# Removal of the physical layer specification

**URL:** <https://forum.opencyphal.org/t/removal-of-the-physical-layer-specification/895>\
**Category:** Specification\
**Created:** [July 30, 2020, 8:51am UTC](https://forum.opencyphal.org/t/removal-of-the-physical-layer-specification/895 "2020-07-30T08:51:22Z")\
**Posts on this page:** 2\
**Page:** 1

<div class="post-metadata">

**Author:** ![pavel.kirienko](https://forum.opencyphal.org/user_avatar/forum.opencyphal.org/pavel.kirienko/32/9_2.png) [@pavel.kirienko](https://forum.opencyphal.org/u/pavel.kirienko)\
**Post date:** [July 30, 2020, 8:51am UTC](https://forum.opencyphal.org/t/removal-of-the-physical-layer-specification/895/1 "2020-07-30T08:51:23Z")

</div>

@scottdixon proposed that we remove the physical layer specification chapter from the UAVCAN Specification document:

> As discussed on the dev call, the physical layer specification for CAN  
> transports is significantly less mature than the rest of this specification and is  
> also the most difficult layer to standardize in generic terms. Proper  
> functioning of the CAN data-link layer requires properly designed  
> electrical busses with properties specific to the physical  
> dimensions of a vehicle, the topology of the network, the number of transceivers connected  
> to the network, the location of the transceivers in relation to each other,  
> and the electrical properties of each transceiver. As high data-rate  
> variants like CAN-FD 1/5Mbs are deployed this electrical engineering  
> will become even more important even for very small CAN busses.  
> Because of this we are removing the physical layer specification from  
> the core document but reserve the right to create sibling documents  
> that standardize physical UAVCAN networks based on vehicle morphology  
> and supported transports.

> <https://github.com/OpenCyphal/specification/pull/95>
>
> As discussed on the dev call, the physical layer specification for CAN
> transpor…ts is significantly less mature than the rest of this specification and is
> also the most difficult layer to standardize in generic terms. Proper
> functioning of the CAN data-link layer requires properly designed
> electrical busses with properties specific to the physical
> dimensions of a vehicle, the topology of the network, the number of transceivers connected
> to the network, the location of the transceivers in relation to each other,
> and the electrical properties of each transceiver. As high data-rate
> variants like CAN-FD 1/5Mbs are deployed this electrical engineering
> will become even more important even for very small CAN busses.
> Because of this we are removing the physical layer specification from
> the core document but reserve the right to create sibling documents
> that standardize physical UAVCAN networks based on vehicle morphology
> and supported transports.

If the change is accepted, the upcoming DS-015 standard will inherit this chapter mostly unchanged.

We will leave this issue hanging for a short while to make sure that every interested party has enough time to raise objections.

FYI @TSC21 @tridge @jani.hirvinen @Gijs_Bruining

---

<div class="post-metadata">

**Author:** ![Kvide](https://forum.opencyphal.org/letter_avatar_proxy/v4/letter/k/3bc359/32.png) [@Kvide](https://forum.opencyphal.org/u/Kvide)\
**Post date:** [July 30, 2020, 9:36am UTC](https://forum.opencyphal.org/t/removal-of-the-physical-layer-specification/895/2 "2020-07-30T09:36:39Z")

</div>

> [@pavel.kirienko](#):
>
> upcoming DS-015 standard will inherit this chapter mostly unchanged

A side note, as I’m not involved in any CANdrone work:

CANaerospace v1.7 suggests using a floating CAN signal pair without (CAN) ground, whereas the now removed UAVCAN physical layer chapter mandated CAN ground to be connected inside the devices (as it is specified in CiA 103 I believe).

For configurations where part of the bus is on the hot side of a spark-ignition engine firewall, floating signals, optionally also isolated, is the sound way implement the bus.

Br,  
Christian
