CAN Bus: A Beginners Guide Part 1

This is the first of a two part series on the CAN Bus. In this video we will take a look at the physical aspects of the CAN Bus and the protocol itself.
Part 2 Here.
• CAN Bus: A Beginners G...

[@BitwiseMobile]

One of most important features of the CAN bus is it's noise immunity. It can operate in very noisy environments where other protocols fail. That's it's primary strength. Cars are very noisy electrically - especially the ground.

[@1963TOMB]

It's a differential pair, similar to RS485 which in turn is RS422 on two wires instead of four, i.e. half duplex instead of full duplex

[@tensevo]

13:24 this is mind blowing, every time it is explained, sheer genius

[@aarongoff3914]

I am very impressed with the way this guy breaks down all of the information in this video. He should teach for a living 👍

[@manitoublack]

Thanks for these short videos. They're a great heads up on how these systems work.

[@brendanmccabe8769]

Really good video, lots of information well explained, delivered in very good style and pace. I’m looking forward to watching more of these.

[@k4kfh]

This is one of the best overviews of CAN I've ever seen. Thank you!

[@netgrok]

Brilliant. Thank you. Just the part about why differential pairs are immune to noise is worth the price of admission.

[@shitmyhellcatsays]

One of the best videos on explaining this topic. Very well done.

[@user-pm7ti4ho4m]

while i just randomly choose your video and going though it for just 15 min, Man you are just amazing , what an explanination. 👏👏👏

[@bandaralfoaim8672]

Thank you George. The vid was very easy to understand for beginners such as me. I appreciate it.

[@robbieraychannel]

Best video I have seen on this subject, Thank you for taking the time to post it. Rob.....

[@user-py2ni5od1c]

You just won my subscribe and my heart with your detailed explanation and the oscilloscope example!

[@shivamtripathi7910]

One of the best video on CAN.👍

[@pete540Z]

Thank you, very good lesson for someone like me that knew nothing about CAN or any other types of bus before watching.

[@Paddington2000]

I don't use CAN bus very often at at all but this was a wonderful and very interesting video from a general electronics point of view. Thanks.

[@alexbolin3159]

This video is awesome thank you so much for your time making it. I'm a industrial highspeed door installer and the manufacturer we work with has just added CAN protocols to there doors for all of the safety, accessories and communication systems for the doors and are having major issues with it and this helps me understand the system better and to help the manufacturers R&D come up with some solutions

[@nigelshepherd2970]

That was absolutely brilliant, well explained... I'm a hgv fitter that has been out of the trade for 20yrs...I feel like I have a understanding of can bus...thank you...👌👍

[@Umski]

Thanks, takes me back to the early 2000s when me lab partner and me were jobbed with linking two Atmel 8051 MCUs on a CAN BUS - can't remember what it actually did but the module was on multi-processor embedded systems in C++ - now trying to get my head around what the hell is happening on all the modules on a modern car 😁

[@danwaterloo3549]

Thanks! great video... and very clear. I like the 'chalkboard' approach. You do it well.

[@RufusVidS]

I first fell in love with CAN bus when I found out a message collision doesn't stop the high priority message from getting through without interference! The lower priority just backs off and the high priority message goes out unimpeded. (Unlike ethernet, where EVERY transmitter in a collision has to back off and wait a random time before retry).

[@mitchs2148]

Thank you so much for such a well done and informative video. Thank you.

[@frasersimpson8264]

Yeah awesome work George..........really informative.

[@barryconaghan7248]

Thank you for explaining the can bus 🚌

[@alibro7512]

Brilliant video, thanks for sharing. In my (limited) experience CAN gets very flaky when we don't follow the rules for termination and stub length. Often it will work with long stubs and extra termination resistors but be unreliable.

[@parrotraiser6541]

Thank you. I've been watching CANBUS for a long time with great frustration at being unable to find proper documentation. Ad hoc rules are all very well, but it's hard to do serious thinking about a system you cannot grok. Just considering wheter it's appropriate to a problem requires understanding.

[@chrisedwards2539]

Really good video. I've been wondering about canbus since I first heard the word but am coming from IT. I'm not that hot on electronics but ive meddled with microprocessors . So i liked the level this video was pitched at. Many thanks

[@gearhead719]

Brilliant explanation! Now I understand.

[@MichaelKingsfordGray]

Well explained. Thank you.

[@PatHaskell]

Beautiful description!

[@Lexusturbo]

Really enjoyed that. Everyday is a school day!

