LC-1 - cannot connect

[derekf]

Okay, I installed my LC-1 per the instructions at http://www.innovatemotorsports.com/resources/efilive_lc1_tutorial.pdf :

Red wire going to a free Acc port in the fuse box
Yellow goes to pin C on a Flashscan v1
Blue, Green, and White go to pin D
Black goes to the LED and pushbutton (in parallel) and then through them to ground.
Brown is not connected.

Wideband is on a separate bung, it will not be emulating an NB sensor.

It doesn't work. The LED is constant, pressing the button to start calibration does not do anything other than turn off the LED. Trying to connect via LM Programmer gives a "please connect the device to a serial port" error message. My laptop is old enough to have a real serial port so I am not going through an adapter.

I saw a thread that looked similar at http://forum.efilive.com/showthread.php?2852-lc-1-help-plz. but frankly it confused me. Are the instructions as given from Innovate incomplete -- should there be a connection between D and ground, or should brown/blue/white be connected somewhere other than D.. or is there a power switch somewhere on the LC-1 that I'm not seeing or something along those lines?

[joecar]

Blue has to go to vehicle's battery ground.

Connect the ground from the bottom of the LED/pushbutton to the same ground as blue.

Try with white grounded and floating.

[joecar]

Sanity check:

red -> power (switched, fused, battery)
blue -> ground (battery)

black-> LED anode, pushbutton
white -> LED cathode, pushbutton

brown -> V1 pin C (signal)
green or white or sheath -> V1 pin D (signal ground)

yellow is programmed for NBO2 signal emulation; yellow can go to V1 pin C, but you have to reprogram the LC-1.

May or may not have to also connect green and/or white to battery ground (mine works regardless).

[joecar]

To program the LC-1 using the LM Programmer software, you have to connect this:
LC-1 SERIAL IN <- terminator plug (looks like a blanked 2.5mm stereo plug),
LC-1 SERIAL OUT -> PC

[derekf]

Thanks, Joe. The ground made some difference; the programmer now can find the WB.

However, neither EFILive nor LogWorks reads anything from it. EFILive just reports 0.0 and logworks reports "No real time data detected".

Hopefully I'm just missing something small.

[derekf]

After a reboot, LogWorks can successfully talk to the sensor. However, I still get no data through either the yellow or brown wires. I've tried with white both to pin D and floating, with no difference.

I have both green and blue connected to pin D, with an additional wire from pin D to the vehicle ground. That does seem to be sufficient for LogWorks, anyway -- is that the correct way?

[joecar]

Try removing blue from pin D (but keep blue going to vehicle/battery ground).

The signal on the brown or yellow wires is wrt green or sheath.

[derekf]

Thanks, Joe. Does pin D also get connected to vehicle ground, or just to the green wire?

[joecar]

Pin D's primary function is to connect to signal ground... so try just the green wire.

[derekf]

Sigh... still no joy (I had such high hopes).

Now LogWorks reports an AFR that bounces around all over the place -- even with power on/engine off -- so now I'm thoroughly confused. Reconnecting blue+ground and/or white to pin D doesn't get it stable again, and of course EFILive still doesn't get anything useful.

[derekf]

Quick response from the Innovate support folks -- redoing the free air cal fixed the bouncy, but not the lack of analog-out joy.

[joecar]

Check the connections:

red --> switched fused power
blue --> battery ground

brown --> FlashScan pin C
green --> FlashScan pid D
yellow --> FlashScan pin E

black --> pushbutton + LED anode
white --> pushbutton + LED cathode

[joecar]

If still fails, then also add this connection (in addition to the above):
green --> white

If still fails, then also add this (in addition to all of above):
blue --> green

[joecar]

Have you done this:

heater calibration (turn on with sensor disconnected, turn off, reconnect sensor, turn on and wait).

[derekf]

Heater calibration - I think I did that by accident when I was first setting it up, but not intentionally or since. Think that might be the problem, given that LW works?

Please sanity-check me on the wiring -- pushbutton/LED cathode is connected directly to ground, yes? White->cathode would be equivalent to white -> chassis ground?

Green->white and blue-> green would be the same net result as green + white + blue + ground -> pin D? I've done green+white+blue+ground to D, and I've done green to D and blue to ground, and green to D and blue+white to ground.

[joecar]

Yes, LED cathode goes directly to ground (if LED was wired backwards, it would simply not light at all, and no harm would be done).

Yes, green->white->blue->ground->pin D, one big ground...

