RomRaider

Just got around to installing two PLX sensor boxes, EGT, and Wideband AFR. I used the daisy chain setup to install 2 DM-6 gauges in addition to sending the analog 0-5v analog outputs of th PLC boxes directly into the Subaru ECU via the left and right Tumbler angle inputs. I wanted to keep the engine bay clean so I installed the sensor boxes right next to the cars ECU. I cut and repinned the connectors for the sensor boxes so i could use an existing grommet to get the wires inside. On the 2005 Turbo legacy I located the input pins as Connector C135, pins C26 and C27.

I even tapped the ECU switched power since these sensors don't use very much power. The power came from Connector B136, Pin B4 for ground (Black with white stripe). Switched +12 volts on Pin B5 (Yellow with light Blue stripe) located right next to the ground. The interesting thing about this is, the gauges work as long as the ECU is on, so they remain on for a few seconds after the vehicle is off due to the timed delay.

In romraider I added a few formula's. Some from this thread and one taken directly from the manuals that came with the gauges. For EGT I did a little math to arrive at the correct formula and the numbers seem to match up perfectly with the DM-6 gauges. The "C2" is the formula for the stock EGT sensor that I threw in just to take a peak at. But the PLX sensors seem to be scaled from 0 to 1500C with the 0-5v swing. More testing when I do some good data logging.



I wanted the EGT sensor before the turbo and didn't feel like taking off the turbo and up-pipe, so I drilled and tapped the stock Subaru exhaust manifold right near the stock AFR sensor. Just used a grinder to make a flat surface on it, turned on the engine and drilled it out with a bunch of sticky tapping oil to make sure the shavings blew out at me rather than go through the turbo.

Other things of note, I also tapped the ECU wires for the tachometer/injector to send to my remote start alarm setup. I even tapped the neutral safety switch which the manual transmissions provide on these Subaru's, however I had to make a transistorized circuit to correct the neutral safety switch values so that the alarm system knows for sure if the car is in gear or is in neutral before the remote start will activate. And finally, the E-brake wire is a secondary back-up for the remote start signal to make doubly sure the remote start will not active if the vehicle is in gear or if the E-brake is not set.

nice work

Has anyone tried using their Rear O2 as an input that can chime in on this? Apparently the Rear O2 input can only read 0-1.5V, unlike the rest of the 0-5V inputs like the TGV's/Fuel Level Sensor Inputs. Which I don't have due to having a JDM EJ207.

I have only found one post confirming this found in the page below.

http://www.romraider.com/forum/viewtopic.php?f=7&t=6430&start=15

My concern is with sampling size. If the ecu reads in 0.02V increments as stated in this thread, I will have a lot less accuracy when I scale my WBO2 from 0-1.5V than 0-5V in the logger definitions.

Any help will be appreciated.

I didn't find an exact thread - but this looks pretty close..
If you have TGV removed, you can wire up a GM IAT sensor between the TGV 5V output and the TGV angle sensor input (0 - 5V).
Basically any 0->5v two wire sensor will work (Oil pressure, IAT, AFR, etc..) and one just needs to update the logger file to map the volt to an actual value.

Correct?

Thread Revival.

Thanks for those that have kept this up to date with their findings.

First attempt at changing my SD-logger XML and would just like to know if my approach is correct on this.

I am using a Zeitronix ZT3 0-5v output into Stock ECU on MAF 0-5 In as I am on SD.

Manufacturer website states formula for voltage to AFR is as follow AFR= 2 * V + 9.6

Does this look correct?

<parameter id="P12" name="Mass Airflow" desc="P12" ecubyteindex="9" ecubit="4" target="1">
<address length="2">0x000013</address>
<conversions>
<conversion units="AFR" expr="x/50*2+9.6" format="0.00" />
</conversions>
</parameter>

Just an update on this, I may need to login and use the zeitronix logger as my gauge is idling at 14.6-14.7 but my RR is showing 14.7-14.84.

Not sure if the gauge is incorrect or the formula.

