electrical engineering student, I think, at UCSD.

The sensitive part of these sensors is that the need to be fully heated up before being exposed to the exhaust, or risk being contaminated. The preheating circuit is a large part of the circuitry that these units are built of, but once the sensor is heated, which takes about 30 seconds at 12.5+ volts, it's very sensitive and ready to measure the slightest bit of CO. Just being in the vicinity of a idling car will read a output variation. I use mine on a short piece of pipe with an O2 bung which I stick up muffler tailpipes. The exhaust flow and velocity from a 911 at full throttle is so great, on a dyno you can feel the hot exhaust shooting out from 40 feet away, so there is no problem with accurate sensor measurements.

[jkeyzer]

There's a DIY kit available now for about $130 plus sensor ($30) from a company called Tech Edge in Australia. I ordered one of their kits and will start putting it together in a few weeks.

http://www.wbo2.com

You can use a Palm pilot to see the AFR and log data. It's probably the cheapest solution around and apparently it works.

[turbo6bar]

I am a volunteer for www.diy-wb.com, the ORIGINAL DIY wideband meter mentioned by Steve W. It is a non-profit effort, and others have copied the diy-wb design and profited immensely.

Anyway, if you would like to buy a DIY kit, let me know. The kit and pcb are around $20, which is essentially wholesale price for the parts.

I credit the DIY-WB as the force that led to affordable wideband O2 meters.

The meter from Paragon Products sounds very nice. If it works as advertised, it is a steal at $150. I would not give Tech Edge the sweat off my back.

The NTK L1H1 is the expensive sensor, and I expect the LSU4 Bosch sensor will eventually become expensive when the cars using this sensor fall out of warranty.

Oh, well, enough yammering.

Jürgen

[Steve W]

Ok, my dumb mistake, it wasn't Paragon Products, but Paradox Engineering - sorry. My brain is tired. Plus we should give Jurgen credit for bringing about the availablity of affordable systems.

[tbitz]

I picked up the DIY wide band kit from turbo6bar (Jurgen). Obviously a great deal at $20. You need to be able to solder, build a case, and get the proper pigtail connector for the L1H1 sensor and ofcoarse get the L1H1 sensor. The above link has a typo. Click here to get more info: http://www.diy-wb.com/

I have it installed in my 911 which has backdated exhaust. I drilled a hole and welded a bung on my headers.

I used it with the megasquirt ECU which does the logging. You could use the WBO2 with the controller Jurgen is selling on its own and simply read the voltage using a multimeter to figure out your AFR. The sensor must be placed after the turbo (if you have one).

[89turbocabmike]

Rick, I have had the Innovate LM1 for about 6 weeks. Price and the fact that the manufacturer/designer is located just 20 miles from me was a big influence. I had been looking at the Tech Edge until I came across the LM1. The units are still under constant development. I had a few small problems initially which have been corrected with updated software. Check the Yahoo group listed on Innovate's support link. Klaus, the designer/engineer knows his stuff and is in constant communication with customers as the product develops. I am waiting anxiously for their "aux-box" which will allow rpm/ign/boost etc logging. Klaus' personal ride/test vehicle is a 3.2 powered Vanagon, so he's one of us!

[RickM]

Thanks guys!

And yes I was reaching out to you Steve (SteveW). I was curious how you made out with your AF meter project. I like your Pipe-in-pipe method of testing. However, I've read that it is desirable to measure before the cat. Can you (or anyone) comment on this?

At this point I think I'll pursue a unit that utilizes or has the capability to use the Bosch LSU4.

From what I understand the TechEdge guy is a former diy-wb member who decided to profit from the project. True?

[RickM]

Does anyone know if Paradox Engineering is still in business? My email attempts have all bounced back.

[kstylianos]

Rick,

I saw this the other day this:

http://www.plxdevices.com/NB_WB_GaugeConv.htm

Its a narrow to wideband gauge converter. Unfortunately, it only works with Autometer gauges. They claim that with a wideband sensor, you can accurately measure (with LED lights) from 10.0-20.0 in 0.5 a/f ratio increments.

