kekek-20

i'm not FI but see this is the forum where most hondata folks are..

i just installed the aem uego wideband and am trying to interface it with hondata. with the car not running, the aem output is correct at around 6V but the voltage shows up in romeditor as 3.83V?

I triple checked my wiring and i'm using a dead O2 plug as suggested. 1k ohm across heater wiring, output and signal ground are correct.

could it be a problem in the lamda conversion table? the table goes from 0-5V and the bottom-most cell is blank at zero. i tried it that way and also tried inputting 6v with a corresponding lambda of 1000 just so it's off the chart.

any ideas? any help would be greatly appreciated!

thanks,
dan

Spunkster

You aren't the first person to have this problem. Sounds like a problem with the AEM setup.

I have no idea what the problem is, but I figured you should at least know others have it.

kekek-20

thanks matt. i spent 2 days stumped so it's nice to know it's not me.

do you suggest that i sell the aem?

i'd be willing to ship hondata my aem wideband setup to figure out how to interface the two. i'm sure others would like to use the aem since it's a high quality inexpensive alternative to the fjo and motec.

Hondata

3.83 volts is typical when there is no input voltage signal to the ECU.

Use a voltmeter and check the output voltage at the AEM wideband is correct. Then trace the voltage/continuity to the O2 sensor wiring and then to the ECU.

true

Add another one to the list. Appearently the AEM wideband does not work with Hondata. I am certain I have it hooked up properly, and at idle/no load conditions RomEditor & HondaLogger both read 3.83v. If I give a rev or start driving even under load, the voltage starts moving around and acting normal by giving me decent a/f's etc.. But just sitting there idling while upping the fuel to the point of about killing the engine, RomEditor shows 3.83v. Testing the connection at the ECU on the o2 signal with my multimeter shows ~4.5v at idle. The voltages are obviously not in synch! How could this be? Now, there are no taps or anything on the D14 ecu input (obd1 wht/red), the wideband output is connected directly to that wire. I used ground from the wideband controller on the Chasis, Direct battery (how the aem instruction tells to ground), and even A6 (the stock heater controller ground). All of these resulted in the same voltage going in of 4.6v on my multimeter, but still only 3.83v in Hondata software. This doesn't make sense. I tried disconnected the signal ground (D22) and the voltage just goes wacky so obviously the signal ground is needed. I am pulling power from the Wideband controller from A25 which is what controls the narrow stock o2. How could a provided 4.6v from the wideband output tested with a mutlimeter come up different in the ECU.......Anyone????? Willing to try whatever it takes. One other thing, before I ever started the car, I clicked the car to accessory..and after about 2 minutes the o2 sensor was nice and hot on a cold engine....I'm positive the heater is working. Thanks.

Also, in response to keke's original post...how could the AEM output be 6v?? Where are you getting this value...using a multimeter? That sounds extremely strange because the wideband output is 0-5v!! You are using the orange wire from the wideband for o2 signal correct?



Modified by true at 6:07 AM 9/9/2003

PHiZ

Are you guys calibrating it like this:

http://hondata.com/techwidebandtuning.html

Does the AEM require you to calibrate it in fresh air?

Widebands need a couple of minutes to heat up too...

-PHiZ

true

Phiz, the calibration really has nothing to do with any of this. The fact is that the voltage from wideband output should match up with what the ECU is reading. Weather the unit is calibrated and is accurate or not calibrated shouldn't cause the voltage differential. I am not using a 1k resistor across the stock heater control setup, but on that page it says the only keeps the check engine light from coming on..doesn't bother me. I have no tried the Diode......and this basically describes my problem, but I do know people who had similar issues with th Techedge and the diode didn't do a thing. Is Hondata still recommending trying this? Would appreciate a response from Hondata. Thanks.

PHiZ

well, I'm saying there is a dialog in hondata, where you can change the o2 sensor voltage scaling...

Has anyone checked the hondata forums? or ask kyle at ks-motorsports if he knows anything?

