Midnight Caller

  • Posts

  • Joined

  • Last visited

  • Days Won


Midnight Caller last won the day on March 10

Midnight Caller had the most liked content!

About Midnight Caller

  • Birthday 03/28/1981

Contact Methods

  • Website URL
  • Yahoo

Profile Information

  • Gender

Previous Fields

  • Location
    Bucharest, Romania

Recent Profile Visitors

4,709 profile views

Midnight Caller's Achievements


Newbie (1/14)

  • Dedicated Rare
  • First Post Rare
  • Collaborator Rare
  • Posting Machine Rare
  • Week One Done Rare

Recent Badges



  1. Hello ! As I've read all over this topic, pin A19 is definitely not recommended for the WB connection. I've completely disconnected mine from the PCB. My AEM UEGO WB has only one ground and I've connected it to the battery minus lead directly. @venderbroeckand @Piet's mods are not public, who has them bought them. I for one bought the WB mod and the Boost threshold mod but still having issues with them. I think I'll ditch them and go for Aaron's ( @Tightmopedman9) solution, as none of them are replying my messages... Didn't really catch your last phrase ( "the cars rear or my external wb ") Later edit : From what I've saw on the PCB, the power ground plane is on the back of the ECU and signal ground plane is on the upper side. They are tied toghether with through-holes, so from my measurements power GND and signal GND are connected to each other. It's clear that I don't recommend connecting anything that draws a high current on a signal GND pin, but for what's worth, connecting something like a boost pressure sensor ground can be done on any ground pin that is closer / easier to solder on. Please correct me if I'm wrong (I've designed PCB's for ON Semiconductor for 8 years, but I still might miss something )
  2. For what I can say, I also have the european model with IMMO3 and had no issues with flashing apart from some 10-20% write fails. I just stop the flashing process, restart it and it succeeds. Maybe worth cleaning your ground points on the chassis too (measure the voltage drop between the battery minus lead and the ECU GND - it should be less than 100mV with ECU operational)
  3. OUT - is connected to the ground plane anyway as far as I remember (connected to IN -) so it shouldn't matter. Anyway, seeing it work perfectly on the bench every time but not working in car baffles me
  4. I don't know what to say here but.... I always disconnect the fan when flashing (I disconnect the black wire connector that goes on top of the fan shroud). When flashing, the lambda sign should stay lit ! If it's flashing, it's gonna fail so you could just stop the process and start over again, no need to wait. Try disconnecting the fan as that draws pretty much current from the battery and (could) influence the flashing process. I myself have like 3 failed flashes on every 10 flashes I do.
  5. I'm not having any differences between the gauge and the logged values anymore, I fixed that yesterday. I am using the right formula which also Piet/venderbroek suggested when they sent me the modded BIN The issue now îs that the logged AFR isn't oscillating around the requested AFR in the VE map (15) but rather around 14.7 (maybe a coincidence ?) I'm.using the AEM UEGO 30-4100 or 30-4110 with Bosch LSU4.9 which outputs 0-5V for AFR 10-20.
  6. Heeeey ! Piet, I've been trying to reach you for a long tine ! Everything ok ? As I was writing earlier, it seems like the WB mod ignores what is set in the VE map. Could you lend a hand please ? I had some issues with the modded BIN from the first release you sent me (car didn't start), and then the second file worked. Maybe You missed something in there ? Can I send you the BIN to take a look ?
  7. Hello ! Thank you for the detailed answer ! Basically shouldn't the wideband mod's purpose be to "move" the oscillation of reach/lean around the point set in the VE map ? If we set an AFR of 15 at the correesponding RPM/LOAD columns, shouldn't STFT flip point be around there ? As you are in the short clip I posted, It seems the changing STFT point îs around 14.7 not 15 (requested AFR, IT seldomly reaches that value). Thats's why I question If the mod is really working as expected.
  8. Well, after 5 years (of which 3 the car was unused because a tree fell over the top), the wideband ordeal IS OVER !!! I want to personally thank each and everyone of you that tried to help me out of this ! You guys are awesome !!! Ttoday, after fixing all vacuum related issues, and still didn't manage to obtain the right AFR, I decided to pull the ECU out and take it apart as I've forgot what I've done 5 years ago. The wiring was like this : - The AEM UEGO has 4 wires - black (-) / red (+) / white (0-5V output) / blue (serial output) - The + wire of the controller was connected to the +15 relay (yellow in the fusebox) to be able to shutdown the controller when I turn off the engine. - The white wire was tied to the Signal (+) wire of the Rear O2 sensor harness. - Because I didn't have two ground wires, I decided to tie the black (-) wire to the Rear O2 sensor harness. In the ECU i've cut the A19 pin (Rear O2 signal -) and tied it to ECU GND (pin A42). This way I would have the same ground for both the ECU and the AFR controller. But this seemed to have been a bad decision, as I've re-read almost all of the 500 pages here and I think that only by chance I've seen that the ADC in the microcontroller has a 0.2V offset in respect to ECU GND. Because of this, ECU measured +0.2V more than the wideband controller's output and hence a leaner measured AFR and thus enriching the mixture which in turn made the wideband read a richer mixture. Knowing this now, I've disconnected the Rear O2 sensor