Aem wideband don't match

  • 22 Replies
  • 3530 Views
*

hakcenter

  • developer
  • Hero Member
  • *****
  • 1785
Re: Aem wideband don't match
« Reply #15 on: November 04, 2015, 08:04:51 PM »
as long as they are very similar throughout the 10-14 range i would consider it close enough
AMP 2 Corinthians 8:21
For we take thought beforehand and aim to be honest and absolutely above suspicion, not only in the sight of the Lord but also in the sight of men.


*

eclipsegst1990

  • WEREWOLF Owner
  • Full Member
  • *
  • 87
  • "IF it can go wrong, it WILL"
    • Galant VR-4
    • 1991
    • Head: 1G (large ports)
    • Motor: 4G63 2.0L
    • Turbo: E316g
Re: Aem wideband don't match
« Reply #16 on: December 05, 2017, 02:25:45 PM »
Ok, well, sorry for bringing this topic back from the dead.  I was wondering if you could program in the ability to adjust lamda values so all of us with these ridiculous AEM gauges could find some peace.  For example, I want to try this solution to the problem in the linked video below.

http://www.dsmtuners.com/threads/ecmlink-linear-wb-calibration.428702/

When i set my idle to run at 11.0afr my voltage was 1.13 or 1.15 volts.  When I set the aem to provide volts for 14.7, I got about 2.46 volts. 

I'm sorry for exhausting this topic but I know there are a few people out there that are going gray with this problem.  I've since got a LC1, but I think the controller is shot (given to me), so I'm back to the AEM. 

If this doesn't work, I'm going to pull the sensor, place it in a gas can for rich voltage, then free air for the other voltage.
FIC-1150s, DeatschWerks 300, EGT gauge, AEM AFR gauge, Evo3-16g not ported, Tial 38mm wastegate on O2 housing, Accel 300+ wires, stock ignition, Arias 8.5:1 pistons, Ferrea valves 3mm over, titanium retainers, dual springs, stock cams, stock trans, 3" turbo back exhaust (Apexi-no cat), Ostrich2.0

*

hakcenter

  • developer
  • Hero Member
  • *****
  • 1785
Re: Aem wideband don't match
« Reply #17 on: December 15, 2017, 07:19:05 PM »
Can you get me the AFR to voltage table for the gauge ? Since it's not linear...
AMP 2 Corinthians 8:21
For we take thought beforehand and aim to be honest and absolutely above suspicion, not only in the sight of the Lord but also in the sight of men.


*

eclipsegst1990

  • WEREWOLF Owner
  • Full Member
  • *
  • 87
  • "IF it can go wrong, it WILL"
    • Galant VR-4
    • 1991
    • Head: 1G (large ports)
    • Motor: 4G63 2.0L
    • Turbo: E316g
Re: Aem wideband don't match
« Reply #18 on: December 16, 2017, 11:59:39 AM »
11.0 afr my voltage was about 1.13 or 1.15 volts.  When I set the aem to provide volts for 14.7, I got 2.46 volts.

**This doesn't match.  The 11:0 at 1.13v kinda matches the P2 column, but the 14.7:1 at 2.46v kinda matches P0 column.  I got this 14.7 voltage by disconnecting the gauge as stated in the gauge manual. 

Maybe I'm reading this all wrong.  Your 64bit brain might brain might be able to figure this out.  My 8 bit doesn't cut it, lol. 

Anyway, should we use my values and do something like what the video in this link suggested?  See reply #2

http://forums.ds-map.net/tuning/has-anyone-got-the-aem-35-8460-to-log-properly/

It should also be noted that there is a theory that the gauge analog output cannot drive the ecu directly or properly and that a buffer circuit may be needed, but I haven't seen anyone post anything about this working.
FIC-1150s, DeatschWerks 300, EGT gauge, AEM AFR gauge, Evo3-16g not ported, Tial 38mm wastegate on O2 housing, Accel 300+ wires, stock ignition, Arias 8.5:1 pistons, Ferrea valves 3mm over, titanium retainers, dual springs, stock cams, stock trans, 3" turbo back exhaust (Apexi-no cat), Ostrich2.0

*

hakcenter

  • developer
  • Hero Member
  • *****
  • 1785
Re: Aem wideband don't match
« Reply #19 on: December 17, 2017, 02:02:09 PM »
I need to find the super super old old aem pdf, cause those linear values don't cut it.

It is just (2 * x) + 10 for p0 / p1, which we all know is wrong.