[@ui4lh]

This video is gold, I needed this so badly. Thanks!!!!!!

[@maystokes1562]

Explained brilliantly.best video out there

[@ELEISONCARABALLO]

Awesome video. Congratulaciones

[@Bianchi77]

Cool video shot, keep it up, thank you for sharing :)

[@jamesgeddes797]

Fantastic explanation for a home gamer. Thank you!

[@anhhuy4466]

Thank you for your sharing, this video is very helpful.

[@SidneyMoropa-iy8cg]

Very helpful thanks for sharing ❤

[@praveenprithiviraj6502]

Very good video with very informative. thank you...!

[@klebermolina2850]

Loved the video, had almost 0 experience at the area, and understoo everything clearly !

[@buniyaminadewusi8593]

This is awesome pls do more of this

[@TiTiNMaRiKa]

Masteful explaination, thank you so much for this video

[@sirousmohseni4]

Thanks for the details.

[@fredsalter1915]

subbed! Great vid. I'm very interested in CAN regarding automobiles. Thanks for posting this excellent vid!

[@AhmedBahgat]

Great video. Thank you

[@fullwaverecked]

Awesome video! Always pondering the canbus when im out first thing in the morning waittfor the bus. Can or can't? Can High or Can Low... Ha! Cheers!

[@donmoore7785]

This is very good. One assumes that the senders synch the beginning of messages with each other, by listening to the bus. Your example of beginning a transmit frame at the exact same time requires this.

[@arduinomaquinas]

Great vídeo 🇧🇷😉👍👏👏👏👏👏thank you bro !

[@arthurfricchione8119]

Wow I’m a new subscriber and have been taking electronic automotive courses but you have presented the clearest explanation of the CAN buss I have watched. Every other video I’ve watched gives you know explanation of or a very simplified explanation. Thank you for sharing I like to know how the system works from beginning to end. One question I have is do CAN systems on automotive use yellow and green twisted pairs and do the stubs coming off the nodes remain yellow and green? I’ll watch again to see if I missed anything. Thanks Again. Artie. 👍

[@steveiak9970]

They can probably be any combination of wire colors. This ford I’m on has white and blue for high speed can. They can use other for can low. Depends on manufacturer you’ll have to study a wire diagram for the application

[@pravallikachennakesavula6667]

Thanks for this vedio .it's very helpful.

[@SunnybraeCroft]

Thanks very informative

[@RicardoMoreiraKrahnin]

Excelent video, I needed this so much

[@seshachary5580]

very educative. Thank you Regards

[@baskarv7975]

awesome explanation

[@user-mq4sn9vj9e]

FANTASTIC VIDEO!!

[@AKKJ420]

Fantastic video.

[@rjmlondon]

Very nice explanation. To put it in to practice, is there a kit available to ty out?

[@kevinikola8003]

Very informative, thank you!! The arbitration is interesting. I work with this on a daily basis and now I know why an ABS module id was changed to 7E6. This gives it a much higher priority as it should be.

[@TestECull]

The ABS module shouldn't need to be on the CAN Bus at all in order to operate. It should be able to operate entirely on its own. Indeed, the only reason it even needs access to the CAN Bus at all is malfunction lights.

[@ethompson526]

It needs to be on CAN to send/recieve signals from other modules that it relies on for vehicle info like vehicle speed, ignition status etc. ABS will not work without these signals.

[@TestECull]

@@ethompson526 ABS will work just fine without those signals. ABS predates CAN by over a decade fun fact. Mercedes was putting it in the S class in the late 70s, GM was putting it in select high end Caddies in the early 80s, and by 1990 p.much anything that wasnt the lowest trim vehicle on the showroom had at least rear wheel ABS. CAN didnt come around until OBD II in the mid 90s, and even then, CAN integration was minimal. ABS didnt really start talking on CAN until the early 2000s when everything started getting other driver's aids that also used wheel speed sensor data. Moreover, there are ABS retrofit kits available for classic cars that are 100% stand alone and are designed to work in a car that is otherwise electrically unmodified. You could put ABS on a 1929 Ford Model A if you wanted to. Ignition status is irrelevant. ABS should be active any time the car is moving whether the ignition is on or not. An ignition switch failure should not mean ABS no longer works. Can also be derived from the simple fact that the fucking module has power in the first place if one wants to avoid another quiescent current draining the battery; if the module is on it knows ignition is also on. ABS is where vehicle speed data is collected these days. It does not need access to CAN to know how fast the car is going. It operates by measuring the speed of each wheel and comparing across the car looking for a discrepancy indicating a locked brake, so by virtue of how it operates, it already has that data. CAN needs ABS to report vehicle speed to the driver but ABS doesnt need CAN since it is collecting that data directly anyway. ABS is perfectly capable of collecting all the data it needs to function without any CAN whatsoever.

