Logging Wideband O2 Via The TGV Sensor Input

[SBT]

Parenthetically, would it be wise to seal-off the unused TGV connectors from the elements? If so, what's the recommended approach? Something like the flexconnector.com 5-pin connectors to close-off the old connection, then seal-off the oppositive end of that new 5-pin connector?

[fahr_side]

I put plastic bags over the connectors and sealed them with electrical tape. In hindsight I could have stuffed them with dielectric grease first.

[Turkeylord]

Since I wired at the ecu, I simply plugged them back in to the motor and sensor.

[SBT]

Since I wired at the ecu, I simply plugged them back in to the motor and sensor.

 

So you detached the motor and sensor from the stock TGV and just left it locally in place?

[Turkeylord]

Deleted the rods, plates, and dividers, welded the holes, then reinstalled the motor and sensor on the gutted TGV housing. Looks stock.

 

I think I might've removed the electrical contacts from the motor and sensor.

[snow05gtRI]

No, they're for the later 5-pin TGV motor and sensor combined.

 

can it be used on the 3-pin?

[fahr_side]

can it be used on the 3-pin?

 

 

 

The plugs are very different.

[stikiller]

Update I've wired everything up the gauge works I grounded it to the battery ground that runs to the left side fender wall right by the fuse box. But now I went to start my motor last night and I get no spark to the plugs. I have fuel and power to the injectors but no spark! Timing is dead on. The crank sensor worked before the build. I don't know where to go from here. I also have the ground for my defi unit going to the same place. Could it be I have to many grounds going to one spot? I'm gonna back up and take the grounds off there when I get home and try that. That's the only thing I have changed since the build. Edited August 24, 2016 by stikiller

[stikiller]

OK pulled the timing cover off and double checked. The timing is good but I do have a broken tooth on the crank sprocket so that's the culprit.

[BarManBean]

OK pulled the timing cover off and double checked. The timing is good but I do have a broken tooth on the crank sprocket so that's the culprit.

 

Need a crank sprocket? PM me if you do.

[stikiller]

Well got it running but it won't idle. Hopefully the tuner will be able to get it straightened out tomorrow.

[Underdog]

I'll be wiring my AEM UEGO and fuel pressure gauge analog outputs to the TGV sensor inputs sometime this week. I have the gauge wiring in the cabin already (at the fuse box) so I'll be running it behind the dash and tapping in at connector B136 at the ECU.

 

Pinout for my '06 LGT B136 connector: TGV RH pin #27 (light green w/ yellow stripe) for the UEGO, TGV LH pin #26 (red w/ black stripe) for the fuel pressure.

 

I have already updated my RR 0.5.9RC3 logger definition (logger_STD_EN_v317.xml) with the following code. Scaling is direct from AEM literature.

 

<parameter id="P39" name="A/F Wideband (TGV R)" desc="P39 - AEM UEGO AFR=2*V+10" ecubyteindex="12" ecubit="0" target="1">
<address>0x000033</address>
<conversions>
<conversion units="AFR" expr="2*(x/50)+10" format="0.00" />
<conversion units="V" expr="x/50" format="0.00" />
</conversions>
</parameter>

<parameter id="P40" name="Fuel Pressure (TGV L)" desc="P40 - AEM FUEL PRESSURE (0-100PSI) PSI=25*V-12.5" ecubyteindex="13" ecubit="7" target="1">
<address>0x000034</address>
<conversions>
<conversion units="PSI" expr="25*(x/50)-12.5" format="0.00" />
<conversion units="V" expr="x/50" format="0.00" />
</conversions>
</parameter>

Edited December 12, 2016 by Underdog

[Tuning Alliance]

Tuning Alliance has come up with a kit to log your wideband via rear 02!

 

https://tuningalliance.rpmware.com/tuning-alliance/rear02-logging-kit-innovative-lc2-sensor/tawb02/i-3708881.aspx

[utc_pyro]

Adding to the logging new sensors over TGV input, I built a plug and play fuel pressure logger last night. It's similar to the Cobb kit, but will actually work with 05-05 LGT's and is half the cost.

 

http://i.imgur.com/X73rGYf.jpg

 

Parts list:

Sumitomo 3 way TS Plug

1/8 npt 100psig 5V pressure transducer

Aeromotive 15120 quick connect fuel tap

 

Installed it tonight.

http://i.imgur.com/lQcD29A.jpg

 

Based on the spec sheet the sensor reads .5v at 0psig, 4.5v at 100psig, and is linear between those points. Without fuel lines it read 0.5-0.52v, so 0psig is close to spec. Thus to convert the logged voltage to psig:

(Left_TGV_V - 0.5)*25psig = Fuel Pressure

 

Attached is a 3rd gear hill pull with this logged. Car is running ~E50 fuel, DW65C pump, DW740 sidefeed injectors, stock rails, BNR16G turbo at wastegate boost. ~14.5psia local pressure.

 

Edit: This is a graph of the fuel pressure through the TGV

http://legacygt.com/forums/attachment.php?attachmentid=243331&stc=1&d=1484244146

romraiderlog_20170111_185208.csv

Edited January 12, 2017 by utc_pyro

[Underdog]

Nice work. I did similar, but via the AEM gauge in the cabin.

