EEVblog Electronics Community Forum
Electronics => Projects, Designs, and Technical Stuff => Topic started by: Simon on August 13, 2024, 10:46:59 am
-
I usually use a common mode choke ( https://www.we-online.com/components/products/datasheet/744235251.pdf (https://www.we-online.com/components/products/datasheet/744235251.pdf) ) and a single termination resistor on a CAN bus.
Lots of designs I see have the termination resistors split into two 60 ohm resistors in serios and a capacitor in the middle to ground. This is apparently for common mode noise rejection, but that is what the CMC is for ?
Do the two solutions complement each other or is it best practice to use both together. I have just found lots of reference to 4.7nF being the value to use and at most 100nF. But most of the designs I have previously seen have used anything from 1µF to 10µF. I realize that this value has no actual effect on the filtering as at all times this capacitor should sit at 2.5V. When recessive both bus lines sit at 2.5V and when dominant the voltages are symmetrical around 2.5V so still 2.5V will be present.
Texas instruments put out an application note about the dangers of common mode chokes but it turns out that they are simply warning that if the CAN bus is shorted to the supply voltage there could be a large back emf spike. Well if you have problems with bus lines shorting to a supply voltage maybe you really need to sort that out rather than blame the CMC for ruining you day.
Maybe the split termination is a more recent development to cope with higher speeds than what the differential inductance of many CMC may allow for?
This seems to be an area where many people are just doing what they always did without question. I don't want to load my designs up with a whole load of parts that I don't actually need.
-
...This seems to be an area where many people are just doing what they always did without question....
This is done by the same people who came up with the polarity of the non-magnetic coil in pulse power supply. Do as usual, with two 120 resistors at the ends of the longest line.
-
Consider a transceiver in two states: dominant and recessive.
In the recessive state, the transceiver is open-circuit, and CM voltage is defined as the difference between instantaneous bus CM voltage and local ground. The receiver is thus susceptible: exceed its CM range and gibberish is received.
When a CMC is used, its voltage drop is small: the impedance at radio frequencies might be a couple kohms, but the transceiver is high impedance as well. Thus the CMC does almost nothing on receive. (At high enough frequencies, transceiver capacitance takes over and some filtering value can be had.)
In the dominant state, the bus voltage is either forced high, or the transmitter acts in parallel with another transmitter, and both share bus current to some extent. Either way, the impedance is low, or at least the voltage difference between bus and transmitter will stay low. The bus CM voltage is thus forced to local ground (or, well, somewhat halfway between GND and VCC, but referenced to GND is the point).
When there exists a galvanic (say DC or mains-frequency) voltage difference between two nodes, their alternating transmissions will pull the bus CM voltage up and down, thus emitting CM radiation coincident with the transitions into/out of dominant state. (This might be contained in a shielded structure, cabling, cableways, etc., or it may couple to nearby cables, or it may be exposed to space and radiate per se; I use "emitted" generally here.)
When the bus is long, its capacitance to surroundings, or more generally its CM transmission line impedance (whatever that may be; in general it varies with position along the bus, due to variable proximity to surroundings), is relatively low, and a CMC has some impedance to work against. Thus transmitter CM emissions can be reduced.
For the special case when a node is near/at the end of the bus, it can be terminated nearby, and that termination resistor can be tapped to obtain a common mode reference to the bus, independent of dominant/recessive transmitter state.
When a termination is tap-grounded (via bypass capacitor) and combined with a CMC, very high CM attenuation is possible (i.e., kohms into 30 ohms), and over a broad frequency range. Conversely, emissions are reduced, as the node presents a fixed (DC) CM voltage, rather than a dependent one.
Since there are at most two terminators on a bus, this isn't a general solution, but in the special case where nodes cluster only near the ends, or point-to-point communication is used, very high signal quality can be obtained even through a rather mean environment. It's basically like Ethernet without the transformers, but with enough CM input range that it can handle most anything else (and is also slow enough, and robust to errors, that the remaining stuff -- EFT and ESD most likely -- can be dealt with; something Ethernet doesn't have a luxury of, hence the transformers -- and CMCs).
