...and it works great! The guage I'm talking about is the Ultralite 2" O2 sensor guage. It's a 20 (I think) segment LED with three colors (red lean / orange stoich / green rich). The way the meter works (which suprised me) is that it samples extremely quickly, and only lights up one LED at a time - however because it samples so quickly, it often blurs a few LEDs on one or two sides of the one that is lit up the most. It's hard to communicate, but basically this helps make up for the problem of O2 sensors being 'rich/lean switches' as many complain.

Before I tested it, I plugged in a wide band O2 sensor with a hand held readout ($1050 setup). They both gave very close and comparable results. However it seems the Autometer setup (non wideband) has a range of around 9.0:1 - 17.0:1, while the wide band reads from < 8.0:1 - 26.0:1. The wide band also reads the ratio down to 1/10, while the Autometer is probably closer to 1 LED being 1/2 - 1/3.

I built a homemade O2 guage with a 10 segment LED readout, but there is no comparison - the Autometer setup works 100x better. Once again, here's a car part not to skimp on. The O2 sensor I use with it is a 3 wire (heated) one, which heats up quite quickly.

To sum it up - I'm selling the wide band setup (I've got better things to do with that money), the Autometer setup is DAMN close to being as good, and a heck of a lot cheaper (and nicer looking). I really think that the wide band setup is overly accurate for a carb'd system, the traditional style is good enuff!

Rob

 

I have the edelbrock system and think it is a very good tool for setting up the mixture. I would highly recommend anyone to install the gage and properly set up their carbs.
Norval

 

How is the 3 wire O2 sensor wired?

I REALLY need to get around to hooking up my O2 gage!!

 

How much did the Autometer gauge cost?

Why didn't you lilke the home made version??

 

The 3 wires are for +12v, gnd, and sensor. Basically, they are self-heated so you do not need to depend on the exhaust gases for this.

I'm also wondering why you did not like the DIY meter? It seems to me that the accuracy is dependant on the sensor, as all the gauge does for you is lights up LEDs based on the voltage of the sensor. I'm planning on building one of the LM3914N based meters. Now I may also by an Autometer from Summit (they have a good return policy) and compare them.

 

Chris - the homemade one is based on a 10 LED segment (verses 20 LED for the Autometer). It also will light up all the LEDs up to the one indicating the voltage level (like a volume level on a stereo). It seems to update slower (as in the LEDs move a little more sluggish, not a big deal). The homemade one is also all the same color, so you will have to count the LEDs to see where you are at on the scale (in the dark anyways). I guess you would have to try them both, but I really think you won't be returning the Autometer

As far as wiring the 3 wire goes, I grounded it to the exhaust pipe and ran a + for the heater element and the other line to the sensor. I will be changing it shortly. The right way is to run a ground wire (and the power wire for me) from the O2 sensor to the guage. I see a noticable difference when I play with power accessories. I highly recommend a 3 or 4 wire one. I went to Pep Boys and had the guy bring out a whole bunch from the back. I took them all out of the boxes and compared them till I found the one I wanted.

Keith - I think the Autometer one is about $50. You can make a homemade one for around $20 and an hour or two. I have four Ultralite 2" gauges mounted, so the Autometer fits right in. My o2 sensor was $60 I think.

Later

Rob



[Modified by robzr, 1:33 AM 8/15/2001]

 

I'm kinda glad to see your complaints about the home made one, as I think I have a fix for both issues. First, there is a LED bar graph from Newark Electronics Part #91F4665 which is GGGYYYYRRR so this fixes the "all the same color" problem. Also, you can set the LM3914N in "dot" mode where only one LED is lit at a time. This is done by not connecting pins 3 and 9.

As the 10 vs. 20 LEDs go, it doesn't matter to me. I don't believe the O2 sensor is accurate enough to make use of the extra granularity and the over-all range is the same. The only thing that still concerns me is the sluggishness.

Thanks for your comments.