Ok so they are pretty close, perhaps there is a delay when using the RS232 converter vs direct to ecu.

Thoughts?

Yup that's normal. Try it on WOT. If it shows the same AFR. Then it's gold. Other wise follow the gauge as it will be more accurate and adjust the formula by the difference constant.

Thanks Welaish, Will Log tonight and have a look.

bringing this back. My old cobb setup had an AEM wideband analog out rewired to the rear o2 sensor input and tuner used that. So it can be done on a 07 STi ECU at least. But currently I'm trying to figure out where do I edit the table to account for this in my new open source tune? I need to get it to use the new conversion based on the AEM vs the OEM.

I think I've figured out logging with the rear O2 for 0-5V.

I know this is possible since COBB does it with AccessPorts. In AccessTuner you can set a scaling for the rear O2 sensor for 0-5V. With the new Cobb green speed update I can't confirm anything with AccessTuner since this feature has been removed, but here's a link from Nasioc showing it.
https://forums.nasioc.com/forums/showthread.php?t=2805570



Based on what dschultz said below for editing the logger.xml, the "x/50" is to convert 0-255 bytes to 0-5V. In the logger.xml the rear O2 is "x/200" which equates to 0- 255 bytes to 0-1.275V, which is similar to what Mike96 discovered, and is in line with the speculated 0-1.5V. It makes sense that subaru would scale for 0-1.5 V if 1.5V is all the sensor would outputs as this gives better resolution

So for using the rear O2 for logging change "x/200" to "x/50" it should scale for 0-5V. I currently can not test this as I'm waiting for my 2007 Legacy Spec.B wagon to arrive from Japan

Original Logger Def


Change Def to the below for 0-5V logging

[Wideband scaling multiplier] and [Wideband offset] will depend on your wideband being used, check with your manufacturer

I've been using the x/200 and it gave me real crappy readings and thought my wideband failed lol, glad I found these threads. Will be testing the parameter below tomorrow.

AEM inline wideband kit.

<parameter id="P15" name="AEM Wideband" desc="P15" ecubyteindex="9" ecubit="1" target="1">
<address length="2">0x000018</address>
<conversions>
<conversion units="AFR" expr="X/50*2.3750+7.3125" format="0.00" />
</conversions>

Did anyone figure out if logging a wideband analog 0-5v output on the rear O2 signal wire is doable?

I connected my analog output 5v (10AFR = 0V; 20AFR = 5V) wire to the rear O2 signal wire and grounded it to the ECU. I set up the formula in the logger xml to x/200*2+10. It seems to be working ok. However what I am wondering is, does that mean it's scaled to 0-1.5v instead of 0-5v? Using X/50, I cannot get it to output properly. Is that because the ECU is only looking for 0-1.5v on that circuit and thus I am restricted?

2007 STI

Take with grain of salt - From my understanding - /200 will give you 0-1.275 scale, and /50 will give you 0-5.1. the rest of the formula needs to be redone to scale properly as well, did your wideband not come with instructions and formula?

Yeah I wish they gave me a formula but I would figure its a simple y=mx+b. With x/50 that would give y=x/50*2+10 for 5V but what do I know?..

WIDE BAND ANALOG OUTPUT 0-5V (0V = 10AFR, 5V = 20AFR).

Linearity Scale
VOLT GASOLINE AFR
0.00 10.00
0.16 10.32
0.31 10.62
0.47 10.94
0.62 11.24
0.78 11.56
0.94 11.8
1.09 12.18
1.25 12.50
1.40 12.80
1.56 13.12
1.72 13.44
1.87 13.74
2.03 14.06
2.18 14.36
2.34 14.68
2.50 15.00
2.65 15.30
2.81 15.62
2.96 15.92
3.12 16.24
3.28 16.54
3.43 16.86
3.59 17.18
3.74 17.48
3.90 17.80
4.06 18.10
4.21 18.42
4.37 18.74
4.52 19.04
4.68 19.36
4.84 19.66
4.99 19.98