Can bus sniffer

[Midnightmoon]

I've been reading can bus issues need an oscilloscope to diagnose properly yet in service manuals I've yet to come across the use of them in diagnostics. I've used diagnostic manufacturer software to save can info and send it to engineers to analyze.

I picked up two oscilloscopes and understand what I'm seeing but without data of a healthy system im not sure of the usefulness.

Can bus sniffers attached to your computer seem to be another tool but has anyone used this for diagnostics.

Anyone build a sniffer or try out sniffers

 

[Birken Vogt]

There is a video on Youtube of a guy using a scope to nail down a canbus problem on a Chrysler car where the electric power steering pump was the culprit I think.

It gives you some idea what is involved and how it is supposed to behave.

I have used my old fashioned CRT scope to look at canbus when I thought I had a problem and it was real hard to see what was supposed to be happening. In the end it was just one of the modules that had glitched when the battery got real low and disconnecting the battery and reconnecting was all that was needed in the end.

 

[BigWrench55]

The can bus sniffer will identify everything that is communicating on the network. Some come with software that can name the component most just give a address. With the one that only gives the component address, you will need to unplug each component to see what it is. As far as the oscilloscope goes I haven't seen any technical info about what your wave lengths should look like. Cat has a excellent video on YouTube explaining how can bus works. It's basically like Morris code with voltage as your dots and dashes. I normally don't need to go to such extremes to identify can bus issues. But I also have access to software. If you are savvy enough the codes will tell the story of what is happening. In my experience it's usually a failed sensor or resistor. Rarely is it a harness issue. I get called out when one of the other techs says it needs a ecm replacement or harness. In my 20 plus years I only replaced 2 harnesses and 3 ecm's. One harness was destroyed by fire and the other warranty department paid for. As for the ecm's one I messed up and the other 2 were actual failures, also determined by a failed 120ohm internal resistor.

 

[Midnightmoon]

https://www.csselectronics.com/scre...g-can-bus-messages-with-wireshark/language/en
I'm looking at these but will probably build my own with a raspberry pi

Never had to use oscilloscope either. diagnostic software gets you to where you need to be.
in 20 years only a few ecm needed replacement had one with a failed barometric pressure sensor. Other one had a welding failure. Guys where I work a quick to reinstall ecm software before they even start diagnostics because they miss understand the diagnostic procedures. They always state is the software up to date. Even if it is they reprogram it. Any experience with software actually being the issue?
I've had a few harness with worn wires from crossing sharp metal.
Yet techs like your saying screem bad ecm bad harness before actually taking a good look. I had bbn one machine with a add on gps harness that had can wires backwards. Other tech replaced harness ecm 2 controllers before I got it. It took a while but found gps would occasionally try and pull info from ecm when it did can high to can low and can low to can high caused all kinds of crazy codes

 

[DMiller]

What we found in the 90s on the Caddy GM digital(Fore runner to Can Bus) systems is had to use a CRT Scope, the designed in delay in the Electronic units would miss the minor fluctuations.

We used a OLD Allentronic Ignition scope for some deep diagnostics.

 

[BigWrench55]

I have seen software cause some of the issues that I troubleshot. But that's after going through all of the diagnostic procedures and not finding anything wrong. Volvo software updates will tell you what it's patching up. It's tedious and largely boring but I will scan through and see if it will fix my problem. And sometimes it will mention the problem that I am experiencing. As much as the software engineers think they are god's they screw up too.

 

[Coaldust]

I’ve connected real scopes and my Fluke Digital storage scopes (DSO) to the J1939 to show my students how the information packets flow, for fun and entertainment.

I don’t know how it could be interpreted or meaningful for J1939 diagnostics. Maybe? I’ll have to do more research. Interesting.

There are other cool things your new scopes will be useful for, like watching sensor outputs.