Tim
-
So essentially both are a good idea which is what I was dreading for my component count :). So midway devices that are not at the end should have a choke anyway. At the moment i am dealing with very short cable lengths although I will be using ribbon which is not the best but with a bus length of under 500mm I figured not an issue. Any cable that would leave the machine would be twisted.
I could say that given the short distance with devices connected by a 150mm cable and the power to these little things being isolated the chokes may be superfluous but for the fact that I will have to attach the ground at one point to a motor driver and um, I don't know if now my ground will be a bit shaky making CMC rather useful.
-
It's worth understanding / appreciating / reiterating that CAN, like RS-485, etc., is a common-ground bus, and not truly differential like we might hope it to be. (Case in point: CM input range.) A receiver can be fully isolated, without using the tapped-terminator-ground strategy, by finding a (much weaker) ground through the input divider resistors, but this is quickly swamped by pin capacitance; it works at DC and low frequencies, but not over a wide range.
500mm is definitely short enough that everything can be treated as a local node, I guess unless you're using an especially high-speed CAN mode. Certainly it's fine up to several Mbps. If those local things can also share ground (in a reasonably low-impedance EMC sense), then a single CMC for the bus entering the local area will suffice, no need for per-transceiver CMCs.
Tim
-
Well Yes CAN Bus is a funny one with it's 2.5V offset. I guess that at the time it was designed either they just did not think of using a full bridge output or it was not feasible. I am certainly not operating above 1 Mbps.
-
This is the TI appnote referenced? https://www.ti.com/lit/an/slla271/slla271.pdf (https://www.ti.com/lit/an/slla271/slla271.pdf)
You should remember CAN bus was developed for automotive use, and thus chassis was always near and often referenced.
No need for a fully balanced system. (it would be expensive)
The common mode choke can be used to block some hf noise exiting the node, radiating with the cables.
I don't think you need a choke unless you fail EMC test on the CAN bus wire and can't find the actual emitter.
-
Yes that one.
-
I meant that the coil CMC after the transmitter looks good, when you design the PCB - it should be routed there, but shorted with lines suitable for breaking. But dividing 120 ohms into 2 x 60 with a capacitor is a perversion.
-
I meant that the coil CMC after the transmitter looks good, when you design the PCB - it should be routed there, but shorted with lines suitable for breaking. But dividing 120 ohms into 2 x 60 with a capacitor is a perversion.
Fortunately, EMC doesn't care what you think of it ;)
-
From an EMC point of view, it is much more efficient to put a ferrite bead on the pulse power MOSFET, which is usually not done and there is talk about the polarity of the coils.
-
The point of the split termination with the bias capacitor is to clamp the 2.5V centre point at 2.5V regardless of common mode transients between the nodes. Whether in the active or recessive state the centre point should always be 2.5V so it does not load the bus at all once the capacitor is charge at switch on unless the bus tries to shift it's common mode voltage, which will have the effect of filtering that out if a mere pulse.
-
The point of the split termination with the bias capacitor is to clamp the 2.5V centre point at 2.5V regardless of common mode transients between the nodes. ....
No. The 120 ohm load is located at both ends of a long line, one of them has a 150 amp motor, the overall level is shifted because of this. As a result, it turns out that the capacitors are connected asymmetrically and create interference.
-
well this is kind of what I need to defend against. I depends on many things, every circuit is a collection of parasitic properties and you have to hope that the property you designed for outweighs them.
-
The center tap capacitor circuit is a fool's idea, don't use it.
Even a schoolboy couldn't come up with something like this, complete incompetence.
It's no wonder you got cognitive dissonance from this diagram.