Try this: using LM Programmer software, program yellow to 2.5V and brown to 3.5V, wait a few moments, then disconnect LM Programmer, cycle power (off-on) to LC-1... then with DMM/voltmeter check the voltage at the brown and yellow wires wrt green and then wrt white.

If it fails this test, then the LC-1 may be defective...

Did you by any chance short the yellow or brown wires to any other wire or to ground (this will blow out the ADC output driver)...?

[derekf]

The result was not what I was expecting - the output voltages were pretty close to what was programmed, using the tiny screws in the 3-pin plug as the test points.

Are there any diagnostics to be done on the flashscan itself? I've seen values for WO2AFR1 and WO2AFR2 ranging from 0.00 to 215ish, with 10.0 being the most common.

[joecar]

Do yo have FlashScan V1 or V2...?

WO2AFR1 and WO2AFR2 are on V2 only (V1 does not have wideband serial data ability), those are the wideband serial data pids.

With V1:
connect the scantool to vehicle, with LC-1 programmed to 2.5V/3.5V, record a log (say 20 seconds), the pids you want are EXT.AD1 and EXT.AD2, these should show the programmed voltages...

post this log if you can.

[derekf]

That's awesome.... (laughing at myself). I'm on V1.

Will try with the correct PIDs. Stand by.

[joecar]

lol I know the feeling :)

the System column should probably mention V1 or V2 for EXT pids.

[joecar]

With LC-1 programmed to default (brown channel), you need these pids:
AFR_LC11 and/or 2
BEN_LC11 and/or 2

[derekf]

Yeah - when I use the right PIDs, I get values, how wierd!

I had already reverted the LC-1 voltage settings back to the defaults, so the log should actually be reflecting reality and not the 2.5/3.5 v values.

Pin C is the yellow and pin E is the brown (backwards, I know; assume that means that for real logging I'd need to use EXT.AD2 or swap them)

8667

[joecar]

That's correct, use EXT.AD2 or swap.

make sure LC-1 yellow channel is programmed to match the AFR_LC11 pid and that LC-1 is programmed with the same stoich AFR as your B3601.

I'll look at your file later on my other PC.

[derekf]

Okay - altered B3601 to reflect 14.12 for the E10 fuel that's all we can get here.

Reset the LC-1 to 2.5v on Analog 1 and 3.5 on Analog 2.

First log is with power on but engine off -- EFILive reads pretty much exactly what it should for the voltages, without much bouncing around.

Second log, I reset the LC-1 voltages to the factory defaults and captured with the car idling; Analog 2 is LC12 (so I assume that that's what I need to pay attention to).

Anything jump out as wrong, or am I headed in the right direction?
8676
8675

[joecar]

Sanity check, is this what you have:

LC-1 Analog 2 = brown -> V1 pin C = EXT.AD2


EXT.AD2 -> pay attention to: CALC.AFR_LC12, CALC.BEN_LC12

(you had CALC.BEN_LC11 in your log... but that's ok, you can add calc pids to the log file after the fact, and then save the log file).


Sanity checks:
- LC-1 Analog 2 (brown) is programmed to default V:AFR points...?
- LC-1 is programmed for stoichiometric AFR 14.12...?


If the AFR_LC12 waveform has too many squiggles, you can program the LC-1 out do 1/3 or 1/6 smoothing.


Looks like it's working... you can further sanity check by allowing Closed Loop which should cause AFR_LC12 to tightly oscillate around 14.12.


That spike in AFR_LC12 is a single misfire... the fuel/air was not burnt, so the oxygen is not used, so the wideband registers lean (wideband looks at oxygen content in exhaust gas... so a misfire is different than a rich burn).

[joecar]

Oh, something is wrong with GM.DYNAIRTMP_DMA it shows 326°F which is incorrect... it should be close to IAT with a little influence from ECT;

it might that DYNAIRTMP_DMA is not defined correctly for 1998 OS's.

[joecar]

Derek,

Is this a 1998 Corvette or Camaro/Firebird...?

[derekf]

Sanity: confirmed. Brown is connected to pin C. The settings are the factory defaults for Analog Out 2, with the exception that stoich is set to 14.1.

This is a 98 Trans Am. DYNAIRTMP_DMA doesn't appear to resolve as a supported PID on this OS, even if I do "validate PIDs" before I start logging.

[joecar]

Yes, that pid is not supported on 1998...

You would have to create a calc pid that using lookup() on a copy of your B4901 table and blend IAT/ECT from this, as you have done in your other thread.