Signal (-) pin A19 and tied it to pin A18 (sensor ground), and tied the AEM controller directly to battery minus pole. The offset seems that's gone now, and the formula in the ADX (X*10/255+9.6) now reports exactly the same AFR as the wideband gauge. I also calibrated the injectors and now the AFR sweeps from 14.4 to 15.0 at idle. I still have to understand how Piet's wideband mod works. In the VE MAP, at 800-1000RPM and 0.4-1 load, I've set an AFR of 14.9->15.6 (see attached image) Still, even if in the dash I see requessted AFR of 15.1, when looking at the STFT, when AFR goes over 14.7, STFT goes positive up to 3%, and when AFR goes back under 14.7 the STFT goes negative to -3%, and the cycle repeats. I wonder how does this wideband mod work after all, if ECU does not target requested the values in the VE MAP ? I've also attached a small movie of the TunerPro dashboard :
  9. Well, 20 minutes to take it down and 20 to put it back is not such an ordeal It seems that the PCV hose was only out of the house coupling. Put it back together and was good to go. Unfortunately that wasn't the reason if my AFR issues. I plugged the BOV, and pressure tested the intake system, no vac leaks. Tried calibrating the injectors again, I used the injector constant and the dead times in the M44 wikia (0.7188 and dead times from the wiki). Vac. is around -10, STFT is oscillating between -6% and 6%, MAF meters 13.8kg/h, load around 1.01. Requested AFR from VE map is 15.1, but AFR on the gauge sweeps between 14.1 and 14.7 I removed the formula in the ADX to see what's in the ADC bits, and it oscillates between 120 and 130 If the wideband is calibrated in the BIN for 0-5V to correspond to bits 0-256, then: 5V/256bits = 0.01953125 volts/bit The middle of the "scale" - 2.5V would correspond to bits 128 (AFR=15 on the gauge) As I wrote earlier, I'm seing values between 120-130 in the TunerPro dash (using no formula, just "X") and 14.1-14.7 on the AFR gauge. AFR readings on the gauge : - 14.1 AFR on the gauge corresponds to approx 2.1V output to ECU - 14.7 AFR on the gauge corresponds to approx 2.35V output to ECU AFR read by ECU (shown in TunerPro) : - 120*0.01953125 = 2.34V -> which corresponds to 14.68AFR (according to AEM UEGO manual) - 130*0.01953125 = 2.53V -> which corresponds to 15.02AFR (according to AEM UEGO manual) So as far as I see I'm having around 0.23V offset between what the AFR controller sends to the ECU and what the ECU translates as in bits (and reapply transformation formula back to volts). I'm pretty sure that I tied internally (in the ECU) the (Signal -) pin of the rear O2 sensor (A19) to ECU signal ground (A18) so.... calibration problem in the BIN file ?
  10. Ok, back to square one. Removed the dump valve, same thing, so I started the hunt for vacuum leaks. And Lo, behold, the hose from the PCV canister to the intake duct îs split in two. Manifold down time
  11. Wait, what ?! You may be right, and the BOV be actually leaking air OUT, not IN the system, as it's before the throttle plate. You deserve a minimum sixpack of beers for that bitchslap you just gave me
  12. Exactly. That's why I think maybe leaning the whole VE Map by the offset I have would "fix" this situation I also did some aproximations regarding the dump valve. (Unfortunately I didn't have the time to take it out today, I helped a "brother" who bought an 1994 850 2.0T5 with 100.000 miles in mint condition, but one of the heater core hoses exploded on the way home. Yeeeey for April fool's ) If the dump valve would leak under vacuum, that would mean that the MAF would record less air than in reality (but mine measures 14-15kg/h at idle). Furthermore, the extra unmetered air should lean the mixture, not enrichen it. If I would've had a lean condition, the dump valve could be the reason but in my case it's the other way around, the mixture is too rich, both at idle, cruising and full boost.
  13. Hello ! Thank you so much for confirming. I was really intrigued by the -100/+99.2 STFT values. You are perfectly right about the dump valve, but the issues I'm fighting now are when idling. Yes, when dumping the extra air after being in boost, extra air measured by the MAF will lead to a pig rich condition for a short time. The issue I'm facing now is when idling or not under boost condition, the VE map cell / requested AFR is (let's say) 14.9-15 (low load, cruising, tuned for improved mileage), but the AFR gauge swings between 14.2 and 14.7. It's this offset I don't understand where is coming from, since with the injector calibrator BIN + software I'm seeing on the PC what I'm seing on the AFR gauge (max. 0.1-0.2 difference, but mean of 10 measurements is what the gauge shows). That's with the dumpvalve in place. I will try to plug it tomorrow and see if it changes anything. The rich condition is reflecred by the plugs (pulled them out yesterday and they were a "puffy" black (indicating a rich mixture) I COULD compensate by further leaning the VE Map, but as Piet or Venderbroek said at that time, it wouldn't work and just lead to futher STFT being unstable. I received my modded BIN from them in April 2016, and as I understood, in September 2016 they solved the AEM calibration issue... that's why I was asking you for a 608 BIN with wideband enabled (I also have the boost threshold mod but not sure if I'll use it in the future). I know the wideband mod was purchased by you as it was by me, so I understand if you don't want to make it public... So please PM me if you can lend me a hand Or I could send you my BIN and see if you can find a difference ? Gdog helped me with his 607 BIN and I thank him greatly, but there are differences I dont have the experience to identify Thank you !!!