DIYPNP not reading AFR from AEM WB02
 

I have seen a lot of threads similar to this, but nothing quite like this.

1990 with DIYPNP built by me. 94 engine with turbo and RX8 yellow top injectors. Starts and runs with basemap, but runs pig rich and eventually stalls. Req fuel has been adjusted for the 420cc injectors. AFR table looks ok, I look at fuel table. I bump it down 10 points across the table (5 points at a time) and that helps it run longer, but it still eventually runs rich and stalls. If I remove too much fuel the car runs really lean and won't start at all. I bump the fuel back up. I use VEAnalyze but strangely, the numbers don't seem to change at any point in the rev range. I go to calibrate AFR. As we all know the AEM WB02 gauges are a total motherfucker to calibrate and I try a range of settings. But no matter what I do, the AFR gauge in tuner studio does not change while the car is running. So Tunerstudio always shows an AFR of 10, even when the gauge shows leaner by several points. I can get Tunerstudio to display a different AFR by changing the voltage settings in the AFR calibration window, but Tunerstudio's AFR gauge does not react or change when the car is running.

Thinking the gauge could be off, I test voltage output. It's within spec. (here's a neat link for testing your AEM gauge: DIY Testing AEM wideband gauges/sensors ) So it seems the ground is ok. It's grounded at the harness at the ECU on the black/green wire. I'm wondering if I have the white output wire from the AEM going into the right place at the ECU. I've tried both 2P (red/green) which is what flyin' miata suggests in one of their PDFs, and I've tried 2N (red/blue) which is what is identified in the m.net garage section somewhere. I have links if you want. Both 2P and 2N produce the same result - Tunerstudio still doesn't ever change its AFR. Which would jive with VEAnalyze never doing anything - if it can't read AFRs I don't think it's going to adjust. I guess it could also be the Bosch O2 sensor unit itself, I haven't tested that. But it was brand new.

What am I doing wrong? Tune file attached.

Your idle cells look very low. Are you sure it runs rich?
Just a pick a port under EGO Settings and feed the wideband output to it.

Which port though? It's set to Local, but there's also Remote ADC0, Remote ADC1, etc.
When you say "feed the wideband output to it".... how's that?

Thanks for replying!

And yeah, AFR is 10 on the gauge itself. You can smell the unburnt fuel in the exhaust. It runs really rich.

Do you have a map Hose connected?
Have you tested ignition and fuel on test mode?

Yes, the hose is connected to the Megasquirt and is routed to the recommended place near the FPR, tee'd into the line. I'm not familiar with test mode. I did get the car to idle last night by adjusting the fuel at low RPMs as you suggested. Tunerstudio still shows an AFR of 10 no matter what else I do in EGO Control settings or anywhere else. I'm wondering now if maybe I did a pull-up wrong when assembling the unit. I know I can keep playing with the fuel table to have a driveable car, but I'm hoping to get this AFR reading issue sorted out so that VEAnalyze Live can see actual AFRs and consult the AFR table to correct fueling for me. Obviously I'm not super great at this tuning business so I thought I'd get the software to help me out. Unfortunately since it's not reading AFRs correctly I'm just sort of bumbling around.

I'm having the same issue my my car. 1993, with a MSPNP2 and Innovate MTX-L WB. The gauge in tunerstudio stays at 10 no matter what fuel load or RPM even though my actual AFR gauge reads different. Gauge is grounded at the engine, timing is good, TPS is good, MAP line is good, but I'm also running pig rich (low 12s at WOT, mid 13s at idle and cruise)

Could it be adjustment of the TPS is needed?

I don't know much about RX8 injectors, but I hear the dead times on RX7 ones were extremely picky and have to be perfect to run smoothly
Could your req_fuel be incorrect for the injectors?
Did you calibrate the afr in TS properly?

instead of randomly connecting wires to random places, why don't you figure out exactly where your DIYPNP is wired to accept the wbo2 -- then connect to that?

It's not random, I researched it. The DIYPNP is set up for a 1990 harness, and I have two conflicting sources of information telling me where the O2 sensor comes in on a stock 1990 harness. I'm hoping someone can verify for me where they have their WB02 connected on their 1990 harness. In the meantime, I'm going to pull my DIYPNP unit and verify that the O2 connection in the unit is good.

I don't think it's the TPS adjustment, I was able to calibrate my VTPS without issue, and my car stays rich throughout the rev range.

Where do you have your WB hooked up to in the harness? Which wire? I have tried a couple, as I posted above. It seems like we're just missing a simple piece of information here that someone could easily provide.

I believe the req_fuel settings are correct. I had the RX8 injectors cleaned and flow-tested so I believe they are ok. I think the primary issue here is the DIYPNP not reading the signal from the gauge, which means the unit cannot be tuned in reference to the AFR table.

why are people connecting widebands to oem wiring harnesses? what am I missing here?
every..single...ms unit ever made has a specific wideband input that you can use specifically for this

Seen this a few times before

Odds are you have the old narrowband sensor connected as well as a break out for a wideband.

