mj1176

Hey y'all. I've been reading til my eyes bleed and I can't figure this out. Here are the relevant specs of the car:

-0'90 NA6, naturally aspirated
-DIYPNP 1.5b assembled without vTPS jumpers, WITH sequential injection, extra outputs have been correctly run from the db15 connector to the injector harness
-Base map from DIYPNP instructions
-Cone-style intake filter fitted to the end of stock crossover tube
-GM AIT sensor from DIYPNP
-RoadsterSport cat-back exhaust
-AEM UEGO X-Series 30-0300 wideband gauge + sensor kit

Otherwise the car is stock.

Tune file here: Tune File
Log file for startup attempt here: Log

I loaded the base map to the DIYPNP out of the car with the wall plug, installed everything in the car, and calibrated the CLT/AIT sensors per the DIYAutotune instructions. Only other changes I made to the base map were the idle control settings from this guide as recommended in an older post here about a similar problem, I didn't try the Spark Advance or Fuel VE Table table changes on that page because I figured I should stick with what's in the base map from DIYAutotune. Anyway, I went ahead and tried to start the car and it starts but doesn't idle properly. As soon as it turns over and catches, it'll struggle for a second at ~400 RPM and then slowly climb to ~1000, then drop back down, and keep surging like that for 10-30 seconds before it dies. I checked the log file and it looks like TunerStudio isn't seeing a AFR signal, or rather, it's seeing a constant max signal as it's flat across the whole graph at ~20. To my knowledge the AEM kit is installed correctly: red wire goes to 12v power (I used the blue wire under the steering column), black wire goes to ground, and I spliced the stock O2 sensor plug onto the white wire and connected it back into the stock harness connector for the O2 sensor. The gauge itself reads fine and fluctuates as I'd expect for surging idle.

I'm not sure what to do at this point. Did I wire the AEM incorrectly? Anyone see any glaring problems with the fuel/spark tables in my tune file, or incorrect settings that would prevent TunerStudio from seeing the correct AFR signal? And if so, would that be an obvious cause for the behavior seen in the log file? I've done a ton of reading so I hope I'm not missing something stupid :(

Also, if there's anyone in the Austin, TX area who's willing to help me with basic tuning in person I'd be glad to pay for their time!

LimblessNinja77

iTrader: (1)

I'm not great with tuning and am a beginner but maybe you could start by checking your WBo2 sensor. One way to test if the sensor is dead would be to take it out of your exhaust (It should read full lean) then soak a rag in Brake clean, wrap the rag around the sensor and it should read full rich, taking the rag away should send it back to full lean. Also, I'm not sure where your ground is, but it's recommended that it's grounded with the ECU. (The ground wouldn't cause a full lean condition but it might effect it a little after you figure out the bigger problem).

mj1176

I don't think it's an issue with the sensor itself, since the gauge reads realistic AFRs as the idle surges. Seems like I just need to figure out why TunerStudio isn't seeing the signal from it, and if that's the sole cause of my idle issues. I do appreciate your input though!

thebigtuna

The DIYPNP might be way different than the MS3 I have, but I have an AEM as well. I had to run the white (?) wire coming out of the guage straight to the ECU. I didn't splice into the harness but rather into a wire coming off the ecu. Did you calibrate the o2 sensor to the ECU? I took this off DIY's site:
"Finally, you should calibrate your O2 Sensor to the ECU. To do this, click ‘Tools > Calibrate AFR Table’.

• Choose your O2 Sensor from the list. Choose Narrowband for the stock O2 Sensor. Or select your wideband and the proper configuration of said wideband from the drop-down list.
• Click ‘Write to Controller’. Once finished writing, click ‘Close’.

Like LimblessNinja said, you basically need to see if the AEM is reading correctly. If it is then you have an issue with Megasquirt seeing the correct signal. Does the o2 sensor read full lean as well? AEM will be more accurate than tunerstudio in a sense. Tunerstudio just sees what is being sent to it from the o2 sensor. Someone let me know if I'm way off.