-
You don't accumulate 17900 posts on an electronics forum overnight. It baffles me why you are struggling with basic termination on a CAN bus. :scared:
Also, beware of connecting tesla coils to your canbus, their back emf is known to cause troubles on a CAN bus. :blah:
-
.... It baffles me why you are struggling with basic termination on a CAN bus. :scared: ...
Firstly, the disadvantages of a circuit that is "basic", and only for you, are obvious, and in industrial devices, of course, no one does this.
Secondly, use technical arguments available to your mind, instead of using the word "basic".
(Because "basic" means appnote from TI.)
-
You don't accumulate 17900 posts on an electronics forum overnight. It baffles me why you are struggling with basic termination on a CAN bus. :scared:
I am in fact not struggling with basic termination, that is a single resistor and nothing else. post count is no indicator of anything. The fool thinks he knows everything, the wise man knows that he can never know everything.
Also, beware of connecting tesla coils to your canbus, their back emf is known to cause troubles on a CAN bus. :blah:
What? and I thought I was going to do that and magically power the whole machine off of free energy, bugger, now what do I do?
-
The center tap capacitor circuit is a fool's idea, don't use it.
Even a schoolboy couldn't come up with something like this, complete incompetence.
It's no wonder you got cognitive dissonance from this diagram.
Split termination provides termination to single-ended signal on the bus, which can happen due to propagation delay mismatch, or simply from the fact that that it's not a true differential bus as Teslacoil said. This is a great technique for on-PCB high speed signals (in the many GHz), and of course for a 3-terminal CAN bus; in both cases it greatly reduces common mode emissions on the cable..
-
when your total cable length is below 2m or so then 1 side termination is not a problem. We use 500kbps CAN in industrial applications 40m total cable length.
I will mention all devices (20+) do all have galvanic isolated can drivers and common mode chokes. The average bus load is about 2000 mostly 8 byte messages / second.
Benno
-
.... This is a great technique for on-PCB high speed signals ....
True. But we are dealing with a long line, with a split load (two ends of line) and unbalanced outputs (low/high).
It is probably easier for you to assess the situation when at one end of a long line the capacitor is connected to ground, and at the other, where the motor is, the capacitor is connected to the power supply of this motor.
-
.... It baffles me why you are struggling with basic termination on a CAN bus. :scared: ...
Firstly, the disadvantages of a circuit that is "basic", and only for you, are obvious, and in industrial devices, of course, no one does this.
Secondly, use technical arguments available to your mind, instead of using the word "basic".
(Because "basic" means appnote from TI.)
Bruh, it's okay to admit you don't know some things. You don't have to rub it in our faces repeatedly. :palm:
Tim
-
...You don't have to rub it in our faces repeatedly. :palm:
Sorry. I misunderstood at first, but then I didn't correct it.
-
I have used CMC with a 120 ohm resistor until now with no issue. I think on my last prototype I may have only had one termination over a short bus. I have in the past used a single termination with no issue but i have never actually looked into the effects and properly evaluated.
My question was more academic. You see as I am not a fool that assumes I know everything because I have not had a problem so far. I know that what I have done so far is relatively tame although having two motors joined together mechanically created some interesting situations transient wise and in the future I will really get going so I was interested in hearing peoples insights into the various termination solutions.
-
Yeah, a single terminator is fine at short bus lengths and low (relative) bitrates -- stub lengths too, for the same reason. Same reason again why nodes clustered together can share a terminator, or CMC or what have you.
Have worked on one application, where just a little more (than normal tolerable) stub length was needed, and communications problems resulted. The bitrate was low enough, but signal quality was the problem. And no, return routing (for a proper chain bus) wasn't an option, sadly. The solution for that particular case was to make a high-frequency line splitter, i.e. put a 60Ω ferrite bead to all lines, in the junction box. Signal level after the leading edge is perfectly fine, but the between-stub resonances are attenuated. This degrades the rising edge, limiting baud rate and bus length, but neither was an issue for that application.