-PHiZ

true

Phiz..I understand. You are able to correlate A/F values based on the voltage input. But again, this doesn't change what the incoming voltage is. It allows you to put in 5 volts = 30 A/f, but that won't change what voltage the ECU is reading from the wideband controller. I posted on the Hondata and AEM board. Someone on the AEM board responded that the AEM has an o2 gain so you can synch up the voltages. I'm wondering if I can run a potentiometer inline or something....but according to Hondata 3.83v usually means no signal in...so something is just goofy.

Pretty frustrating.

racerxadam

true, sorry to see you still fighting this. I made the switch to aem and haven't looked back. The gain function works very well to sync the voltages and to tell the truth the gain that i use is no different than any other honda that i have had my wideband on. it acts the same as in a newer 92-95 car.

true

Yeah adam still having the same grief. I had some types of issues with voltages being off with the TechEdge sensor when I was using with the 4b. I know you got the AEM and I decided to give the s200 a shot instead. Ton of improvements over the 4b, but still having some problems with voltage differential. I was hoping someone would have gotten to the bottom of this problem by now..it's been over a year. The strange thing here is that at idle it just shows 3.83v.....when i had my narrow sensor disconnected a few months back..that's the same thing that hondalogger read. So basically, if the ECU gets no signal..I guess it displays 3.83. With the techedge it seemed to have always been reading a voltage...(albeit wrong) The fact is, my wideband output has voltage...the ecu is just interpreting it wrong......... does anyone here have a wideband using the Hondata where the voltages match?? any brand?? Arg. I've tried all kinds of grounds..everything...can't get them to match. I would assume my ecu is fried, but other people seem to have the same 3.83v problem. Sounds like the Hondata would really benefit from a o2 gain system, if possible.

true

One other thing....the AEM o2 is the Bosch unit. I know most other personal ones are the NTK....I was told hondata worked fine with the MoTeC unit..which supposedly used bosch, but a quick search on their site shows it takes both bosch and NTK. Anyone know which one the MoTeC PLM ships with...or anyone know anyone using a bosch o2 to tune a Hondata....

JDogg

when the aem wideband is used with the aem ems there is a option in the ems software called o2 gain that allows a correction factor if the output voltage from whatever wideband you happen to have is different from what the computer reads. not sure if hondata has something like this or not.

true

I think I'm onto something. I have attached my datalog from a little drive around the block. http://www.boosted.org/aem_wide.rec. This is a recording with the Diode in place, which didn't seem to do a thing other then change my input voltage from 3.83 to 3.81. It's probably easier to follow the voltage when reading this recording in RomEditor. Now, you should notice that throughout the ENTIRE run, voltage in NEVER gets above 3.81. (for AEM stoich is OVER 4.0 volts). So while im idling it appears that the aem feeding 4.3 or 4.5 or whatever is out of range! You will also notice that most of the light throttle light load driving shows 3.81v....this should be an a/f of over 14, so again this all would be out of range of that 3.81v. A very interesting thing to note: Fastforward to 3:10, from this point I am simply sitting there idling...and start decreaseing the fuel table.....you can see the RPM's dropping..and at around 3:25 the car about dies from being so lean. Next I start going rich at about 3:35. Now you will clearly see that it will start adjusting the voltage once the A/F corresponds with a voltage UNDER 3.81. So basically, I believe that the Hondata is simply NOT working with any voltage over 3.81...it's acting like its out of range. Once i lower the voltage past 3.81, it seems to read fine. Anything above 3.81 it simply shows just that. I bet the other o2's that are working have a stoich and interesting voltages under 3.83/3.81? Is there a cap/bug in the software for voltages over that?!

true

Ok, I just did a perfect test, I'm now certain that it's NOT the AEM. Here's what I did......idle the car...shows 3.81. Keep upping the fuel untill it comes down to 2.8volts. Now it's showing me a very rich A/F and guess what the output of the controller is??? 2.8volts!! So the voltages are in-synch to anything UNDER 3.81 volts...anything OVER 3.81volts on the controller (perfecly usable on the AEM) is clipped off and displayed as 3.81!! What gives? I posted this in the Hondata forum..can anyone else supply a voltage of over 3.8 volts to their hondata and see if it reads it???? Now it makese sense..........

