Talk:RS-485
This is the talk page for discussing improvements to the RS-485 article. This is not a forum for general discussion of the article's subject. |
Article policies
|
Find sources: Google (books · news · scholar · free images · WP refs) · FENS · JSTOR · TWL |
Archives: 1Auto-archiving period: 365 days ![]() |
![]() | This article is rated C-class on Wikipedia's content assessment scale. It is of interest to the following WikiProjects: | |||||||||||||||||||||||
|
![]() | The contents of the DH 485 page were merged into RS-485 on 2016-06-22. For the contribution history and old versions of the redirected page, please see Error: Invalid time. its history; for the discussion at that location, see its talk page. |
Merge request
editI'm requesting that this article have its history merged into EIA-485, where this article now is. (Someone else has improperly cut-and-pasted the article instead of moving it.) I made the request on the Helpdesk but got no response. –radiojon 02:00, 2005 Jan 31 (UTC)
Huh?
editThis seems like a nonsense sentence ... "between the ribbon cable?
This allows EIA-485 to implement linear topologies using only two lines and between the ribbon cable.
2-Wire and 4-Wire Implementations
editMight be good to discuss the differences of a 2-wire (half-duplex) over a 4-wire (full-duplex) implementation. This seems to have become an increasingly popular marketing point with various vendors.
multi-dropped, multi-point?
editIn the box at the start of the article there is " multi-dropped, multi-point", what is a difference between the two? Afaik RS-485 is usually done via master-slave, with only one device sending a request on a bus, and then this device responding back within specified time. Can RS-485 be made multi-master? Some form of CDMA, protected by checksums to detect collision? — Preceding unsigned comment added by 2A02:168:F609:0:6EB4:8A5F:9FC4:E8FB (talk) 14:47, 22 July 2019 (UTC)
- RS-485 is only a physical layer electrical interface spec. The choice of protocol is left to the implementer.
- —Trappist the monk (talk) 15:24, 22 July 2019 (UTC)
- The RS-485 physical layer has no provision for collision detection so is really only useable for multidrop protocols that avoid collisions. Master-slave is one example of such a protocol. Token passing is another. ~Kvng (talk) 14:48, 25 July 2019 (UTC)
- When RS-485 allows to pass link drive to another master station then for the smooth operation there has to be defined behavior of undriven link. All datasheets which I have read define level reported by receiver when link is not driven to be steady logic one. Idle state stabilyty can be even supported by high impedance pull-up and pull-sown resistors. Such arrangement allows to define dominant (driven logic 0) and recessive (undriven idle logic 1) states and deterministic arbitration and collision resolution is possible then same as it is used on CAN. The simple collision avoidance protocol requires propagation of signal through whole cabling forth and back and switch to "recessive state" slope is limited by cabling impedance and capacitance so the reasonable arbitration code rate is equivalent to 30 MHz/cabling length in meter (same as for CAN). After arbitration phase the full RS-485 data rate potential can be used for frame payload. We have designed such open-source protocol (uLAN) at PiKRON company in 1990 year and it serves in many our instruments and even projects of other companies for decades. May it be, it would worth to include idea in the RS-485 Wikipedia page. Compared many proprietary and non permissive consortium controlled (like Profibus) solutions this project is available including sources for GNU/Linux, Windows, NuttX and bare metal drivers and infrastructure sources for decades, it is documented by articles presented on multiple conferences for more than 20 years which ensures that no patent bomb can be hidden by us or by others to suppress uLAN use and demand payment like in the case of Profibus Interantional damage control of the attempt for libre implementation. Advantage of uLAN protocol is no need for building of active stations lists, no problem with lost token situation and long time recovery. Full multi-master stack has been implemented even on systems with 256 bytes of RAM memory (only very limited single and short Tx message buffer used in this case and Rx has been limited as well). The full featured stack using 8kB of RAM can support 2kB messages and throughput can be better than CAN FD or even be comparable to future CAN XL standard. When number of the master stations does not exceed 16 then distributed arbitration method ensures round robin access of stations which intend to transfer message. (quote of Byron Jeff about uLan from some discussion between some group looking for suitable RS-485 multi-master protocol in past, intellectually they did not contact us to unite forces: From first glance it's a rotating reservation slot system that uses the break character to seize the line and manage line arbitration. Slot reservation based on the previous master makes it pseudo token passing when all nodes are busy. Not a bad design overall). Ppisa (talk) 09:17, 24 January 2021 (UTC)
- The RS-485 physical layer has no provision for collision detection so is really only useable for multidrop protocols that avoid collisions. Master-slave is one example of such a protocol. Token passing is another. ~Kvng (talk) 14:48, 25 July 2019 (UTC)
More polarity confusion
editThe diagram labels the two signals U+ and U-. I have added a caption and updated the text in the section to clarify that U- = A and U+ = B. If someone wants to update the diagram replacing U- and U+ with A and B respectively so we have consistent signal descriptions in the article, that would be good.
Before correcting me on this, please read previous discussion above and RS-485#Signals. ~Kvng (talk) 17:22, 11 October 2019 (UTC)
Hi Kvng. I believe that the section RS-485#Signals is actually incorrect. If you are available to discuss I am happy to do so. My logic is as follows:
- Understanding that a microcontroller UART outputs a logic 1 when 'idle' (mark), and a start bit is a logic 0.
- ON THE WIRE, the sense of 1 and 0 is reversed. i.e. "UART Logic 1" == "Idle" == "Mark" == "ON" == "Binary 0 'on-the-wire'"
- Therefore in the diagram U+ should be marked A, and not B as you proposed.
- Closest reference, without buying the TIA-485 specification, is this TI document that has an excerpt of the TIA-485 specification — Preceding unsigned comment added by Edward Hague (talk • contribs) 14:28, 18 September 2023 (UTC)
Everybody but Wikipedia is wrong...
editJust dropped by to see what the current state of the article is, and notice that the article is still / again, insisting that the people who use and make RS-485 devices are "all incorrect", and that only the author is correctly understanding the standard. This is laughable, but un-encyclopedic, and unhelpful. — Preceding unsigned comment added by 121.200.27.15 (talk) 05:23, 10 December 2021 (UTC)