[utc_pyro]

Nice work. I did similar, but via the AEM gauge in the cabin.

 

Underdog, how much does the data point jump around in your logs? Mine isn't exactly smooth, but I don't know if that's due to a sensor/electrical noise issue or actual fuel pressure fluctuations.

[Underdog]

It definitely doesn't stay as steady as I would have expected, having used mechanical gauges for diagnostic purposes in the past. I would say it has a similar scatter to the data in your log, but I've only tested for functionality and haven't logged any pulls. The pressure displayed on the gauge also jumps around, and has been that way since before sending the 0-5V output to the TGV sensor input on the ECU. I also don't have any pulse dampers, so it will be interesting to see how the pressure on my car compares when I have a chance to pull a similar log.

 

Given the way the scatter increases towards the end of the pull I would guess we are seeing the effect of a smaller relative line volume as compared to injector demand.

 

BTW - how did you end up with 45% and 85% for FP duty cycle steps?Also, any idea why your log dropped out twice in that run?

Edited January 12, 2017 by Underdog

[heiche]

utc_pyro, have you done the mod where you re-route your FPR signal hose from the stock single intake manifold runner to the BPV signal hose (via a tee)? It made the fuel pressure reading a bit smoother, although still more "noisy" than I'd expect it to be.

 

The more consistent fuel pressure actually helped driveability a bit, the better looking logged fuel pressure was just a by-product of that.

[utc_pyro]

It definitely doesn't stay as steady as I would have expected, having used mechanical gauges for diagnostic purposes in the past. I would say it has a similar scatter to the data in your log, but I've only tested for functionality and haven't logged any pulls. The pressure displayed on the gauge also jumps around, and has been that way since before sending the 0-5V output to the TGV sensor input on the ECU. I also don't have any pulse dampers, so it will be interesting to see how the pressure on my car compares when I have a chance to pull a similar log.

 

Given the way the scatter increases towards the end of the pull I would guess we are seeing the effect of a smaller relative line volume as compared to injector demand.

 

Interesting... Sounds like it's OK electrically, just Subaru fuel pressure is really unstable. I'd be interested to see what yours looks like, because I’m kind of surprised these cars run as well as they do if this is accurate.

 

On the fuel line size thing, are you saying that the lines have too much of a pressure drop for the demand or there is a dynamic effect going on as the injectors open/close? My knowledge of fluid dynamics only covers steady state.

 

BTW - how did you end up with 45% and 85% for FP duty cycle steps?Also, any idea why your log dropped out twice in that run?

 

There is a table for FPDC that NSFW found, bumped it up based on DW recommending 6v minimum on their pumps. I suspect my openport's USB port is failing or a software issue on my tuning laptop.

 

utc_pyro, have you done the mod where you re-route your FPR signal hose from the stock single intake manifold runner to the BPV signal hose (via a tee)?

 

Yep, it's actually fed off the vaccume distribution block on the IC bracket (blue/silver thing in second picture). It also has a vacuum signal filter similar to the one on the MAF sensor to smooth things out even more.

Edited January 12, 2017 by utc_pyro

[snow05gtRI]

those dropouts may not be related to the openport or the laptop. I get similar dropouts on occasion. They last 1-2sec, like yours. I did a bunch of reading on it, other people have had similar problems. The belief is an electrical gremlin causing the data transfer to hang up in the ECU. Supposedly someone replaced their ECU and that fixed it, and others claimed relocating the grounds fixed it.

 

does that connector you're using for the FP sensor plug right into the TGV plug? I've still not wired up my WBO2 and that might make it much cleaner...

[utc_pyro]

those dropouts may not be related to the openport or the laptop. I get similar dropouts on occasion. They last 1-2sec, like yours. I did a bunch of reading on it, other people have had similar problems. The belief is an electrical gremlin causing the data transfer to hang up in the ECU. Supposedly someone replaced their ECU and that fixed it, and others claimed relocating the grounds fixed it.

 

does that connector you're using for the FP sensor plug right into the TGV plug? I've still not wired up my WBO2 and that might make it much cleaner...

 

Do you remember where the grounding issue was? As in ECU ground, ODBII port ground, motor ground, ect? Several of my sensors appear noisy, so there may be something to grounding issues.

 

Yep it plugs right in, you just have to remove one tab from the bottom with a knife. If you already have a wideband analog output in the cabin it’d be quicker to tap into the ECU wiring harness directly though. With a posi-tap it’d be a sub $10, sub 15 minute job.

[Underdog]

Interesting... Sounds like it's OK electrically, just Subaru fuel pressure is really unstable. I'd be interested to see what yours looks like, because I’m kind of surprised these cars run as well as they do if this is accurate.

 

Yeah I don't think it's signal noise, but it's possible the ADC inside the ECU is not set up for the kind of resolution warranted for this type of sensor. I'll log and post up next time I have a chance. Perhaps I'll try hooking it up to my AQ-1 and see if that looks different.

