Wideband read differently on your MS? This might be why.
 

So I purchased my Miata pretty stock, it did however have a narrow band gauge installed. During my AEM UEGO installation I saw that most of the required were already there from the previous gauge- My Power, Ground, Illumination. Why run new wires right?

My output wire from my AEM WB is a 0-5v signal, this means that with 14.68 AFR my output voltage should read 2.34v. Due to the previous gauges poor grounding, my output at 14.5 AFR was only 1.2v. Offsetting my entire tuning efforts when it came to my VE table and AFR targets. I just figured the configuration for the gauge was incorrect so I was trusting the AEM gauge face and ignoring the rest.

I got an itch :nuts: and pulled the gauge from the dash. Stripped a bit of wire while I was pulled off making some changes to the AFR target table, held it to my cigarette lighter ground and BOOM goes the dynamite, the gauges were in sync.

The dynamite, it goes "boom".

I get quite a few calls from locals asking for help with this shit, which is a scary thought but that's another discussion all together. My first inquiry is always WBo2 wiring. My schedule and availability are subject to this answer/description. I'd say about 75% of the locals don't read the directions and tell me that "ground wires are ground wires, run them all together!!!"

Someone suggested grounding the wide band and ECU to the same location. Seems like a good idea.

Grounding them in the same location is a good idea, as long as you don't need to extend the wires very long. It's the same concept with car audio, the shorter the power wires, the better.

I tried grounding at the ECU with the AEM I used to have on my near stock 94, it did no go for the matching with MS. Hell I just sold the AEM and bought another LC-1 no problems yet, should have never sold the first LC-1.

"Grounding" is such a mystery to most, even to engineers.
I used to teach a course in college regarding grounding and noise in electronics.
I deal with this grounding stuff all day long at work.


The problem with the AEM gauge/WB02 is that it has a single ground wire that carries both the high current for the heater and the ground for the output signal. This wire will have a 60 mV drop or so due to the high current, and thus the signal output voltage you measure at the end of said wire will have a 60 mV error.

This is why the LC1 has a separate high current power ground for the heater, and an output signal ground. This is good, but then they fracking tell you in the manual that the ends of these wires have to terminate at the same point on the engine. It's a bit dumb in that it should be designed such that it allows you to ground the high power ground to a nearby convenient chassis point, and then the output signal ground should go to the ECU signal ground.

One way to mitigate the AEM WBO2 problem is to parallel a thicker wire to augment the existing ground wire. The 60 mV drop may reduce to 20 mV or so.

I shake my head whenever I see newbie-engineer errors such as what AEM has done to their gauge and what I found when hacking their miata ECU.

Innovate used to have 3 grounds--heater, system, and sensor (analog), and the directions clearly stated to ground the sensor ground with the ECU grounds, and the other grounds on the engine. The problem was that it confused people. I can only assume that it was because people think ground is ground and wire stuff up without reading documentation beforehand. They now ship with 2 ground wires and the directions state to ground them to the same point on the motor. I believe they simply combined the system and sensor grounds together. At least it's on a separate wire from the high current ground.

True story - I work for a technology company and I see simple mistakes like this made all the time. Luckily it's solved usually before production versions. Just thought I'd post this up to help out. Anywhere for "important information" like this on the site?

my dash gauge and MS output different AFR values. Both are grounded together and are supplied the exact same 0-5v AFR range.

I had to tweek to MS's AFR calibration (offset by .2v) to have it match the gauge, which read richer, and was correct.

In my world, sharp analog engineers who understand grounding and especially grounding in a PCB layout, are rare.

70% of the problems I help our customers with, are PCB layout and grounding related.

AEM et al would do well to hire a consultant just to go over and review their analog designs, if they don't have a person with that kind of experience. But then, it could be a case of they don't even know what they don't know.

In for magical grounding locations that make MS, LC1 gauge, and Logworks all read the same.

That's funny, I think that's less confusing than the current instructions.

Hmmm . . . . I've got EXACTLY the same issue with my new LC-1 install . . . my dash gauge reads a bit richer. And, I took care with my grounds. Glad it's not just me (misery loves company).

Brain, how do you tweak the MS's AFR calibration? I used the configurator, and it seems to only give you a choice between several pre-configured input characteristics. As you know, I'm an MS1 (MSPNP).

You'd have to hand edit the file in a text editor (such as notepad).

You may wish to alter the LC1's output configuration instead.

I have MS3, so it's as easy as telling the MS that 0.2-5v = 7.35-22.23 AFR or whatever the scale is.

My gauge, MS and Logworks all read within 0.2 AFR of each other.

.2 difference here too.

Last two times I was on the dyno, the thing was pig rich. After going over logs and acutally watching the gauge on the dash I realized the discrepency. MS was fueling at what it thought was 11.8:1, where the dash gauge and dyno WB would show high tens.

