SRS Airbag system

[onyxnz]

Hi all!
I have a Toyota Granvia '96 KCH10W (basically a HiAce) which is throwing an error with the SRS airbag light staying on permanently. These are Jap imports which are next to impossible to find real service information on, and I don't have the thousands of dollars I have been told to expect to fix this issue at an auto sparky. So, as a qualified computer techie, I should be able to handle it myself, right?
Using the onboard diagnostics port in the engine bay, the ECU (and a chart of error codes I found on the net) tells me that there is a "short to positive" at the spiral switch (Toyota's name for clock spring).
I did the following:
Remove negative terminal from battery, wait 15 minutes.
Disassemble the steering wheel, being very careful with the airbag, as per this video :

Re-attach negative terminal to battery.
Measure the voltages pre clock spring and post clock spring -> srs airbag plug. Both identical; one wire at 1.64V +ve, the other at 13.45V +ve.
No short found. Disassembled and checked the clock spring anyway, no fault found. Reassembled and tested again as per above results.
Made a test shunt to emulate the airbag, as per How To Make: DIY Airbag Tester for airbag diagnosis. - https://www.themechanicdoctor.com/how-to-make-diy-airbag-tester. No change in results. (therefore no fault in airbag?)
So my questions are:

• What voltages should be there, entering and exiting the clock spring, and are they the right polarity?
• Is there another way to test the system?
• Is the ECU even telling the truth? Can it be reset without having to re-tune everything in the car that relies upon it?

Cheers!

 

[James]

I would not recommend doing any more testing of voltages on the airbag circuits.
it is possible to set off the airbags and this could lead to serious injury.

this might not be the response you were hoping for but.

DO NOT MESS WITH SAFETY CRITICAL SYSTEMS
unless you have the technical knowledge and experience required.

 

[onyxnz]

I would not recommend doing any more testing of voltages on the airbag circuits.
it is possible to set off the airbags and this could lead to serious injury.

this might not be the response you were hoping for but.

DO NOT MESS WITH SAFETY CRITICAL SYSTEMS
unless you have the technical knowledge and experience required.

Thanks for being candid, James. And yes, it's not what I want as a response; the airbag is disconnected, therefore it cannot go off by itself. The cost to repair such a problem at a local auto electrician is more than the "book value" of the vehicle, which is perfectly good otherwise, so I cannot justify writing it off, nor can I afford a replacement to cart the kids around. Failing the ability to diagnose this ECU fault I may just modify the vehicle by removing all vestiges of airbag system, and get the necessary certification to say the vehicle is modified, so it can pass the warrant of fitness.

 

[snowhead]

The problem with diagnostics is that they are only a guide.
Regardless of any voltage readings, personally I would suspect the Clock spring as it's just a long piece of wire that gets wound up and unwound as the steering wheel turns.
It may have a break that's making / shorting in certain positions and the diagnostics have detected that.

Unless there is a fault detection line at the spiral switch then the diagnostics are only monitoring the line to / from the switch
"short to positive" at the spiral switch would suggest there is an issue anywhere on the spiral switch line from the control module or maybe even the airbag control module is at fault.

I'm certain it will need resetting when fixed.
Most makes have aftermarket software to access at a greater level than a standard code reader.
Ford have Forscan, Audi / VW have VagOBD
There must be something for Toyota

 

[onyxnz]

So I just found this, which shows that of the two wires connecting to the spiral cable, they should be opposite polarity. They aren't; as measured without the spiral cable in place, and same with it in place. Both readings being +ve.
So bad sensor assembly?