Miata Turbo Forum - Boost cars, acquire cats. - War and Peace . . . Me and my LC-1

As the title implies, I've struggeld with my LC-1. The two main issues I've had are:

1. The dreaded "Error-8"; and
2. Even when the LC-1 wasn't keeled over with Error-8, accuracy as read by my MSPNP was suspect.

So, hoping that others might gain from my struggles . . . :

Error 8:
Error 8 is pretty nasty. It essentially puts the LC-1 into a failsafe mode where it outputs 0VDC on its analog output wire. This, of course, means it is useless for tuning or closed-loop mixture control (probably the reason that you bought it, right?).

The LC-1 documentation describes Error 8 as a "timing error" resulting from either overheating or sensor damage. Taking Innovative at their word, I dutifully had a new sensor bung welded in just upstream of the cat and moved the sensor. Alas, that had no effect.

Moving on to cause number two, I dutifully replaced the sensor. Again, no effect. Me not happy. Abandoning Innovative as being inept due to downsizing, I moved on to the forums . . . hmmm . . . I'm not alone. It seems that Error 8 is a common error with firmware version 1.10. That firmware has shipped with every new LC-1 since 2007, and is still shipping today. The fix for many . . . revert to firmware version 1.00.

Recently, Innovative has acknowledged that there is an Error-8 problem with firmware version 1.10 and has been working on a solution. This has resulted in Beta firmware version 1.20. This is the firmware I've been running for the past week. I've had no further Error-8 issues and everything else seems to work well.

So, if you have Error-8 and firmware 1.10, the first thing you should do is a firmware change, either to 1.00 or Beta 1.20. I've attached all of the LC-1 firmwares to this post in a zipfile.

Accuracy:
Since installing my turbo, I've noticed that my MegaTune AFR readout generally does not agree with the AFR reading on my VEI gauge. As a rule, the MSPNP seems to read high (lean) with the result that I've been running richer than desired (which, at least, is safe).

Anyways, I was tired of the situation, and decided to get everything calibrated and matched. The calibration tool I used was a $3.50 linear taper 50KΩ potentiometer from radio shack.
Attachment 187703

I used test cables to connect my high-tech calibration tool to my MSPNP and VEI as follows:
Red Cable -- Common 12VDC source for MSPNP and VEI, connected to one of the outer tabs of the pot.
Black Cable -- Common signal ground for MSPNP and VEI, connected to the other outer tab of the pot and the negative cable of the multimeter.
Green Cable -- Common input line for the MSPNP and VEI LC-1 analog signal, connected to the middle tab of the pot and the positive cable of the multimeter. Note that the LC-1 needs to be disconnected for this calibration.

Here's a picture of the setup:
Attachment 187704

With this setup, I was able to input a reliable DC voltage into the signal input of the MSPNP and VEI. I could read the DC voltage on my multimeter. After setting up my MSPNP for LM-1 emulation (nominally 0VDC = 10:1 AFR and 5VDC = 20:1 AFR), I recorded the following linear relationships:

0.0VDC MSPNP=10.0:1 VEI=10.0:1
5.0VDC MSPNP=20.8:1 VEI=20.4:1

So, both devices were off a little with pretty normal component variances (in other words, yours WILL be different). Easily fixed though. For the MSPNP, I'm not aware of an adjustment that can be made . . . however both the VEI and LC-1 can be adjusted. I adjusted the VEI endpoints so that I got an exact match on the VEI display and Megatune -- confirmed by swinging through the range of input voltages with my high-tech tool. I then setup my LC-1 to output 0VDC at 10.0:1AFR and 5VDC at 20.8:1AFR. Viola! Calibrated.

Now to do some autotuning.