I since reclaibrated the sensor, made sure the outputs were equal and got them much closer. Offsetting the 0v read to .2v got it damn close for me.

fuel mileage has improved :giggle:

I grounded my AEM at the recommended ECU grounding point and it was off by a steady 0.7-0.9 Grounded it to a point under the dash on the firewall and it's been perfect ever since. I just decided not to question it and enjoy.

Although I did discover a little trick to see how close they are (since it's sometimes hard to eyeball). I put a single large AFR gauge on a Tuner Studio custom dash and put the laptop as close to the physical gauge as possible. Then I started snapping pictures with the cell phone camera. Bingo, moment to moment comparisons.

That's a good idea, sure beats looking back and forth.

hmm my AEM is off by .1-.2 from MS to the AEM, currently its grounded to ECU adapter harness, but I'm gonna try firewall and parrlel wire running with it as well

My AEM is off by this same amount

What did you find by hacking their miata ECU? Anything that should/could easily be "fixed" by the end user?

I've tried several times to resolve the problem with my car; by changing grounds, adjusting the MS's AFR calibration, and even bought a 2nd wideband that did the same thing. My gauge is reading almost a 1 point discrepancy, gauge reads richer than MS.

I just wanted to clarify... In the event that there is a discrepancy between the AEM gauge and MS, it is assumed that the AEM gauge face is correct. Is that right?

I finally gave up on actually fixing the issue and adjusted my target tables a point leaner than they should be. I'd hate to think I'm tuning to a 12.5 afr at wot.

A multimeter tells the tale. MS has an offset.

Bunch of wimpy confused ground connections between the various ground pins on the connectors and the circuitry. I beefed em up with some 20 gauge wire. Fixed a problem where there was a 0.2V offset on the MAP sensor reading, which was worth almost a 10 kPa error (lame!).

Pulling a Jesus on this thread, but I now have the same issue. The discrepancy in my readings is of ~1.3 points. Which gauge reads correct? I dont see an answer for Pdexta. My WB is a PLX300 btw. :wavey:

I solved my problem by running the ground to the same source as my ECU. It still wasn't perfect but it was damn close. I also suggest compensating for the reading in your AFR target table.

DMM's solve all voltage offset problems...

I recently soldered all ground wires to ring connectors (rather than crimping only) and suddenly, the old offset in my WBO2 is gone. :wiggle:

Well, then who is the correct reading? TS or WB?

WBO2 101 here, how do I measure voltage differences. Also how to I "offset"... is there a table for this. I just went throught TS and I couldn't find such thing. My WB does not have the calibration software LC-1s have.

You're so gangsta

Thanks

WB is the correct reading
if there is different reading in TS thats because of a voltage difference from the output of the wideband to the megasquirt.

Okay, thanks. I now remembered that when I had the ground to a chassi port the difference was minimal. This time I kept the cable in the cabin. Time to re-ground.

You need to measure the voltage input into the MS. But you need to reference it against what the PLX is sending.

for exmaple, if you know the PLX should be at 5v and the MS is only getting 4.5v, you can recalibrate the MS to properly read and interperate the signal.

How? is there a voltage/afr table somewhere that I overlooked?

Me is a :noob:

oh youre still on one of those acient msI machines.

When there is a conflict between what is seen on the gauge itself and what is displayed by the ECU, always trust the gauge, and assume that there is a calibration or ground offset problem.

All else being equal, the analog ground for the wideband controller should always be connected to the same ground as the ECU uses for its analog sensors. Ideally, the wideband ground will actually be connected directly to a ground pin at the ECU.

Why not give it the "best" ground possible? Because if the ECU itself is experiencing any kind of ground offset, you want the wideband controller to experience the same offset. That way, the two devices are "in sync" with one another. On my cars, I have always either spliced into an ECU ground wire, or double-crimped the wideband analog ground directly into one of the ECU's own ground pins.

In cases where the wideband system has a seperate ground for the heater, this can be run directly to the engine. This is contrary to Innovate's instructions, however as Ben stated earlier, I believe they state this only to make the instructions simpler for people.

No, I have MSII.

Thank Joe

Logworks is the most accurate correct? I have TunerStudio and LogWorks displaying within 0.1 of each other, but my analog gauge is reading ~0.3/0.4 off. My grounding is superb. Thanks WB gurus.

so alter the output voltge range on that channel until it matches better.

Yes that would be quite easy to do. I just didn't want to alter the gauge if somehow it was more correct than LogWorks. This wouldn't make sense though as LogWorks should be the most accurate reading. Tonight I'll alter the gauge output by a bit to get it to match LogWorks and the MS reading. Thanks.