CAN diagnostics with a test light?

Got to attend a local class on CAN testing and diagnostics recently. Good class overall, teacher is a regular, well versed in the subject.

The problem of short to ground and short to power testing came up, and the teacher advocated using a regular incandescent test light for both. :blink: No disconnecting of modules, just going straight to both wires with your test light connected to B+ or B- as needed. Obviously, the light lighting in either situation represents a problem.

This caught my attention, since I've been taught previously that 12V on a CAN wire will fry the node in whichever modules are connected at the time. However, the teacher insisted this method is safe, and won't fry anything.

No CAN on my personal vehicle, and I'm too unsure of the test to justify trying it on others. Thoughts on this? Just to be clear, I'm all for using the good ol' scope-on-a-rope wherever possible! Just not quite ready to go probing networks without some more verification.

I am pretty sure each modue on the CAN bus is pretty well protected but I would think a DMM would be a more suitable tool.
Ivan MOTOYAM82 has a worthwhile video on chasing down a faulty CAN module with a DMM

Oh I agree, I've never had a problem finding CAN problems with a scope or DVOM. I think this whole thing came up when discussing 0V on one or both wires, as an alternative to resistance testing.

I guess I had the same reaction that some people had to using a test light on O2 sensor signal circuits. We were always told, "NOOOOOOO you'll cook the computer!" But the reality is different :lol:

It's funny. When I started 20+ years ago it felt like it was the beginning of the led test light era. My original tool 'kit' came with a incandescent test light but was quickly told that's useless and to get an led equivalent.
After watching Paul's videos amongst others I recently bought my first test light in way over a decade. How things come full circle. Haha

matt.white wrote: It's funny. When I started 20+ years ago it felt like it was the beginning of the led test light era. My original tool 'kit' came with a incandescent test light but was quickly told that's useless and to get an led equivalent.
After watching Paul's videos amongst others I recently bought my first test light in way over a decade. How things come full circle. Haha

Truth! My tech school preached how a incandescent test light was a big no-no on anything even remotely computer related. Nowadays, my expensive Snappy LED light sits on the hanger, while the incandescent lamp does all the work.

I've always wondered how this whole anti-test light scare thing got started... Was it during the introduction of fuel injection? :huh:

Andy.MacFadyen wrote: Ivan MOTOYAM82 has a worthwhile video on chasing down a faulty CAN module with a DMM

I've been meaning to watch his channel ever since the ETCG 2016 meet up video. Seems like a cool guy!

Was watching a newer South Main Auto video when I got the idea to start this thread. Not that a test light would have assisted his diagnosis, just got me thinking...

I have watched Ivans channel, he has a logical approach and a good grasp of the theory, I would recommend having a look.

This caught my attention, since I've been taught previously that 12V on a CAN wire will fry the node in whichever modules are connected at the time. However, the teacher insisted this method is safe, and won't fry anything.

I was told on a recent CAN training that they can take up to 17V :woohoo:

That's a tad more than I would have assumed :ohmy:
I need to get me some CAN training!

These days I think designers realise that automotive systems have to be pretty robust because wiring gets crah damaged and often water enters connectors and of course add the idiot factor. CAN is now used in industrial automation because it is pretty bullet proof.

A few years back I tried to explan to a shady tree mechanic who asked why just cutting the wires on the brake pad wear sensors on MINIs won't put the pad wear warning light out as a stop-gap(?) until you fit a new sensor. Instead I tried to explain simply joining the wires would because it is a pull down design, the guy totally refused to believe it wouldn't destroy the computer. BMW went for pull-down sensor design because it is more robust in the harsh environment it is exposed to and fails-safe if the wiring breaks. The pad wear sensor simply has a brass conductor embeded in it that wears through when the sensor rubs on the disc and breaks the circuit.

A few years back I tried to explan to a shady tree mechanic who asked why just cutting the wires on the brake pad wear sensors on MINIs won't put the pad wear warning light out as a stop-gap(?) until you fit a new sensor. Instead I tried to explain simply joining the wires would because it is a pull down design, the guy totally refused to believe it wouldn't destroy the computer.

:lol: :lol: Like that story. I guess he isn't the first one who doesn't get it. Mercedes also uses those sensors.
In my opinion there's a lack of decent education, training AND I've met a lot of guys in the field with BS stories, cause they watched Fast & Furious movies instead of taking some time to understand the basics and building further on that knowledge. It takes time and a lot of effort, sacrifice to become a good technician.

Dylan wrote:

This caught my attention, since I've been taught previously that 12V on a CAN wire will fry the node in whichever modules are connected at the time. However, the teacher insisted this method is safe, and won't fry anything.

I was told on a recent CAN training that they can take up to 17V :woohoo:

Holy crap, they're built stronger than I thought! All my previous training made the nodes out to be pretty frail. Just goes to prove you can't get enough training.

Holy crap, they're built stronger than I thought! All my previous training made the nodes out to be pretty frail. Just goes to prove you can't get enough training.

:lol: I wouldn't try it on a customer's vehicle :whistle:

Dylan wrote: :lol: I wouldn't try it on a customer's vehicle :whistle:

Noooooo way, I just wish my car was CAN so I could try some of this stuff out

Noooooo way, I just wish my car was CAN so I could try some of this stuff out

Just buy another car Tyler

By the way, do you have books or other material on FlexRay?

Dylan wrote: Just buy another car Tyler

By the way, do you have books or other material on FlexRay?

LOL, clearly this is the solution! I am currently shopping for another car, but haven't found the right one yet :whistle:

No FlexRay material Is this something you run into regularly?

Tyler wrote:

No FlexRay material Is this something you run into regularly?

I didn't exactly run into it. But I've heard GM uses FlexRay. The BMW X5 was the first in 2006 and used it for adaptive suspension. Faster than CAN & MOST, up to 10 Mbit/s. Can't find much about it, except for some basic crap