kewilso3

iTrader: (4)

i know i know another one... i just got my lc1 wired up, all grounds from lc1 and gauge go to ecu ground under TB, shared bolt, different nut. (didnt disturb factory grounds)

I have been using TS a few days on narrowband, running good. I snipped the narrowband pink wire to the MS and wired in the extra analog there.

the gauge works, but i'm not sure which settings to choose is ts to get that gauge to work. I cant hook up the lc-1 to my computer because i dont have a serial port on my laptop, so i need to know the default? I chose all combinations in TS and none resulted in gauges that agreed with each other.

The gauge says about 14 at idle, TS says 10-12 in the closest configuration I can muster... which one is correct? Can I drive like this? I need this figured out obviously so I can tune.

I feel like i've seen this problem before but there's so many lc1 problems i cant find it again.

kewilso3

iTrader: (4)

according to innovate, analog 2 is 0V=7.35afr, 5V=22.39afr

TS project properties / settings give me these choices:
0-5V=.5-1.5 lambda
1-2V=10-20:1 AFR
PLX 0-5v=10-20:1 AFR

When calibrating afr, i have these choices:
LC1 default
0-5 10:1-20:1
1.0-2.0


Do I need to do a custom Wideband calibration and set it up as 0V=7.35afr, 5V=22.39afr?

kewilso3

iTrader: (4)

well damn i tried all this and nothing worked... TS still reads one or two points lower than the gauge. I'm using analog 1 to the MS as per instructions.

I think the gauge is right (around 14 at idle), but the signal to MS/TS is what really matters, FML.

Cococarbine3

iTrader: (1)

You really need to get a computer to hook up to your LC-1, so you can change the two analogue output settings on the controller. I believe it is set up to simulate a narrowband signal from factory, and you need to connect to a computer using a the serial cable and the LogWorks software to change the values to a wideband signal (0-5v .5-1.5lambda), for whichever analogue output is connected to your ECU. Then in TunerStudio you want the 0-5v .5-1.5lambda setting.

For the gauge analogue output, I believe the values are 1v = 10.1afr and 2v = 20.1afr.

You should use the LogWorks software to find out your TRUE afr by recording while your car is on, and then compare to Tunerstudio and your gauge. For mine, my ecu is spot on, but the gauge is always ~.8 value under. I suspect it may be a grounding issue

Some info here: https://www.miataturbo.net/showthrea...t=44332&page=2

kewilso3

iTrader: (4)

ok i'm gonna hook it up to the computer, totally forgot about the adapter i had from megasquirt.

Shouldnt have to change analog 2 (brown, to gauge), it's set up to read default signal.

kewilso3

iTrader: (4)

oh and thanks a bunch, one day i wont be a newb

popopopop

Is your gauge digital or analog? I concluded that my analog gauge just doesn't keep up with how fast the lc-1 updates, especially at idle. At boost though the gauge holds pretty steady at the right AFR, so I know it's working.

kewilso3

iTrader: (4)

analog... but both remain very steady, just 2 points apart

hustler

iTrader: (6)

put delay in the lc-1 putputs. I have 1/16 on the MS output and like 1/6 on the gauge so I can read it.

hustler

iTrader: (6)

You need to:
grow a brain
open the lm-programmer
set the outputs to report the correct range
put some delay in them
do a free air calibration
kill yourself


How are you connecting to MS? I use the same USB/serial converter. Don't expect it to work right any time soon if you can't connect to either.

kewilso3

iTrader: (4)

got it worked out, mostly. programmed lc1 output to analog 1 as 0-5v = .5-1.5 lambda

still off, by less than .5, but i think this is because in calibrate afr there's not an exact match?

therieldeal

iTrader: (2)

so LC1 delay = smoothing of sorts? i will have to try this, my wideband signal is always pretty wonky

hustler

iTrader: (6)

Yes, it takes the sample period and averages the numbers. When I had the stock 1.6 in my car with the sensor in the stock location I could see the individual cylinder puleses I think. 12.9-12.4-12.9-12.4-12.9-12.4-12.9-12.4-12.9-12.4.

gospeed81

iTrader: (51)

OMG I bet you can see flakes in your oil too.


OP. I have this happen every 4-6months. Only fix is what hustler listed.

Reset your outputs to correct range (LM programmer)

Free air calibration.

hustler

iTrader: (6)

When I use a magnet in a plastic bag...I still can't see flakes, lol.

kewilso3

iTrader: (4)

Thanks for the thread and help, I feel glad I wasnt in that situation.

1)It bothers me that the gauge and TS dont use the same settings... Could I just program the gauge to also use 0-5v .5-1.5 lambda?

2)I would log, but i cant get MLV to work lol, and I only have one serial-usb so i cant run both at once... Why would logWorks be most accurate?

3)as of right now, car is running fine, but the gauge is still .5 richer than TS... messing with my head for tuning...Should I run logworks and compare that to my gauge real-time? I'm assuming the signal to logworks will be most accurate for some reason? Does it automatically know I'm running 0-5v .5-1.5 lambda?

Cococarbine3

iTrader: (1)

1) Lol it's fine. Different products are programmed differently. I suppose you could make your own wideband gauge and program the code to recognize the same output megasquirt uses, but that would be very time-consuming and unnecessary since you can use LMprogrammer to adjust the LC-1 output values to the values that the device recognizes. I see you mentioned you had an analog gauge. You want to set whichever LC-1 analog output you have to your gauge {I would assume 1, since 2 is probably the one you're using for MegaSquirt since its default values are already set to 0-5v .5-1.5lambda} to:

1v=.554 lambda and .0v=1.224 lambda, according to http://www.innovatemotorsports.com/s...ual/Analog.pdf

It's possible your gauge reads something different; if so, use what is specified in the manual it comes with. I am using an el cheapo ebay digital voltmeter to measure air-fuel ratios, which is a little different at 1v=10afr and 2v=20afr.


2) Logworks is getting data DIRECTLY from the LC-1 controller via the serial cable. It eliminates the possibility of error such as poor grounding or wrong output values, causing trust issues with the mismatching values you are having in #3. My prediction is that your Tunerstudio value is the correct one, while your remote analog gauge is a tad rich.


3) I am having the same problem. I suspect it's a grounding issue for me since my gauge is not grounded to where everything else is grounded, but I'm not sure. I know that for my setup Logworks showed the same exact afrs as TS, and the gauge showed about .8 richer. Right now though I have a whole different set of engine problems that has me preoccupied at the moment.

Cococarbine3

iTrader: (1)

Oh, I mixed lambdas and afrs all throughout there, sorry about that... They're the same-ish. Just kinda like using Celsius and Fahrenheit in the same post.

kewilso3

iTrader: (4)

The gauge is accurate according to logworks, not TS... I tried calibrating afr in TS to "lc1 default", 0-5v 10:-20:1, and did a custom linear option to 0-5v .5-1.5 lambda... lc1 default is still closest, about .3-.5 leaner in ts compared to gauge. I thought maybe the voltage was dropping on the way to the MS but that would make it seem richer, not leaner.

gospeed81

iTrader: (51)

Is your gauge grounded in the same place as the controller and ecu?