I took my car to a dyno day yesterday and was pretty disappointed. The car didn't make quite the amount of power that I expected. The dyno operator said that it was because the A/F ratio was too rich. According to their WB it was at 11.6 at 6000 and dropped down to 10.2 by redline; 8200rpm. I have datalogged a few runs before and after the dyno, and my gauge consistently reads 11.9-12.3 throughout the entire pull. I know that cars read leaner off the dyno than on, but it seems like a bit much to me. Unfortunately, with it being a dyno day and not an actual tuning session, I didn't get to datalog the dyno runs and didn't think to ask the operator what my guage was reading. Anybody have any ideas?

The car is a stock Gs-R engine and running 12psi. It only made 245whp/162trq. A few people at the shop said that the dyno typically reads 10% lower than other dynos, but that doesn't explain the A/F ratios.

Also, will running at 10.2:1 damage anything? I feel happy with the cars power, and so long as it will not hurt anything I will continue trying to tune it myself. If running that rich is bad for the engine I guess I will suck it up and pay to get it tuned.

Thanks,

David

 

So this being just a dyno session you didn't tune your ignition timing either? That combined with your A/F ratio would significantly lower your power. You need a real tune.

 

was the wideband the kind they stick in the tailpipe or downpipe?

 

It was the kind that goes into the tailpipe. ::EDIT::

 

It was the kind that goes into the tailpipe.

 

?

 

I would trust the PLX
Ive found it to be very very accurate

and if you read up about o2 sensors you will find the location is close to the engine around 26-36 inches from the head depending on NA or FI
so in the tip of the exhaust is inaccurate
check the plx site for detailed info

 

Whoops. And that is why you shouldn't post before leaving the house in the morning. I wrote downpipe and hit send but thought I had hit the stop button before it went. Fixed it to prevent confusion.

 

Thanks for the tips. I'll check out their website when I get home.

 

I agree that tail pipe sniffers, unless they are of a scientific testing quality $$$$, are not as accurate for real time data. However, they are usually pretty close for an overall AFR reading. The closer the sensor is to the combustion chambers the more reliable the real time data would be for tuning.

Do you still have a catalytic converter installed? It's odd to see a tail pipe WBO2 sensor read richer than the upstream WBO2 sensor. Specially if you have a catalytic converter still in your exhaust.

I have had pretty good luck with the setups I have tuned that were using PLX sensors. I would probably trust it more than I would an unknown WBO2 up your tail pipe.

What do your spark plugs look like? For help in reading AFR's and ignition timing that is as close to the combustion chamber as you can get.

 

The dyno operator said that their sensor was $3000. I would think that it is pretty reliable, but idk. I do not have a convertor on my car. The plugs that I pulled out the day before the dyno looked a tad rich, but they were in the from when I first started it up with a super rich basemap.

I played with the laptop some last night. I noticed that hondata was reading .3 leaner than the gauge so I fixed my offsets and that explains some of te difference. I also noted that at WOT it stayed at 11.79 even though the gauge read lower. I leaned it out some, relying on the gauge, and it seemed to more accurately follow the gauge. I don't know why but will continue messing around with it to get better results.

 

In your OP are your numbers from looking at the gauge or from the datalog? It's pretty common to have an offset between the gauge and the ECU. It doesn't necessarily mean that anything is "wrong," just that you need to calibrate it.

In my car the offset also changes with load or engine speed (I can't remember) -- at high load it tends to read leaner IIRC. If the same is happening to you, you might have inadvertently richened out the high rpm portion of your map, which is why their sensor reads you as being extra rich up top.

That's why direct serial logging is the best. I only direct log now for my street tuning in Neptune and I don't even look at the AFR signal the ECU sees.

 

Where is your 0-5v wire hooked up? D10?

 

Their dyno showed 11.6 and dropped to 10.2. My gauge showed 11.9ish. I checked the offsets Monday night and they were off by about .3. That took up some of the difference, but it was still about a point off.

Forgive the ignorance, but what is direct serial logging? I read so many articles before putting this all together that I either read it and forgot it instantly, or just never came across it.

Thanks

 

Yes, it is wired into D10.

 

If you didnt remove the two resistors in the eld circuit on the exu it will resd fine for everything except wot. I beleive ita r138 and r136. Remove these and it will read perfext with little to no offset

 

I am almost positive that they were removed. I bought the ECU and wideband from phearable at the same time and he asked a bunch of questions to set up the ECU for me. The next time that I work on the car I will pull it out and check to be sure. Thanks for the help.

 

From experience, I know how it reacts when the eld resistors are not removed.

 

I just took it out and they are still in. Are there any bad effects to cutting them? I found on Hondata's website where it says to remove them, but am a little nervous about cutting stuff out of my ECU?

 

U arent using the eld circuit as its designed so you are modifying it to fit your needs.

 

FYI Bosch sensors fail! i would trust his NTK Sensor. dyno i use uses a tail pipe sniffer (NTK Sensor) and i have matched it perfectly with my bosch sensors in both my plx and my LM2. a lab grade sensor is just gonna win no mater what.

As far as your readings you probably dont have the wideband properly grounded to the ecu logic ground (thermostat housing). I have tons of customers with incorrect wideband readings do to this. Does your gauge match the reading in your tuning software (didnt see it listed, i just skimmed)?

 

IDK if Hondata can do it.

On Neptune I can do it with Innovate widebands. The way it works is that you connect both your wideband and ECU to two different ports on your computer (USB probably). The computer recieves the signal from the wideband and sends it to Neptune, which logs it along with the signals it logs off of the ECU. That way you're logging EXACTLY what the wideband sends, with no offsets, interference, etc.

 

Gotcha. I'll google it and see if there is anything similar for hondata.

 

At idle they were matching, but under full load they were off a little. I grounded the wideband behind the right side headlight. I looked at a wiring diagram and thought that that was the ECU ground. Maybe just not the one used for logic?

 

I cut them out today and didn't get the results I expected. I thought I would have to lean it out some to get to 14.7 at idle. But when I started it up it would barely idle and read 17.9:1. I didn't have time to drive it before work, but is this normal? I had to change the offsets from -.37 to -.1 so I guess it got it closer. I have to get the rear motor mount on tonight and then will go drive it and let you know how it does after playing with the fuel numbers a little.