The tapped terminator, with CMC, I would expect is generally better than not using it; it's just not generally applicable because you have at most two terminators on the line.
As far as academic questions, academic answers should be fine, so the "it depends" (but, more importantly, how and when!) is a natural fit. :)
Tim
-
Do it this way:
(see picture)
... The tapped terminator, with CMC, I would expect is generally better than not using it; ....
Your opinion is as always valuable.
-
And just for completeness, the proposed alternative.
[attachimg=1]
(*) Intermediate nodes can't be treated the same way, so default to the base case regardless. (Note this means the proposal is inclusive, it's a superset of the base case; only the end nodes can be terminated in this way, and it is optional whether one or both do.)
V_N shown to suggest noise; note that in the terminal cases, the split terminators with CMCs are highly effective against interference (large CM attenuation ratio), but the intermediate nodes see no such benefit while in recessive mode.
Also consider the case with terminator inside the CMC (notice the CMC is itself a stub length!!) versus outside, or if the distance between say nodes 1 and 2 is small enough and they share GND so that they can be behind the same CMC, which then realizes the grouped-nodes case I mentioned.
To be perfectly precise, note I'm not saying one way is strictly better than the other; I would need to know all possible use-cases, sources of interference, loads, etc., and that's just not possible, I don't have nearly enough experience in this (and, I would suspect no one else yet in this thread does, either). But I can claim (with some justification) that the tapped-terminator case is generally better (lower emissions, higher immunity) -- when the optional is available to do so -- and that this will be true for more cases than the default connection is.
Not even sure what circumstances I could craft that would make it strictly worse; maybe in a double-isolated case, where the bus itself is floating, and now the fact that it's ~statically anchored to one ground more than others see a different impedance, and, maybe it's still fine for the given node, but the others are worse or something? I'm kind of grasping at straws here, honestly.
It's great when you can do it is the point, you can just do it at only two points.
Perhaps it's bad for bulk current injection; automotive heads might offer some insight here. BCI is a weird test; they don't care how the current gets in, they just crank up the test level until it's met. Scary, probably wildly unrealistic when impedances are extreme, but if one wants to play against such tests, a lower impedance is desired. Mind, that's the whole point, the split terminator does better with BCI in and of itself, but the real question is the CMC, does it saturate, does it catch fire, when that much (10s mA? at RF) current is applied?
Tim
-
And just for completeness, the proposed alternative. ...
"CAN transceiver" and "ESD protection" are essentially one chip. Diodes are added very rarely. The parameters of the internal protection of the transmitter comply with ISO requirements.
https://cccsolutions.eu/wp-content/uploads/2017/08/ISO-7637-22011E-STANDARD-CCC-Solutions-AB-Sweden-.pdf
-
Well, transceivers are all over the place. Maybe it's integrated, maybe not. Certainly, TVS for the purpose remain available in great quantity, implying a need somewhere.
Tim
-
And just for completeness, the proposed alternative. ...
"CAN transceiver" and "ESD protection" are essentially one chip. Diodes are added very rarely. The parameters of the internal protection of the transmitter comply with ISO requirements.
https://cccsolutions.eu/wp-content/uploads/2017/08/ISO-7637-22011E-STANDARD-CCC-Solutions-AB-Sweden-.pdf
That standard is about the resilience of the device to disturbances in the power input. It won't cover "what do you do if for some unknown reason out of your control on one specific installation due to cabling or equipment not designed or supplied by you the can bus has some static or other transient on it?". For this reason people put TVS diodes, because why have to argue with a customer about their shitty use of your equipment that they now say is failing and it being your fault.
-
.... people put TVS diodes, ...
I have seen industrial device with TVS on CAN bus only once. I was very surprised.
I have never seen a 60 + 60 + capacitor terminator. I would wash my hands after touching such a device.
-
Guess you have to wash your hands every time you touch something with Ethernet...