 

On the fuel line size thing, are you saying that the lines have too much of a pressure drop for the demand or there is a dynamic effect going on as the injectors open/close? My knowledge of fluid dynamics only covers steady state.

 

Was just shooting from the hip - thinking perhaps the scatter was a cumulative effect of injector pulses within the sample interval, setting up a kind of phantom harmonic. With larger lines the effect would be less noticeable, but I haven't done any kind of sanity check on the premise.

 

There is a table for FPDC that NSFW found, bumped it up based on DW recommending 6v minimum on their pumps. I suspect my openport's USB port is failing or a software issue on my tuning laptop.

 

I mapped the table but couldn't find a recommendation for the DW pumps. Do you recall where you found the 6V recommendation?

Edited January 12, 2017 by Underdog

[snow05gtRI]

Do you remember where the grounding issue was? As in ECU ground, ODBII port ground, motor ground, ect? Several of my sensors appear noisy, so there may be something to grounding issues

 

When I was researching it for myself, I had PM'd seabass07 because I saw him comment he had the same WOT logging dropout issues that I had in post#4 of this thread: http://legacygt.com/forums/showthread.php/vag-com-cables-chipsets-ft232rl-ch340-214994.html

 

seabass07's response to me was:

I never really found a solution. Strizzy ended up getting a new ecu and that solved it. Some fixed it byvreolacing the grounds. I just stopped messing with it since my tune was good.

 

post #22 in that thread, iNVAR says:

I actually used to have this problem on my cable too but it seems to have gone away completely for the last few years. Don't know what ended up fixing it. It definitely was NOT the adding of the ferrite core although I don't think that hurt.

 

The consensus is that it is some sort of electrical interference, whether a bad ground, or voltage spike, or EM noise or something else.

 

A few people have modded their cables using a capacitor I think and it seems to have worked.

 

the rest of the thread is just about the VAG-COM chipsets. This was as far as I got with my research on WOT dropouts. I haven't fixed mine. I had some engine grounds reworked recently, but I haven't had a chance to test if anything is better yet since one of the wires on my EBCS mysteriously, very cleanly, became severed. I've also been on vacation.

 

 

Yep it plugs right in, you just have to remove one tab from the bottom with a knife. If you already have a wideband analog output in the cabin it’d be quicker to tap into the ECU wiring harness directly though. With a posi-tap it’d be a sub $10, sub 15 minute job.

 

I don't have one in the cabin. I have the AEM UEGO WBO2 sensor and sending unit. It doesn't have any gauges. Just the WBO2 sensor which goes to the processor that turns it into a 0-5V signal. I bought the very simplified version for the purpose of wiring it up to the TGV.

 

Any suggestions on proper ENGINE-ON 12V (I think it's 12V...) to use in the engine bay to power the WBO2 sensor? I was reading that the WBO2 sensor can overheat if the sensor is powered on without exhaust gas flowing over it. Otherwise I was figuring I'd just do the switched 12V that goes when you put the key into the START position. Don't want to chance some shop having my car and keying it on for a while and never starting the engine. Then again, I could just disconnect the sensor when I bring it in for service... Either way, question still stands, any suggestion on engine bay power source?

[utc_pyro]

Yeah I don't think it's signal noise, but it's possible the ADC inside the ECU is not set up for the kind of resolution warranted for this type of sensor. I'll log and post up next time I have a chance. Perhaps I'll try hooking it up to my AQ-1 and see if that looks different.

 

Was just shooting from the hip - thinking perhaps the scatter was a cumulative effect of injector pulses within the sample interval, setting up a kind of phantom harmonic. With larger lines the effect would be less noticeable, but I haven't done any kind of sanity check on the premise.

 

Yeah low sampling rate may be causing some strange signal aliasing. Also not sure if the ECU even free-runs this input or only samples it once per cycle. The TGV's on the early cars were just on-off anyway, so it's not like it needed a lot of resolution.

 

I'll try logging at a higher rate (less parameters) and see how the scatter changes. Cobb used an AQ-1 in their testing and that showed a lot of high frequency variation.

 

I mapped the table but couldn't find a recommendation for the DW pumps. Do you recall where you found the 6V recommendation?

 

I first ran across that somewhere in the massive fuel pump controller thread, but they also list it on the spec page. The 45% came from (13.8V(runing)-0.5v(controller+wiring loss))*PWM_value = 6v. The 85% was just arbitrary chosen, but something between 73% and 78% would make more sense. I built this logging rig actually to verify those chosen values weren’t causing fuel pressure issues.

 

snow05gtRI, thanks for pointing me in the right direct. Looks like I have a lot of reading to do. For switched 12V for the wideband I'd tap the V+ off the rear O2 sensor. It's harness is the green plug near the turbo on top of the transmission. Then use the motor ground on the driver side TGV. If there is a seprate signal ground, I'd tie that into the ground on the TGV sensor plug.

Edited January 12, 2017 by utc_pyro
Typo correction

[Flinkly]

you guys and your simpler TGV wiring...

 

Does anyone know where to get the connector for the 07+ TGVs? they have both motor and sensor, so i believe 5 wires.