JDogg

if you dont have any luck getting the hondata to read over 3.8 volts you could use the 1volt output...

PHiZ

Is the O2 sensor/sensor controller, and ECU 02sensor signal ground all the same?

I.e. Are you using the same signal ground for everything? I could see that cause the ecu to read a slightly diff voltage.

-PHiZ

true

Hey Phiz..did you read my other posts..specifically where I said ' I used ground from the wideband controller on the Chasis, Direct battery (how the aem instruction tells to ground), and even A6 (the stock heater controller ground). All were the same'......

Anyways, I'm confident I figured it out. I'll let you guys know the status of Hondata's reply...

PHiZ

Whups, missed that... just brainstorming...

-PHiZ

kekek-20

sorry for the late reply- i was away for the weekend for the touring car races at laguna seca.

i came across this post on endyn's forum.. posted by methods:

Can you hook an AEM WB02 up to a Hondata directly? NO

The Honda ECU (that I have) will not recognize any voltage greater than about 3.8V (Around 14.4:1 A/F….. i.e. it will always read rich). The solution is to run a double buffered voltage divider between the WB02 (controller for you B20 :-) and the ECU input. The reason you have to use a buffered voltage divider is to supply the appropriate impedances to both the WB output as well as the ECU input. The circuit consists of a Voltage follower OpAmp followed by a resistor voltage divider followed by another Voltage follower OpAmp.

seems like he found the answer. too bad i'm an Mech E not an EE so this stuff sounds like blah blah blah voltage blah blah to me. i emailed methods for specifics and i'll let you know if and when i get a response.

-dan

true

keke, link to thread? I couldn't find it on that board. I'm pretty sure I did a test that shows there is just a bug in Hondata displaying voltages over 3.83v, not an ECU limit of 3.83. I used auto-tune and it was able to still tune with a proper wideband voltage while still only displaying 3.83. I have sent all my info to Hondata and am awaiting a reply.

kekek-20

http://theoldone.com/forum/top...age=2

methods has emailed me back as well as posted the diagram in the thread.. i'm going to be calling him later today to get a better understanding. i'll post the info after i talk to him.

thanks methods!

-dan

kekek-20

ok, spoke to methods. very knowledgeable guy.. designs missile control systems or something for the gov't. half the stuff he was saying was way over my head but i now i understand what needs to be done.

basically, the aem wb is putting out a high impendence signal and the ecu likes to see a low impedence signal. we also need a voltage divider to step down the output so it's not going over the 3.83V threshold.

look at the diagram in the thread on endyn's forum.

you'll need to buy a chip that has 2 opamps on it from radioshack. lm series should be fine. you'll also need a pair of 5k resistors that are at least 1% (might not carry em at the shack). good resistors are important. wire it up according to the diagram.. you can use a breadboard which will then fit in a 2" x 1.5" box from the shack which will package it in a nice little black box.

make sure you're using ecu ground in the diagram.

since the voltage will now be halved from the voltage divider.. you'll need to halve your lamda table values. i extrapolated the values from the aem spec sheet for my conversion table.

does this all make sense? hope this helps the aem uego/hondata users out there.

methods is the man. he got the idea for this ckt off one of the missiles!


-dan

JDogg

that seems like alot of work to me... id just use the 1 volt output. the 1 volt output from the aem uego is still mostly lenear and will work fine to tune with

kekek-20

i don't think the resolution of 0-1v would be so hot.

methods response to using the 0-1v..

ABSOLUTELY NOT!

Do not use that 0-1V signal. If you look at the graph, you will see that
the 0-1V signal looks nothing like the true 0-5V output. The 0-1 is
actually purposely skewed to appear like a regular O2. In fact, you would
get better readings off of a regular O2 than you would get off of that
simulator.

It would solve the Voltage issue, and it would solve the impedance issue.
But, if you tune your car with that signal you are gona be a very sorry
engineer.