 

[Midnightmoon]

this one has a nice software screen that graphs. The issue is you need healthy system info to compare with.

https://www.saleae.com/home2/

Without a database you wouldnt know if something is working right or wrong
I have saved system data on new machines so I have something I can compare a problem machine to
like idle boost and full throttle boost ect. Some of the manufactures I work on dont provide this base info
like kubota. Kubota has turbo issues yet no boost specs under certain rpm ranges. So I made my own.

 

[Coaldust]

That is really cool, midnightmoon.

How do you save and organize the known good wave forms? To make your data base, with a set up like that?

 

[JD955SC]

At work I have been trying to get into an advanced electronics class where they teach you to use a scope to diagnose issues but it’s full every year. I use a multimeter to get by right now.

 

[Midnightmoon]

I think I confused you I make files using data logging on current manufacture diagnostic tools. I save these from new machines so I have a baseline to look back on when I have a similar machine with a problem.

Ive saved a couple wave forms with a digital scope I recently got playing around with it but Im not sure how useful it will be thats a neat little bluetooth tool I got mainly because it logs any info a scope can or a multimeter. for instance I just used it to catch a voltage drop that was happening everytime I walked away. Im not sold on it yet it but worth the $90. Its a Pokit meter can be found on amazon. Snap on sells one too for $500

I want to buy a can sniffer and do the same thing. this one comes with a j1939 spn database for addition cost. Im curious how good the database is.
https://www.csselectronics.com/scre...,CAN hardware to try and control your vehicle.

But this one looks to have a better software user interface

https://www.saleae.com/home2/

 

[Midnightmoon]

Just ordered this one
https://www.saleae.com/home2/
If you look through the site you can get it to your door for 209 with a discount code for enthusiasts, students, startups. Mechanic passes as enthusiasts

https://www.saleae.com/home2/

 

[Cmark]

In my experience, 95% of J1939 network problems can be pinned down by checking the resistance across the pair using a DMM. With the ECMs powered down, you should see the sum of the two 120 ohm terminating resistors which is roughly 60 ohm. If you see anything other than that, you can start disconnecting ECMs until something changes.

 

[excavator]

At work I have been trying to get into an advanced electronics class where they teach you to use a scope to diagnose issues but it’s full every year. I use a multimeter to get by right now.[/QUO

I wish my local tech school would actually offer anything close to electronics classes.

 

[excavator]

An interesting thread. I haven't had to deal with CAN issues that much but have been learning a bit along the way. There's still a Deere 310SJ TMC backhoe sitting in the yard that hit an underground power line and killed the boom and backhoe hydraulics. Anything I can learn on these systems is appreciated. Thanks.

 

[Midnightmoon]

It's not my intention to use this initially to diagnose can bus issues but to start making databases of healthy systems and smart sensors. Like the def quality sensor for instance. Every 20 seconds it sends a can signal to the ecu. I'll always grab my fluke first but down the road many years I hope to find new ways to help in diagnostics via can bus signals.

 

[Midnightmoon]

On the backhoe find the melted wires and you will find the problem. Or maybe burnt marks on the hyd valve block where it melted the housing...........

 

[excavator]

On the backhoe find the melted wires and you will find the problem. Or maybe burnt marks on the hyd valve block where it melted the housing...........

It would be nice if I could find a melted harness, this is a TMC (total machine control) machine. No hydraulic pilot system, all electric with 8 or 10 different controllers. Throws approx. 10 codes, took the first one that Service Advisor said to replace, did all the harness tests ect. Talked to the local Deere service tech, after doing his research he agreed that the controller needed replacing. $2300.00 for the controller, $500.00 for him to program it and no change. At what point do you say the machine isn't worth it? I'm the 3rd person to work on it and I probably have 50 hours in it.

 

[BigWrench55]

At what point do you say the machine isn't worth it? I'm the 3rd person to work on it and I probably have 50 hours in it.

I will fix it for the low low price of everything you spent so far and whatever else it will need.
Post all of the codes you have with serial number on another thread and we will point you in the right direction.

 

[mg2361]

Post all of the codes you have with serial number on another thread and we will point you in the right direction.

We will certainly try anyhow!