Converter $35
Sensor $175
Gauge $50

To be honest, the "Innovate Motorsports' unit looks more appealing

[jkeyzer]

turbo6bar - How much do the sensors cost for the DIY-WB?

The big benefit I saw from wbo2 was that the sensors are so cheap. It's too bad the wbo2 kit is more expensive, but last I checked the NTK sensors were well over 100 bucks.

[turbo6bar]

Originally posted by jkeyzer
turbo6bar - How much do the sensors cost for the DIY-WB?

The big benefit I saw from wbo2 was that the sensors are so cheap. It's too bad the wbo2 kit is more expensive, but last I checked the NTK sensors were well over 100 bucks.

NTK sensors are between $140-225, if you are lucky to find them. They're occasionally found on Ebay.

The LSU4 sensor sounds like the hot ticket right now. If I were buying, I'd look for meter that supports the LSU4 or the LSU4 AND NTK sensors. The diy-wb does not support the LSU4, unfortunately. A few guys were planning to make a new diy-wb, but so far nothing has materialized.

Tech Edge basically ripped the diy-wb design to get their show started. Now, they have (re)designed a meter to work with the LSU4. I would encourage anyone to look elsewhere for a meter. The guy lacks ethics, in my opinion. If the guy was a Pelican member, I would expect him to copy advice from this forum and sell it in a book or another forum for profit.

I have an assembled diy-wb that is not being used (d@mn CIS is just too good ). If anyone would like to borrow my setup, just let me know. It does not have a wiring harness to permit use inside the car, so you would need to do a little work(basically lenthen the wires).

Yes, Rick, you do need to measure pre-cat.
kstylianos, the converter you mention is not the real deal. It is just an interface to let a wideband controller talk to the Autometer gauge. You need to add in the price for a control box.

Later,
Jürgen

[Steve W]

Hi Rick, Paradox Engineering is a very new company so it seems his site is not up to date. I met the guy, his name is Chris Williams, try cswillia@ucsd.edu . If that doesn't work, call Huntley Racing, that's where he did some of his testing and development. I was told the afr output readings from his sensor were compared to Huntley's old $10,000 wideband system, and they were identical to less than a percent. He actually has a lot of products as I have his product flyer here and not all of them are in production yet. If you want, email me and I can send you a scan. And yes, you would want to place the sensor before the cat if you have one because the cat does that whole conversion thing with the exhaust gasses, and that throws the mixture reading off.

[Jack Olsen]

I'll speak for the idiots, since I want to put one of these in my car. What's the difference between these setups and the inexpensive off-the-shelf air/fuel ratio gauges?

[turbo6bar]

Originally posted by JackOlsen
I'll speak for the idiots, since I want to put one of these in my car. What's the difference between these setups and the inexpensive off-the-shelf air/fuel ratio gauges?

When the off-the-shelf meter says 13:1, it is really just a guess.

When a wideband meter says 13:1, you can bet it is pretty close. I would say most meters should read within plus/minus 0.1 or 0.2. In the past, the biggest disadvantage to the wideband was cost. Until recently, a wideband setup sold for at least $1000. Now, the prices have fallen, and there is no reason to trust a high dollar engine with tuneable fuel/ignition to a $25 narrowband O2 meter.

Find a quality commercial setup that will fit your needs, unless you adore doing it yourself.

Jürgen

[jkeyzer]

I believe the wbo2 board has a simulated narrowband O2 output for use with a narrowband gauge or a programmable FI system like Megasquirt that only has a narrowband input (for now).

From what I hear you can set the "stoich" point to whatever you want and the wbo2 module will convert the wideband output relative to that stoich point.

turbo6bar - Sorry to hear that the Tech Edge guys ripped off your design. I am an engineer and I can understand your frustration. Unfortunately when I surveyed the meters out there, theirs was the best suited to my needs. We'll see how it works out.

[TimT]

The paradox engineering AFM is a crapshoot. Seems like a cool thing.. then the Huntley association popped up.. I tried the web contact thing with paradox, then used mailto:dkhajavi@nethere.com which is remarkably like Huntleys address

can you tell Huntley isnt on the top of my vendor list?