[Modified by Chris A, 3:20 AM 8/15/2001]

 

robzr
your wide band comparison certainly proves the common sensor is fine. How many inches from the head is it? What circuits were you testing? ie idle, WOT, cruise?

chris http://community.webtv.net/MATTGRU/AFtesting

 

Drives61 - the bung is welded in the collector. I have the 1 5/8" dynomax headers. There is a slight delay until you get a reading, but it is very short. The wide band sensor now that I think about it didn't seem to have a noticable delay. At any rate, that is not really a problem. I tested idle, cruise, WOT and transition. With the wide band setup it was convenient because it was a backlit large LCD readout, so I could hold it in my hand next to the steering wheel and it was easy to see while accellerating.

Chris - you really know your stuff about this circuit. That multicolor LED readout definately sounds like the way to go. In that case I would probably not make it a single LED - it's easy to pick out '2 yellows' whereas if it's just one LED you don't know where in that range it is (and it's definately accurate enough for that to matter). Also, one other nice thing about the Autometer one is that it is in a circle configuration so you can quickly glance at it and see the position of the LED in the circle. The only downside would be that it is 2" diameter so the only place I found to mount it was under the dash (which is good for the sleeper image anyways). I enclosed my home made one in a tiny Radio Shack project box (a little bigger than a 9 volt), so it would be easy to temporarily put it on the dash which would be easier to see. As far as the sluggishness goes, it is not slow by any means. Compared to the Autometer though, the autometer one actually blurs the LEDs next to the one that is working. Now that I think about it though, that probalby wouldn't be from updating so fast, maybe just the way the circuit is setup (to allow partial voltage to nearby LEDs). The homemade one as you probably know, the LEDs are ON full brigtness, or OFF.

Later,

Rob

 

Rob, just a couple of comments;

First, a 20-led version can be made at home with two LM3914's - just cascade them. National Semi's website shows how to do this.

Second, there is a cheap (working!) Wide-Band setup available, but it's completely DIY (make it at home.) Sensor is about $150 (can be found for a little less) and the circuit is around $50 or so to make. Much cheaper than the commercial units. Contact me via e-mail if you're interested.

What I'm really curious about though is, does it still give you the large swings that a normal O2 sensor gives? (Attributable to it being a switch?)

Ken

 

I'm planning on welding on a bung. Did you weld a bung on each side? I have headers; would the collector be an appropriate location for the O2 sensor? Thanks,
jcj

 

Ken - The normal O2 sensor is pretty touchy, but it has more range than I expected. IE I was reading right on the rich side, then I went in 1/8 turn on my four corners (idle mixture), and I was reading on the lean side of stoich. However, where the wide band was reading down to 11:1 - 12:1, the autometer still has a couple LEDs to go. The key seems to be to keep it right on the rich side of stoich, and it is not too 'switchlike' to do that at all.

The wide band setup seemed to sense the mixture very linear like (if that makes sense). However, that precision is wasted with a carb. My idle circuit fluctuates significantly, I'm guessing from engine heat, engine compartment heat, etc. If I had EFI, I would have kept that wide band setup. But after seeing the Autometer function beyond my expectations, I am much happier with that and $950 in my pocket. I paid $1050 for the wide band setup BTW. I'm glad I found a buyer for it. The quality of the handheld unit was not very good (crummy button, cheap case). The sensor, readout and wiring was very nice though. BTW the sensor was a 'Bosch fast action wide band', and the guy sold replacements for $350.

I'd really like more info on the DIY wide band. I'll email you.

Rob

 

robzr
kudos for doing the comparison of wide vs. regular. i will add your results to my article this weekend if you don't mind. That way if the forum melts down, it won't be lost.

 

Haha that sounds like a wise idea ;)

Rob

 

Chris;
Thanks for finding the multi color LED's :D

I think the 'sluggishness is due to the capacitor, it is expressly there ( according to the designer) to filter out the noise from the O2 sensor, so effectively averages many samples over time. If you were to remove it, or use a smaller cap it would react faster.

Cool stuff:cool:


[Modified by 427V8, 9:44 AM 8/16/2001]