[@boldford]

@@TestECull With respect, I think you are looking backwards too much. I agree ABS in one of its various guises has been around for many many years. That said each system was quite proprietary and implemented entirely according to the vendors whim. I see the introduction of CANbus as an enabler to bring, not just ABS but, all those vehicle management functions together over one common resilient transmission media, I'e. Why have multiple systems?

[@TestECull]

@@boldford Part of why I refuse to buy modern cars is how ridiculously interconnected they are. I should not have to worry about a blown out speaker causing a no crank no start fault, and it won't in older cars because the radio is completely standalone. But on these modern ones the radio is on CAN and if the blown speaker takes out the wrong chip in the head unit it will flood CAN with garbage data that drowns out all the other sensors and such the vehicle needs to run. erego, a speaker failure can take the entire car offline these days. ABS has no need of CAN and should be perfectly capable of functioning on its own as it used to do. If we're going to insist on coddling the morons on the road by throwing ADAS at their bad driving instead of actually solving the root problem, let's at least make sure those various systems can and will function entirely independently of anything else on the car.

[@joekrim6557]

Easy to fix. Splice zones are sometimes hard to find when theyre😢not where repair manual states. Usually it's a module that's bad.

[@Bianchi77]

Nice info, thanks :)

[@isra4339]

Thank you very much ! I have one naiive question though at 10:18, is there a particular reason why we consider the zero volt difference as 1, and the 2 volt difference as zero, and not the other way around?

[@benhetland576]

Yes, the 0 needs to be a "dominant" bit during arbitration or packet collision. One way of achieving this with potentially multiple device driving the bus is to let the recessive nodes with their 1 remain passive (i.e. no voltage difference, or hi-Z) while the dominant node(s) force a difference with their 0 bit value.

[@Kaediirus72]

Curious as to what program you are using for the illustrating? Also an excellent explanation of CAN.

[@TOMTOM-nh3nl]

Thank You

[@pyronitro]

Thanks for this. i'm trying to understand so i can fix my own vehicles. thinking out loud here... Terminator resistors absorbs the electrical energy of the signal as it reaches the ends of cable and avoids reflection of signals. So it doesn't become noise. this is similar to BNC type networking cabling back in the 80s-90s. usually theres a limit to the length bec the signal can degrade to unusable or too noisy beyond that length. twisted pair cabling is whats in use in networking today. as he explained. twisting it helps cancel out the EM field each wire generate and make the signal cleaner and go further. right hand rule and such of EM fields moving in a direction. differential signalling is brilliant! i believe it would be educational for people to learn networking concepts as they are very similar.

[@1963TOMB]

Reflections are minimised by having a terminating load that matches the characteristic impedance (a.c. resistance) of the cabling. In radio systems, this effect is specifically tested for during cable installation to ensure maximum forward power, i.e. minimum return loss.

[@Lvvcassss]

What are effects of having more than 2 termination resistors? Eg. devices on the ends of the bus have one each, and by mistakte threre's a third one somewhere in the middle.

[@jeyanthitherce7274]

Thanks sir lesson

[@tomaszstaniewicz7123]

There are to things which are interesting but where not covered in the film (which was very cool and informative btw. :)) 1. How is the bitrate/sampling time aligned? 2. How does the fault detection work and when retransmission is forced?

[@ccarlock8537]

Very goooood

[@theangelofspace155]

If every node needs to be within 30cm (12 inch) how does the OBDII can works? It is far from the ECU, does that mean the man can bus wire run all the way to the ECU? Also how does rhiw change with ECU (motec) that have multiple can ports? Can you make multiple can bus branch? Like, a pair lf teeminating resistor in each for port? Or the port are internally comnected at inside the ECU and you still only need 1 pair of resistor for the whole network?

