mickmoignard
commented
3 years ago - Yes, it was the March 8 version you reviewed. Now updated to 22 March in the header. Draft removed fron "NMRA Standard". My bad.
- RCN-122 Appendix C is a good candidate for a tech note. However, as S 91.1.4 specifically excludes Plux12 I. less certain that appendix D should be carried across, I will create a draft Technote. See also 8 below.
- Using Para 1.2.3 matches the S 9.1.1.3 standard for the 21MTC interface. Indeed I used the general format and layout of S 9.1.1.3 as a guide when drafting S 9.1.1.4,
- Line 23: seven base outputs, 3 logic level. Up to 9 function outputs requires GPIO/A and /B to be full power and thus there would be 9 function outputs but only one logic level output. I don't see an issue?
- I've left "loudspeaker" in 1.2.3 as it is the first such reference. textual references are "loudspeaker" (2 of them but table 2 does use "speaker" and I've left that too. I don't think it confuses.
- RCN-122 specifies the 1mm thick system board and so we should too.
- RCN-122 specifies the pin at exactly 3mm. There are diagrams of the connector in massive detail in RCN-122 which I did not bring across, and they're very clear at 3mm exactly. Is it possible that this actually comes from the supplier of the likely connectors used? I see https://www.mill-max.com/products/socket/852-xx-xxx-10-001000/852-10-050-10-001000?s_type=Quick%2FProduct%2FPart+Number+Search as a possible example (0.118 inch = 2.9972mm). I think we should stick with 3mm to match RCN-122.
- RCN-122 specifies "The impedance of the loudspeaker is determined by the manufacturer of the decoder and must be documented." (my translation). We can go with the 4-8 ohm idea, but we'd then differ to what RCN-122 says. Maybe the Tech note could say that there is a strong recommendation of speaker impedance of 4-8 ohms?
- I've picked up and implemented the other suggestions and the current draft has track changes turned on.
Though the document is labeled with a date of Feb 11, I believe I actually have reviewed the version that was committed on March 8 (26a16f2). Please update the Published Date field each time the document is updated. Be mindful that the Document History table also contains a date entry that may need updated.
The header table does not need a draft label after "NMRA Standard". This is redundant since the "Title" field already contains a draft label.
It looks like there are extra formatting lines that are unnecessary at lines 20, 29, 30, 40, 60, 83, 88, 125, 140, 151, and 153. Sometimes extra spacing is intentionally added in the form of page breaks in order to improve document readability.
The Document History table overflows onto page 7 unnecessarily. Deleting the last unused line is acceptable in order to reduce the document size.
Line 12: TN-9.1.1.4 reference. The content of RCN-122 Appendix C and D are good candidates for a Technical Note. If a Technical Note is not created, remove this reference. I acknowledge that section 1.4 Requirements contains notes similar to RCN-122 Appendix D.
It seems odd that section 1.2.3 Description of the Interface is hierarchically under section 1.2 References. Consider putting section 1.2.3 content in its own 1.x section or perhaps simply fold it into the section 1.1 Introduction.
Lines 22 - 23: I'm counting 12 total outputs (9 amplified/open drain, 3 logic level).
Line 22: Instead of "loudspeaker", consider just "speaker"
Line 27: ...into the vehicle system board.
Line 47: do not abbreviate "maximum".
Table 1: Remove dimensions for "Decoder circuit board thickness" and "Max thickness of components on the underside of decoder. These are entirely up to the manufacturer so long as "c" is met.
Table 1: Free pin length from decoder connector - should this be 3.0mm min, 4.0mm max?
Line 104: do not abbreviate "approximately".
Line 111 - 112: Group 6. We decided with the 21MTC that we should adopt specific impedance of 4 to 8 ohms. I strongly suggest we do the same for the PluX and take the working directly from the 21MTC standard.
Line 148: Remove "are": Which will be "are" connected.