I took this from another thread about hooking up a wideband to a diypnp:
"(wideband output) must be connected to the DIYPnP's O2 sensor input, and you must cut the signal wire from the stock O2 sensor rather than just splicing into it. The location of this line varies by model year, which you have not told us. "

mj1176

Thanks for the reply tuna. To my knowledge there should be no difference between running the white wire from the AEM gauge to the harness via the plug or by splicing it directly into the harness wire at the plug, but if I cant figure anything else I suppose I could try that. I did calibrate it in TunerStudio using the AEM X-Series setting (1.0 = 9.7, 4.0=18.7), and again the gauge itself reads completely realistic AFR numbers, completely different from the full lean that TunerStudio is seeing. So at this point I'm fairly certain either I'm missing something in terms of wiring or there's a setting in TunerStudio that's preventing it from seeing/interpreting the AFR signal correctly.

mj1176

Also I forgot to mention, I tried pulling the fuel pump relay to check the timing without starting the motor and TunerStudio wouldn't see the megasquirt when the fuel pump relay was out. Put it back in, TunerStudio sees megasquirt again. So I connected the timing light and got it to idle just long enough for me to check, and it's actually pretty dead on at 10 degrees. It would fluctuate a little bit during the low-side of the idle surges, but as soon as it got back up over ~500 rpm it would sit right at 10.

pakmx5

iTrader: (2)

on my DIYPNP and AEM units, i wired the white signal wire directly to MS as suggested in the setup manuals and from previous reading. It appears thebigtuna did the same thing. i would just stick with this since it is what's recommended.

mj1176

When you say directly to megasquirt you mean by splicing the white wire into the stock harness on the wire that o2 signal is jumpered to on the diypnp board? In my case it's 4N if I remember correctly

mj1176

UPDATE: I disconnected the white aem AFR signal wire from the stock harness, removed the stock plug, and connected the white wire to a pin on the DB15 cable. Then I opened the DIYPNP back up and changed the O2 jumper from O2 -> 4N to O2 -> pin 11 on the DB15 where I put the white AFR signal wire from the AEM gauge. I also enabled the "Incorporate AFR Table" setting in TunerStudio. Tried to start again and now instead of surging then dying it starts low, stumbles a bit, then surges up to about 1500-1700 rpm before coming back down and dying. I've attached the new log file below, I can see now that at least the AFR signal in TunerStudio is fluctuating, rather than reading a constant 19.1 flat across the graph. But the AFR gauge on the dashboard in TunerStudio still doesn't fluctuate and the car obviously still doesn't want to stay running. All I can think of at this point is that this model of UEGO has to have the brown signal ground wire connected, which seems simple and obvious enough, I just didn't use it because every install I read about with a UEGO seems to have only used the white wire for signal. I already have both SG points on the main board run to the adapter board per the instructions from DIYAutotune, if I want to send signal ground to the DB15 as well I can just add another jumper from one of those mainboard points to the pin I want to use on the DB15 right? Or do I need to find an unused signal ground point on the board to jumper from?

Log file: Here

mj1176

UPDATE ROUND 2: Hope yall aren't sick of me by now!

I went ahead and wired the brown (analog signal negative) wire from the AEM into a pin on the DB15 and jumped that pin to SG on the main board, tried to start the car again and now I have AFR signal going to TunerStudio! the car still doesn't idle but it's seeing AFR now at least so hopefully I can begin tuning the idle. Here's my final log file for the day, from when I tried to start it after connecting the signal ground...

Log File

I don't expect anyone to hold my hand of course, but now that TS is seeing a proper AFR signal if anyone could take a look at my datalog and point me towards a good place to start learning what I need to know to get it to idle happily in open loop mode I'd really appreciate it. I read through the tech articles on DIYPNP and I've seen the blogspot post everyone recommends about setting idle but I'm having trouble figuring out where to start.