-
.... people put TVS diodes, ...
I have seen industrial device with TVS on CAN bus only once. I was very surprised.
I have never seen a 60 + 60 + capacitor terminator. I would wash my hands after touching such a device.
LOL... I guess you have to wash your hands every time you drive a modern car
-
LOL... I guess you have to wash your hands every time you drive a modern car
I don't really have any experience with car automotive electronics for some reason, it would be nice to see a photo of the unit with such a terminator if you have one.
-
LOL... I guess you have to wash your hands every time you drive a modern car
I don't really have any experience with car automotive electronics for some reason, it would be nice to see a photo of the unit with such a terminator if you have one.
Here’s a gateway module from a Ford truck, this has up to 6 CAN buses going to it. They also have 27v TVS on each CAN wire.
Transceiver “TJA1042”
TVS marking “27A”
-
... Here’s a gateway module from a Ford truck, ...
This is the first time in my life I see this!
Thanks for the photo, very useful.
Here is the CAN of the Liebherr port crane, 4xTJA1050.
-
... Here’s a gateway module from a Ford truck, ...
This is the first time in my life I see this!
Thanks for the photo, very useful.
Here is the CAN of the Liebherr port crane, 4xTJA1050.
Tesla AP ECU. Check top left, 5 CANs populated. 62R split to cap and CMCs..
-
... Tesla AP ECU. Check top left, 5 CANs populated. 62R split to cap and CMCs..
I see 3 out of 5 have a terminator. It looks like it's an automated design system that has had this termination scheme added.
-
More likely they're put on by default, and then depending on where along each bus the module is wired, the terminators are used (or not). No need for jumpers because Tesla's buses are all designed (famously eschewing wiring for a printed/fabbed bus). Again, termination works fine for terminal nodes, N/A for middle nodes.
Like, I'm not just saying a bunch of things, there's a reasoning process underlying all this; and that reasoning is universal, it's not just read arbitrarily out of an arcane tome that everyone seems to have but no one seems to know (but, there are those that come close, like Ott's Electromagnetic Compatibility Engineering), it is possible to learn these ground truths (electromagnetic, transmission line and a little network theory) and derive further ones (how to design a bus) from that.
Tim
-
It turns out that in the past there were only fools, and the bus was connected in one way, and then 20 years later, especially gifted people appeared and "improved" it. Well, maybe.
-
It turns out that in the past there were only fools, and the bus was connected in one way, and then 20 years later, especially gifted people appeared and "improved" it. Well, maybe.
You know, the world isn't all black and white. It's a continuum. Maybe a simple resistor termination passes EMC in 99% of cases. Maybe split termination passes with 2dB more margin, and in 99.4% of cases. Therefore it probably isn't "wrong" to do it either way.
-
Well most of the discussion or papers I found online say 4.7nF -100nF for the capacitor. But I have seen 1µF and 10µF in some designs.
What I am actually considering now as I have two CAN bus controllers at my disposal is to have one "fast" CAN bus and a slower single wire CAN bus. As I think about other projects I need to work on and the evolution of dealing with each project by having a collection of little boards that all talk to each other I will need a board to just connect a trigger control to the system, at this point a transceiver on every board starts to become cumbersome. I would assume that this would make my transceiver a simple open collector transistor with a pull up resistor at each end. This makes small local systems easier with a "fast" CAN bus connection on one device to connect the local system to other parts of a larger system.
Sort of the LIN bus and CAN bus combo logic as why write all that LIN code, my employer would kill me, they are still getting over me digging my feet in the ground and insisting on doing CAN bus (I have not told them that I am rewriting that hehe :) )
-
In the Tesla ECU pic, I also note a pair of capacitors on both sides of the CM choke.
I imagine an EV is an (electrically) extra noisy environment, each motor/VFD might be CAN.