[@MaPf818]

There are 2 different physical layer CAN types: high and low speed. Would be great to (at least) mention this fact. You talk about high speed CAN. The colors are not standardized or specified. Yellow/Green might be used often. 5:37 Single wire is misleading as ground is still needed -> 2 wire. How the data is detected is the difference. BTW 12V as a solution is a strange idea. 11:30 before, there was the interpretation of the states and now the way how it is done? 15:04 CAN transceiver which might be in a CAN controller but usually there are much more CAN controllers in uC while the CAN transceiver is external 15:23 Not the twisted pair of wires is essential, the differential signal is, which works on 2 lines, here a twisted pair of wires. You can use twisted pair on "single wire + ground" and this will not make the transmission more resilient to external electrical interferences. The differential signal is the key. Stub length (in theory) depends on the CAN speed -> 30cm for 1Mbit/s. In practical applications much longer stubs are common. Best would be to request to move the termination to the longest distance (the bus ends) then. Also there is a weak termination possible. See CAN application note NXP.

[@gerardlunow567]

I bought a BMW Motorcycle and had to learn about CAN-Bus. In inutes I learned that this is very old technology as in Thick-Net over 1/2 coax from 1983. At is a little different but works the same.

[@tomknud]

Why use the jargon PDM without defining it? otherwise great explanation of Can Bus spec, thanks.

[@BBB.oo7]

Hi, do you know what is that ?

[@DodgyBrothersEngineering]

The RTR seems redundant if the DLC is going to specify the length of the data. A DLC of 0 would indicate that only the identifier is being sent and anything >0 tells you that it is identifier plus data.

[@timelapsechannel2573]

Thank you for this video. Maybe a stupid question, but a device that publishes on the bus should also be able to receive its own message, right?

[@bluedark7724]

If you really want to understand what's going on with the twisted wire pairs then the Altuim talk by Rick Hartley is critical. It has to do with electrodynamics in that electrical energy travels around the conductor, not inside of it.

[@franktran3967]

Hi everybody, I have a Honda civic 2010 coupe is false error link when I setup ODB2 scan then I probed CAN-H and CAN-L look ok but L-line is a straight line 9 volt dc, it has no wave form during ignition is on or engine is running. I am not sure what it should look like before I continue to chase it down to find the problem. Anyone has any suggestion please!!! thank you every much. I don't want to bring it to the shop or dealer because the value of my car might be cheaper than the cost to repair.

[@alet1490]

Can nodes other than the PDM and ECU nodes have their own termination resistors?

[@hardwireelectronics3835]

Yes no problem, you just need to make sure that a node with a termination resistor (if enabled) is placed at the end of the bus.

[@atiq19851]

Can some one explain how to write data frames in the PlC, specially in Siemens

[@mba2ceo]

How is control of BUS determined ?

[@riyazahmednavalur566]

How to bypass can wire discanect scr system

[@tomaszstaniewicz7123]

And what is the voltage of idle state - when no device is transmitting? Is that 0 or 2.5V?

[@1963TOMB]

The High line will be steady @ 2.5V and the Low line will be steady @ 0v

[@stanewstanew]

@@1963TOMB Any reference to official doc? According to the guide both lines should be at 2.5 V

[@speggeri90]

Nice video! Although I think you made a slight error here with the maximum 30cm as a maximum distance. I always thought the maximum was 30 m, not 30 cm, and only with a specific transmit rate (1Mb/s baudrate in this case). In many applications the baudrate is often 500kb/s (with maximum distance of ~100 m) or 250kb/s (with maximum distance of ~250 m). And also in short distances the second end resistor is not even necessarily needed, and the bus runs alright. I am open to be corrected.

[@hardwireelectronics3835]

I recommend reading the TI application notes SLOA101B and SLLA270. It is heavy reading, but full of gold if you dig into it.

[@hardwireelectronics3835]

The 30cm refers to the maximum stub length and not the overall bus length.

[@donmoore7785]

@@hardwireelectronics3835 30 cm seems awfully short even for a car - about a foot. If that is true, these devices need to be carefully used.

[@Slicerwizard]

@@donmoore7785 Why would a stub need to be longer than 30 cm?

[@adrianc6534]

@@donmoore7785 30cm is the max distance of any node to the bus, not the maximum length of the bus.

[@prehanramsamy6728]

When you are stoned and click on the wrong video...

[@bobb.6393]

Sobering

[@readme765]

Good times.

[@peerperemans897]

When you're not and clicking on the right video...just as much fun.

[@adamgietzen5009]

What doesn’t kill you makes you stronger I guess

[@micahbuchheit1283]

I became an engineer by tokin' and cruisin' the internet.

