SN65HVD234: ISO 11898-5 Compatibility

Hi Eric,

Thanks for the information.


Here's a feedback from our customer:

The latest ISO 11898-2 mentions "biasing active" and "biasing inactive" as transceiver modes. Does the SN65HVD234 conform to the updated 11898-2? When I measure our unit with the SN65HVD234, I notice that when the EN pin is low (disabled), the CANH and CANL outputs seem to float to about 3 volts. The state of the RS pin makes no difference when EN is low. Our customer believes these outputs should go to 0V when disabled.

Regards,

Cedrick

Hi Cedrick,

The low-power mode of SN65HVD234 does conform to the requirements of ISO 11898-2. Keep in mind that this standard does not specify characteristics for 3.3V supplied CAN transceivers. For this reason, no 3.3V device can be compliant with the standard. We claim to be compatible with systems that currently follow the ISO 11898-2. This will include behaviors like low-power modes and compatibility with 5V transceivers. 

When SN65HVD234 is in Sleep mode (EN=0), the driver and receiver are both off and the bus is left in a high impedance state. At this point, the state of the bus will be primary controlled by other connected devices or transceivers which may have a different bias point than SN65HVD234. 
A passive recessive voltage of 3.0V is uncommon. Is this instead close to 2.5V? Perhaps with some ground offset from the biasing transceiver? This would be a typical recessive bias for a 5V transceiver.

Can you share a block diagram of the system and all connected nodes? I would like to know what other devices may be including the state of the bus under these conditions. Please include placement and type of termination in the network.

Regards,
Eric Schott

Hi Eric,

Just got a feedback. Kindly see below:

Our circuit is pretty simple. I have attached the schematic in a Word document. Pins 1, 5 and 8 are connected to the main uC, and I can control the state of these pins using some debug code. IO Pin 11 and IO Pin 12 go to the main connector on the product. There is a 120 ohm resistor connected between those pins. I have tried removing the transceiver, and there is no voltage on CANH or CANL in that condition, so I assume it's the transceiver that floats to 3 volts. When two modules are connected to each other via this CAN circuit, the bus will be at bias voltage unless both transceivers are disabled (EN = low). In that case, the bus floats to 3 volts. The state of the RS pin does not matter. CAN circuit.docx

Regards,

Cedrick

Hi Cedrick,

I don't believe I understand this schematic fully. Can you please clarify a few things?

What are IO Pin 11 and IO Pin 12? Are these connected to a uC? Or are these only connected to other CAN transceivers and termination resistors? Can the connections on these pins be shared? How many other transceivers and termination resistors?

The capacitor and inductor network shown on the CAN bus here is uncommon. Could you describe the purpose of this circuitry? 

There is a 10k-ohm series resistor (R155) on the RXD signal line. This is a very large value for a resistor on a digital signal line. This could cause a significant voltage drop on the output signal and increase the transition time of the signal, effectively increasing the propagation time of the signal. As CAN protocol is particularly sensitive to propagation time, this should be avoided. Values from 10-33-ohms for digital line series resistors are more typical. 

Is the intention of this CAN transceiver to drive standard CAN data? I ask because a few aspects of the schematic are not consistent with a typical CAN circuit. I would recommend referencing Figures 38 and 41 on the device datasheet for reference termination and layout examples. Also see the TCAN1042EVM for a full example schematic of a CAN bus which contains several configurability options. Let me know if you have any questions on these. 

Regards,
Eric Schott

Hi Eric,

I got a feedback from our customer:

I guess the schematic could use some explanation. IO Pin 11 and IO Pin 12 go to the main connector of our product. A wiring harness connects the CAN leads to the vehicle bus usually through the OBD connector. There will be at least one, probably two 120 ohm termination resistors on the CAN bus in the vehicle - we do not terminate inside our module. There could be any number of other nodes on the CAN bus. The capacitor and inductor shown is actually a feedthru cap. It is there to prevent RF generated in our module from leaking out onto the CAN lines. R155 was originally there to limit the current from the RXCAN pin into our microprocessor, which could be powered down when not used (the CAN xcvr is always powered). Since our power management strategy has changed, I will change this to a small value.

Regards,

Cedrick

Hi Cedrick,

Thank you for relaying the feedback. This all sounds good. There are a few things I would like to address, but I don't see any major issues here.

CAN transceivers expect to have very close to a 60-ohm load present on the bus. Having only one 120-ohm termination resistor between the CAN lines may decrease performance and slow dominant-to-recessive edges. I would recommend ensuring that two 120-ohm termination resistors are included on the CAN network. 

RF generation reduction for electromagnetic compatibility (EMC) is typically addressed in CAN designs with the use of a common mode choke (CMC). These passive components filter out common-mode noise and ensure that the majority of the energy that travels through the CAN network is differential. This reduces the energy radiated by the setup. A common choke we recommend is ACT45C-510-2P. I'm not personally familiar with a CAN system that has used a feedthrough capacitor for EMC. It seems to me that the capacitance added by such a component would negatively impact a CAN signal in any sizable network. 

Regards,
Eric Schott

Hi Eric,

Thank you for your response.

I've got a feedback from our customer. Here it is:
There is usually two 120 ohm termination resistors on the vehicle bus. We don't provide one internal to our module. You're right, the common mode choke is the more common way to reduce RF emissions. The caps work fine with standard CAN, but we have to remove them on CANFD to get maximum performance.

Regards,

Cedrick