One thing is never assume CANBUS is perfectly balanced. I had some crappy shielded cable with non-symmetrical capacitance to shield, on the twisted pair. Drove me nuts but it was the insulation thickness different between the conductor's two jacket colours.
The SOT-23 might be a Bourns TVS for CANBUS (https://www.bourns.com/docs/technical-documents/technical-library/chip-diodes/application-notes/bourns_can-bus_circuit_protection_appnote.pdf) and they use split term in that app note.
-
...Therefore it probably isn't "wrong" to do it either way.
The whole point of the CAN bus is the ability to shift the ground. I don't understand how a capacitor can have a positive effect on different potentials. Let's assume for simplicity that the ground shift is 5 volts, as if it were a snapshot of interference. Then the CMC coil is already unbalanced, and it can be thrown away. The signal is modulated, the fronts are shifted. What's the advantage here?
-
I'd guess the split termination with capacitor is meant to reduce common mode RF noise from the bus, while the common mode choke is meant to reduce emissions that result from an asymmetric delay in the bus driver output. It's digital, basically no failure caused by noise. Maybe some retransmission.
CAN is somewhat similar to USB, except lower bandwidth. Often the bus connectors include power and ground and often the bus cable is similar to a USB cable: shielded with four wires of controlled impedance. Power on the bus can be used for isolated bus interfaces.
Regards, Dieter
-
The whole point of the CAN bus is the ability to shift the ground.
Is it? I thought that is was a differential signal. You will notice that any digital signal of some frequency or having to travel some distance is differential over a twisted pair, this is so that the inexpensive cable will carry the signals easily. The twisted pair means that for complementary or differential signals the inductance of the cable is greatly reduced so the signal can be of a higher frequency and/or travel further. The drivers and receivers themselves are pretty much single ended but act together.
The only way you can guarantee different ground potential is with isolation as for any differential bus type. Most transceivers have a common mode rating of several volts as an absolute max rating to make sure they deal with power voltage spikes. Typically you see something like Vcc+7V, well 5V is the normal supply for CAN bus and putting 7V on that takes you coincidentally to 12V.... the supply voltage of most automotive systems, even on an EV there is a 12V subsystem.
-
I'm using RS485 for my home network (common mode very similar to CAN). and I connect both power (+24Vdc + GND) and data through CAT5 cable. (Still have two spare pairs) On long cables and high load current (bunch of nodes on the far end) there can be a few volts lost over the cable. The data is low current, so it does not have this shift. As a result, Some nodes receive date shifted by a few volts up compared to their local GND, while other nodes receive data that is a few volts shifted in the negative direction, because each node that sends data uses it's own local GND reference.
The current though the power supply wires does not only drop the +24V wire, but it also shifts the GND wire, and thus the local GND for all nodes connected to the cable. And this is the reason why both RS485 and CAN have a common range for their data signals 7V outside of their GND and (usually) 5V power supply. (Some RS485 and CAN transceivers work with a 3V3 power supply. Data transmission only needs 100mV or so differential so this hardly changes reliability.
-
Yes there is some allowance for common mode shift, but then in the same way any digital input will have a tolerance. My micro controller inputs will see a 1 if the input is more than 0.7*Vcc, this means that any voltage drops on the ground as well as from the high side driver are allowed for.
I don't think you normally want to have a signal ground also carrying enough power to significantly shift the reference of the data lines. Again, for this reason most people use isolated connections when this is a high risk and again they do this because most people are designing something for mass implementation where the exact installation of every single device cannot be predicted. So they take a significant step to make their devices resilient to this if there is a risk of ground shifts that are significant. The ability of the transceivers to deal with some offset is usually reserved for the unexpected situations. If you use all of your margin for normal use case you have nothing left for when you have an unexpected situation.
It's kind of why I created the topic and asked. Because so far what I have done has always worked, but I not know how marginal my designs have been. You have created a bespoke system in your house where you know the exact characteristics. Would you sell the same design to the mass market to do what they liked with it?