Disconnect the narrowband and see if it helps

it is random until you open up the unit that someone built and see EXACTLY where the wbo2 input is wired.

I could put a wbo2 input on the purge canister solenoid if I wanted to...

I'm probably not making myself clear here. I built the unit. It was earlier this year and I don't recall exactly where I set the 02 input, though at the time I thought I was wiring it to the stock location on the stock harness, which would make installation of the WB02 as simple as just finding that particular wire in the harness and grabbing a power and ground. Anyways I'm going to pull the DIYPNP and check and see. That was always going to be the next step, but I put up a thread to see if there was anything else I'm missing.

Maybe that's my mistake and it's ME missing something. I thought the process here was to build the DIYPNP like it was going to use a OEM O2 narrowband, and then just take the WB02 output from the gauge itself, and run that wire to where the narrowband would normally be sending its signal, if that makes sense. I just tapped into the OEM harness right at the ECU connector. This is my first DIYPNP and I couldn't find great instructions for doing things the tidy way with the DB connector. I'm not a wiring whiz. I did a lot of reading but after a while it all sort of blends in together. I PM'd Brain for one of his units many times and never got a reply so just doing the best I can. Thanks to everyone for your responses!

I'd say unless you plan to use the unit with the stock narrowband at some point, just yank it out, figure out where your wb input is, and run it separate and not through the oem harness. This way you can easily troubleshoot, and have a strong, clean, uncompromised input to the ms from the wb.

Getting ahold of Braineak is impossible. Just leave a message with one of his cats :likecat:

ahhhh ok. I'll check for that too while I'm in there. I don't think I understood that I needed to pick and choose which one was connected. I probably did both and have been sending the WB gauge signal into the narrowband connection, where the DIYPNP isn't looking for it.

My confidence level while building the DIYPNP fluctuated between "this is simple, I can build computers, this is easy" to "OMGWTFBBQ HOW DO I MAKE A WIRE TALK TO ANOTHER WIRE"

I will probably never use the stock narrowband so that makes sense. I'm guessing I need to use the DB connector to do what you're describing so I guess I'll have to figure that out.

Also, lol @message with cat. Maybe I'll send my cat over with a tiny scroll or something.

I have the MSPNP2, not the DIYPNP so idk about the wiring. On mine the o2 plugs directly into the WB gauge, and then a wire from the gauge goes into the stock wire for the o2 to go into the ecu. Stock location of o2 sensor right on the manifold, but on my stock ecu it runs fine, and then with the MS plugged in it runs rich. I'm also at a loss

Read the instructions

Per DIYAutotune's instructions, O2 on the DIYPNP mainboard is wired to 4N on the breakout board. This should be analogous to position 2P on the 1990 ECU connector - a red/green wire. If you're looking at the ECU connector, it's the bottom row, 7 holes from the left side. And I've already tried splicing the output wire from the AEM into that red/green wire. Never had any luck with it, getting Tunerstudio to show AFR other than ten, even with this setup. I guess I should note that while I did cut the stock 02 sensor off at its wire in the engine bay, that wire is still hanging out in there. I didn't cut it at the ECU, just spliced into it. Is that extra length of wire preventing the 5V output signal from getting into the DIYPNP somehow? Please forgive my extreme lack of analog electronic knowledge. I'm also happy to open the DIYPNP up again, remove the O2-4N connection, and wire the O2 straight to the DB connector, but I'd like to first understand why my current setup isn't working. Thoughts?

bump

I followed DIYAutotune's instructions when I built mine for my 2000. I just hooked up the stock narrowband input wire to the AEM gauge's output wire and that was correct. I ran the AEM wire into the engine bay, spliced the stock connector (from a dead O2 sensor), onto it, and then plugged it into the stock connector (harness side).

Also, it should be a red blue wire, unless you're blue green color blind :) . http://www.yorbalindamiata.com/images/wiring/90diag.pdf see page 4

I'll try the red/blue wire again, but I'm not getting why the AEM wire has to go all the way to the engine bay instead of just being spliced into the red/blue wire at the harness. Can you enlighten me on that?

Doesn't matter where it connects, just made more sense in my head to install it there. Let's say I needed to unplug the O2 signal for whatever reason, instead of ripping apart the harness again, I can just unplug a connector :)

SOLVED
 

I ended up setting up the DB15 connector to have 12V, ground, signal ground (in case I need it in the future), and 02, and went through that. Bam, Tunerstudio AFRs actually reflect what the AEM gauge is telling me. Calibration was slightly off but I fixed that quickly and now it matches perfectly. Now I can actually tune the car. Thanks everyone for all the help!