My old Innovate LC-1 is doing my head in. I last connected via serial about 13 years ago and despite way too many hours of effort so far, I have not been able to connect again. I don't have the original computer nor cables, so I can't really rule out causes, but I have been slowly making progress. Here's bit of a summary:

I bought a generic USB to 2.5mm cable (with FTDI chip). I opted not to buy an Innovate DB9 to 2.5mm cable as they're hard to get, cost ~$60 and how hard is it to copy a RS232 cable?! Turns out, quite (more on that further on). Every single reference on the internet said that the tip of the 2.5 aux plug is Tx, middle is Rx and base is ground, so I bought a cable to match that.
I installed the lasts LM Programmer and LogWorks3, plugged my terminator into the "IN" cable, my USB to 2.5 cable into my laptop and the "OUT" cable and had no luck with comms. I proved that the USB to 2.5 serial cable actually worked by doing a loopback test and it was fine.
I suspected that my LC-1 cables and terminator may have gone open-circuit and it turns out that my terminator has done so or is temperamental. I cut the IN cable and tied two of the three wires together, based on a thread I found on a Fiat Forum where they cut open an LC-1. I still had no comms, so I cut my own LC-1 open.
Turns out the poster was wrong, despite having cut their LC-1 open. For everyone's reference:
IN cable: Red = Tx, Black = Rx, Green = not fully tested, but NOT ground. Possibly floating, which wouldn't be surprising.
OUT cable: Red = Tx, Black = Rx, Green = Ground. These are expected results - these are standard colours for RS232

Anyway, this meant that I needed to tie the red and black IN wires together to guarantee that I had working termination on the cable. Still had no luck after this.
I then opened up my USB to 2.5 serial cable and did continuity tests for the three wires. Turns out, everyone on the internet is wrong: the tip of the 2.5mm Aux pin should be Rx, not Tx (and therefore the middle section should be Tx). I resoldered the USB wires to correct this and still had no luck!
I tested the new setup trying to connect to the LC-1 via PuTTY and it worked! At 19200 baud, or course. I was receiving characters: I can't prove if it's exactly what I should be receiving, but there is a clear difference in characters received during the sensor warm-up phase and during normal operation.
Still, LogWorks nor LM Programmer would not connect. I decided to uninstall the Innovate software and install the version available in 2012 (thanks to Internet Archive). I then got LogWorks to connect for maybe one second before it disconnected. I should have recorded the exact message I received when it disconnected but it was to the effect of MTS data transmission stopped unexpectedly. Check your wiring.

This brings me up to today - I can recheck my wiring once again and I can try installing other old versions of the Innovate software, but are there any other things to try? Are there any secret serial port settings I should configure in Device Manger other than setting to 19200 baud? I can't understand why PuTTY seems to be happy but LogWorks and LM Programmer are not. It should have been a 15 minute task to connect, configure and be set for the next few years.

I hope you all enjoyed my verbal diarrhoea/story.

 

Update: final issue was my FTDI USB to 2.5mm aux serial cable, despite that chip being known to work for other people. I think mine was faulty. I bought a random $4 USB to DB9 adapter and jury-rigged a new set up which then communicated to the LC-1 without issue. I did install and use the 2014 release of Logworks 3, but suspect that any version would have worked.
I'm now working on getting the analogue WBO2 signal I'm logging through the ECU to more closely align with the serial comms data. I always knew that my analogue signal jumped around a lot so I figured that there was electrical noise on the power being fed to the LC-1. I came across an install guide that suggested installing a 4700uF 25 or 35V capacitor and have ordered one now. After that, I might reduce the range on the signal from the default 7.35 to 22.39 down to maybe as far as 10 to 17.35, which would double the resolution and allow some very fine tuning.

FWIW here's some pics of the LC-1's guts:

 

Conventional ftdi chips (as ft232) is a UART, all Innovate products use RS232 proto, with ft232 u need also uart-RS232 converter

 

Thanks, that's good to know. I specifically picked the FTDI-based adapter as that was what was recommended by someone on a forum (unsure if it was EvoM, but I don't think so). Once again, the info on the internet was wrong! Glad we've got all the right info in this thread now