-
The whole point of the CAN bus is the ability to shift the ground.
Is it? ...
Yes.
-
For example if they would use the car chassis as a common ground to distribute energy in an EV like they do for the starter of a conventional car, this could become a disaster for a CAN bus.
I am currently testing a USB device that needs about 120 mA and its 5 V supply drops to only 4.3 V over the cable from the hub. Half of the voltage drop - that is 0.35 V - could be along ground, unless the cable uses its shield to strengthen the ground connection.
CAN transceivers might use voltage dividers on their inputs in order to extend the common mode input range beyond supply rails.
Regards, Dieter
-
... CAN transceivers might use voltage dividers on their inputs in order to extend the common mode input range beyond supply rails.
This is true.
-
Yes the CAN bus voltage ground will shift. But to rely on that too much as part of design is a bad idea. What happens when there is a transient in the separate power circuit whose ground is also your can bus ground, ah, that common mode input range helps it cope, but if you are using that range by design then you can't use it to deal with the unexpected.
So I have a 30A motor driver that can do 60A briefly, what sort of shit will that throw around my ground as the designers have chosen to not isolate the CAN bus, because of an amount of common mode range that won't be a problem, but now if I put my other can bus nodes very far away and carry power on the same cable that uses the can bus ground as the negative supply and use up that common mode range in voltage drop now what happens when there are motor driver transients?
You can use the common mode range for whatever you like, just don't use it twice....
-
Yes the CAN bus voltage ground will shift. But to rely on that too much as part of design is a bad idea. ....
Well, the CAN bus was invented by fools. A new generation grew up and came up with the idea of putting a capacitor there.
... that common mode input range helps it cope, ....
With a capacitor present, the input signal is out of phase because the load cannot conduct current between the drains while being pulled up to a different level.
... what happens when there are motor driver transients? ...
Without a capacitor, the exchange on the bus will proceed normally and nothing will happen.
I also assume that a capacitor with a capacity less than 1/4 of the line capacity will not create a negative effect, but then it turns out that it is not needed at all, since the introduced imbalance of the CMC coil will negate the removal of interference to the near ground.
-
Common mode chokes were introduced for CAN to handle the mismatch in timing on CANH and CANL due to difference between N and P channel transistors in the transceiver. As transceiver got better, but still maximum 1 Mbits/s, it was possible in some cases to remove the choke based on the EMC test results, but it also depends on different OEM's requirements. Some specify the CMC as a must, some as optional if it passes EMC without it. With higher bit rate, like CAN FD, the CMC might be needed depending on the application.
Termination has also changed over time. 10 years ago it was most common to leave it out and the OEM handled it on the vehicle harness. One German OEM had a specific requirement for a 120 ohm termination with an open end to be bridged in their harness if the unit was located at the end of the bus. A few years ago some OEM's started to ask for high impedance split termination with 2*1.3k.
Being able to handle the full supply voltage on CANH and CANL for more than 1 minute was also mandatory. This was also part of the miswiring test in ISO 16750-2 were all combinations were tested in a sequence. This test has implications on both transceiver selection (CM voltage range), TVS breakdown voltage and termination resistor power rating.
TVS diodes were not optional for the product I worked with as most transceiver handled up to 8 kV, but the OEM requirements were 15 kV for ESD.
One thing to look out for is connector pinout and how the CAN bus wires are located with respect to other wires with transients. On an older product our pinout on an internal harness was Vbat, CANH, CANL, 0V. During the ISO 7637-2 Pulse 1 immunity test (-600V, 50 ohm on Vbat) we got error frames on CAN. The product had a CMC and 120 ohm termination. Problem solved by twisting the internal harness to minimize the loop sizes and hence the mutual inductance coupling between the supply voltage pair and the CAN bus.
Also the wire insulation can have an impact as the harness impedance will change with material and thickness. Not sure how much it will affect immunity, but it will impact emissions.