[@brothertyler]

Reminds me of 1553B

[@jakesteampson7043]

2:48 And then Windmöller arrived like: "Screw this, both green and yellow will be tied together for ground, we'll add white for high, brown for low, and we'll sometimes use the shielding around the cable as another connection for no apparent reason"

[@Howdy513]

A white pointer on your mouse would be great

[@ExplainingMathematics]

Signal processing, noise, transmission. huhhh all in one intro. you are a gun man.

[@edwinlipton]

I know what a can of tuna is for. A can of oil,, but how do you get a BUS in a Can?

[@sinanbozkurt1823]

Can you add Turkish subtitles?

[@adriancarboni8546]

When you mention the single wire system and noise problems, you show a 5V level having a potential 5V glitch. As a solution you then raise the signal level to 12V to overcome the threshold so the 5V isn't seen anymore. This is wrong! Glitched don't confine themselves to 5V. If there is a 12V system for signals, the glitches more more likely be 12V as well. So that is not a reason to raise signal levels to 12V. Pedantic mode off....

[@marian-gabriel9518]

I don't think that is necessarily pedantic; it's a fair point to raise. He gave that example, I think, because he's referring to external induced noise (EMF) not inherit line noise (which would be symmetrical on both lines by design if the noise is internal to the transceiver or just a fault if one line is electrically impacted by physical short-circuit or open load type faults), raising the line active level far away from the noise floor actually does mitigate this external noise but in a naïve way (no disrespect intended), if the actual operational context of the ECU is not considered. It's much, much more complicated than "just two wires" might lead some to believe, however this isn't hobby level electronics we're talking about and communications in an automotive setting need to actually be guaranteed to a minimum automotive standard (can't remember the ISO number of the top of my head) and depending on the use of the actual signals traveling on those lines, the SW part, at least, of that CAN implementation is held to higher and higher standards such as levels on the ASIL standards (safety standards) as well as (newer) cyber security standards which in the EU at least, will be mandatory by law, in any new ECU sold after mid 2024, if I remember the year correctly. And, regardless, if the operating conditions of a vehicle exposes it to such a high induced external noise on balanced lines, then, even though I'm no expert, it may be that the wrong CAN type is used (in this case "normal" CAN or CAN-HS - high speed or CAN-FD - flexible data rate) so one might use the CAN-FT - fault tolerant - bus that has the lines go from -27V to +40V or even a terminating bias CAN that has power and ground cables besides the two data lines that provides the same electrical bias and terminations for each bus segment, and this is just scratching the surface of the physical layer of the standard with no mention of errors/collisions/bit timings/ bit sampling/faults mitigation happening in all the other layers of an actual CAN com stack.

[@misterzia01]

19:50

[@JENSATAAU]

Dear all. Regarding the canHigh and canLow please take a looke at en.wikipedia.org/wiki/CAN_bus and a figure of canHigh and canLow from the same page: en.wikipedia.org/wiki/CAN_bus#/media/File:ISO11898-3_Waveform.svg This is different from explained here. The whole idea is that (canHigh - canLow) is "1" and recessive when canHigh i low ~ 0v (and recessive) and canLow is high ~ 5V (and dominant). Neither canLow og canHigh will never be 2.5V ! This is a very common mistake.

[@hardwireelectronics3835]

Hi, Thank you for the comment. I believe that you are confused between high speed CAN ISO 11898-2 (the subject of this video) and low speed CAN ISO 11898-3.

[@fu1r4]

After watching all these car issue videos, the can bus seems to be very vulnerable. One tiny fault make the whole can bus go down.

[@clockworkphoto]

I'd you could mute out mouth noise between sentance in your future videos. Make it very hard to lisen too thanks

[@99john58]

RS485 - what you say

[@jeyanthitherce7274]

Post video new basic lessons sir

[@briankleinschmidt3664]

We spell it Cannabis, but I think it's pronounced the same.

[@mxcollin95]

Like the green kind? Lol

[@mohammad45h]

سپاس درود وهزاران لایک بزرگوار محترم استاد تدریس شما بهترین است اما ترجمه بفارسی ندارد لطفا ولطفا اقدام نمایید .همه زبانها وجود دارد بجز فارسی ممنون هستم کمک نمایید 🙏🙏🙏👌👌💝💝😘😘😘😘

[@danbrit9848]

so if a car dosnt work is it a ....cant bus...lol

[@davidhall8874]

Can bus do what?

[@pjmoran42]

Thank God the ethernet doesn't use a CAN bus. We'd still be stringing copper.