Using 1.15 / 2.46 really makes it like,
(4.86586386*ln(x))+10.31993741

I have some old mmSD code that I used from the super old pdf and inputted everything by hand, in a 256 byte array from 0v to 5v.
Basically 0.01953125v per 1 unit, starting at 0
Code: [Select]
//AEM UEGO Unlinear
if(WB02Unit == PREFS_WB02_AEM) {
static const UInt32 interp[256] = {
8410,  8423,  8437,  8451,  8465,  8479,  8492,  8506,  8520,  8535,
8550,  8565,  8580,  8596,  8611,  8626,  8641,  8663,  8684,  8705,
8727,  8748,  8769,  8791,  8812,  8834,  8855,  8876,  8898,  8919,
8940,  8962,  8982,  8996,  9009,  9023,  9037,  9051,  9065,  9079,
9094,  9115,  9137,  9158,  9179,  9201,  9222,  9243,  9265,  9288,
9310,  9333,  9356,  9378,  9401,  9423,  9446,  9467,  9488,  9510,
9531,  9552,  9574,  9595,  9617,  9638,  9659,  9681,  9702,  9723,
9745,  9766,  9787,  9809,  9830,  9851,  9873,  9894,  9916,  9937,
9958,  9980, 10001, 10022, 10044, 10065, 10086, 10108, 10129, 10150,
10172, 10193, 10215, 10236, 10257, 10279, 10301, 10323, 10346, 10368,
10391, 10414, 10436, 10459, 10484, 10512, 10539, 10567, 10595, 10622,
10650, 10678, 10706, 10735, 10764, 10793, 10822, 10851, 10879, 10908,
10937, 10966, 10995, 11024, 11053, 11082, 11111, 11140, 11168, 11197,
11226, 11255, 11284, 11313, 11342, 11371, 11405, 11441, 11478, 11514,
11551, 11587, 11623, 11660, 11695, 11731, 11766, 11801, 11836, 11871,
11906, 11942, 11978, 12014, 12051, 12087, 12124, 12160, 12196, 12233,
12274, 12317, 12360, 12403, 12445, 12488, 12531, 12573, 12616, 12659,
12702, 12744, 12787, 12830, 12872, 12915, 12959, 13003, 13047, 13091,
13135, 13179, 13223, 13267, 13317, 13367, 13417, 13467, 13518, 13568,
13618, 13669, 13726, 13784, 13842, 13900, 13957, 14015, 14073, 14131,
14195, 14259, 14323, 14387, 14451, 14515, 14579, 14644, 14716, 14787,
14859, 14930, 15002, 15074, 15145, 15218, 15297, 15376, 15455, 15534,
15613, 15693, 15772, 15852, 15939, 16025, 16112, 16199, 16285, 16372,
16459, 16547, 16640, 16733, 16826, 16919, 17011, 17104, 17197, 17295,
17411, 17527, 17642, 17758, 17873, 17989, 18105, 18231, 18388, 18545,
18702, 18859, 19016, 19174, 19331, 20430
};
UInt32 v = (interp[d]) / 100;
return StrPrintF(s, "%d.%02d", (UInt16)v, (UInt16)(v % 10000));
}

I'll try to convert it today and see what I can come up with.

Basically unit, 58.88 which lands at 9.51:1, don't think that matches up to the gauge :/

V * 2 + 9.3 is closer but also not it. If you can... get more points so I can plot it out, it should start to match up with some equation.

Also something quite important about matching gauges, they must share the same ground as the ecu. So I would highly suggest grounding the gauge to the ECU ground, and not any random frame ground to help get consistency between the 2. The same goes for measuring voltage output from the gauge itself, make sure your ground on your voltage reader is to the ECU ground

I can tell you from personal experience working with my he351ve turbo controller, when I am programming and monitoring the device through the USB connection, my ISR for counting engine rpm, is different, when I am just connected through bluetooth (not using a different ground ie PC), so sources and grounds really matter in this type of stuff.
AMP 2 Corinthians 8:21
For we take thought beforehand and aim to be honest and absolutely above suspicion, not only in the sight of the Lord but also in the sight of men.


*

eclipsegst1990

  • WEREWOLF Owner
  • Full Member
  • *
  • 87
  • "IF it can go wrong, it WILL"
    • Galant VR-4
    • 1991
    • Head: 1G (large ports)
    • Motor: 4G63 2.0L
    • Turbo: E316g
Re: Aem wideband don't match
« Reply #20 on: December 17, 2017, 03:54:46 PM »
I am grounded to the ECU ground.  I tried that a while ago, but didn't fix the problem.  I will try to get a few more points.  Now the voltage readings are coming from the Werewolf logger.  If the analog gauge output is wrong with the AFR values, wouldn't the voltages be wrong too? 
FIC-1150s, DeatschWerks 300, EGT gauge, AEM AFR gauge, Evo3-16g not ported, Tial 38mm wastegate on O2 housing, Accel 300+ wires, stock ignition, Arias 8.5:1 pistons, Ferrea valves 3mm over, titanium retainers, dual springs, stock cams, stock trans, 3" turbo back exhaust (Apexi-no cat), Ostrich2.0

*

eclipsegst1990

  • WEREWOLF Owner
  • Full Member
  • *
  • 87
  • "IF it can go wrong, it WILL"
    • Galant VR-4
    • 1991
    • Head: 1G (large ports)
    • Motor: 4G63 2.0L
    • Turbo: E316g
Re: Aem wideband don't match
« Reply #21 on: December 17, 2017, 08:12:02 PM »
Thought this article on the AEM site might help a little.

http://aemelectronics.com/?q=forum/aemtuner-and-guage-not-reading-same
FIC-1150s, DeatschWerks 300, EGT gauge, AEM AFR gauge, Evo3-16g not ported, Tial 38mm wastegate on O2 housing, Accel 300+ wires, stock ignition, Arias 8.5:1 pistons, Ferrea valves 3mm over, titanium retainers, dual springs, stock cams, stock trans, 3" turbo back exhaust (Apexi-no cat), Ostrich2.0

*

hakcenter

  • developer
  • Hero Member
  • *****
  • 1785
Re: Aem wideband don't match
« Reply #22 on: December 18, 2017, 10:48:51 PM »
That guys was always off by 0.6, if this remains true to yours, you can setup a programmable, and make it match your gauge output, by varying the low / high until you reach a good linear line.
AMP 2 Corinthians 8:21
For we take thought beforehand and aim to be honest and absolutely above suspicion, not only in the sight of the Lord but also in the sight of men.