This document is a good idea, and overall I like what I see in this first draft.
Overview
The first thing I would consider is: Who is the audience? I propose that there are two intended audiences:
average non-technical users
advanced technical users
I think near the beginning of the document the two different audience types could be acknowledged with some recommendations on how each should read the document:
average: Use the provided recommendations as is and expect trouble free operation
advanced: Use the theory provided in making the recommendation in order to understand how the recommendations may be modified for special use cases.
In the body content of the document, there should be a clear separation between stated recommendations and the theory from which the recommendations are derived. The separation could be as simple as a new paragraph or as complex as a new sub-section at the discretion of the author.
The text body contains a number of second person statements. While these are not necessarily bad, using third person statements tend to give a more formal feel to the document.
Detailed Comments
In the Terminology Table, consider the following changes:
Change "Bus" to "Track Bus" with the following definition: "Pair of wires ... to the track. Sometimes abbreviated in this document to "Bus".
Change Power Station definition to match the definition in the draft S-9.1.2 standard: "A device that amplifies the low current DCC electrical signals transmitted by a Command Station for the purpose of providing high current DCC signals with sufficient power to operate model trains and any accessory decoders that are connected to the track. Also known as booster or power booster."
Line 23: than the copper wire
Line 24: remove we will use and replace with is typically usedLine 29: should be V = I x RLine 33:The operating voltage is different for each scale is a somewhat subjective statement. I use and regularly recommend 15V in N scale when many HO scale modelers use 14V or less. In Europe, almost all DCC systems default to 18V for all scales. Perhaps instead you could state: Refer to S-9.1 for recommended voltage ranges in various scales.
Line 37: The maximum recommended current ... varies between scales. (remove much)
Line 38: station limit of 3-5 ... for O and larger scales power...
Graphs: The graphs are blurry and difficult to read. Consider extending the bus length to 100 feet and also adding a duplicate x-axis scale for meters. For the 10A graph, extend the length to 200 feet, not uncommon in larger scales.
line 46: ...closer to the track section or the center...
line 50: ...to each rail is typically smaller than...
line 57: I don't agree with the 3ft / 1 meter recommendation. 6ft / 2 meter should be more than adequate in the case of soldered rail. ...of course if we provide a table of typical nickle silver rail resistance per meter at various codes, we will have the theory to confirm the statement.
Table 2.1 should be tied to booster current capacity, not scale.
line 81: ...should never loop back on itself, nor should...
line 83: ...reflections in the DCC...
line 94: It is no longer recommended to use...
line 84/85: Consider abbreviations C/MRI, LCC.
We should add a discussion about booster commons: what they are, why they are needed, and when they are and are not needed.
Theory
All of the specified recommendations should also have the theory from which they have been derived added. A good start is the addition of the resistance per length specs for wire gauges and rail codes and NA electrical code safe current carrying capacity for wire gauges. Table 2.1 seems somewhat arbitrary without this data.
Bus wire size and voltage drop is important because the booster must be able to fully drive its rated current for the overload protection to trip. Too much voltage drop (inline resistance) may limit the drive current of the booster in a short condition leading to excessive heat built up and damage to equipment.
Twisting of track bus pairs provides all of the following benefits:
Reduces inductance (reduces overshoot and ringing)
Increases capacitance (reduces overshoot and ringing)
Reduces susceptibility to common mode noise (interference)
The give the characteristics of the DCC signal (low frequency, high power), 1 and 2 are important, 3 is not so much, but is still worth mentioning
Twisting of track bus pairs does not reduce noise emissions. Running the two bus wires in parallel with very little separation does reduce noise emissions. The track bus pairs will naturally happen to be in parallel as a consequence of twisting.
The transmission line characteristics of a typical DCC bus are such that ringing/overshoot typically has a characteristic frequency above 200kHz. A 150ohm + 0.1uF capacitor has a corner frequency of about 10.6kHz. This seems a bit low to me. I need to look into this further. The capacitor needs to be rated for 50V or higher. If a surface mount ceramic capacitor is used, it should be a C0G/NP0 type so as not to cause an audible piezoelectric effect with the Bi-Directional cutout.
This document is a good idea, and overall I like what I see in this first draft.
Overview The first thing I would consider is: Who is the audience? I propose that there are two intended audiences:
I think near the beginning of the document the two different audience types could be acknowledged with some recommendations on how each should read the document:
In the body content of the document, there should be a clear separation between stated recommendations and the theory from which the recommendations are derived. The separation could be as simple as a new paragraph or as complex as a new sub-section at the discretion of the author.
The text body contains a number of second person statements. While these are not necessarily bad, using third person statements tend to give a more formal feel to the document.
Detailed Comments In the Terminology Table, consider the following changes:
Line 23: than the copper wire Line 24: remove we will use and replace with is typically used Line 29: should be V = I x R Line 33: The operating voltage is different for each scale is a somewhat subjective statement. I use and regularly recommend 15V in N scale when many HO scale modelers use 14V or less. In Europe, almost all DCC systems default to 18V for all scales. Perhaps instead you could state: Refer to S-9.1 for recommended voltage ranges in various scales. Line 37: The maximum recommended current ... varies between scales. (remove much) Line 38: station limit of 3-5 ... for O and larger scales power... Graphs: The graphs are blurry and difficult to read. Consider extending the bus length to 100 feet and also adding a duplicate x-axis scale for meters. For the 10A graph, extend the length to 200 feet, not uncommon in larger scales. line 46: ...closer to the track section or the center... line 50: ...to each rail is typically smaller than... line 57: I don't agree with the 3ft / 1 meter recommendation. 6ft / 2 meter should be more than adequate in the case of soldered rail. ...of course if we provide a table of typical nickle silver rail resistance per meter at various codes, we will have the theory to confirm the statement. Table 2.1 should be tied to booster current capacity, not scale. line 81: ...should never loop back on itself, nor should... line 83: ...reflections in the DCC... line 94: It is no longer recommended to use... line 84/85: Consider abbreviations C/MRI, LCC.
We should add a discussion about booster commons: what they are, why they are needed, and when they are and are not needed.
Theory All of the specified recommendations should also have the theory from which they have been derived added. A good start is the addition of the resistance per length specs for wire gauges and rail codes and NA electrical code safe current carrying capacity for wire gauges. Table 2.1 seems somewhat arbitrary without this data.
Bus wire size and voltage drop is important because the booster must be able to fully drive its rated current for the overload protection to trip. Too much voltage drop (inline resistance) may limit the drive current of the booster in a short condition leading to excessive heat built up and damage to equipment.
Twisting of track bus pairs provides all of the following benefits:
The give the characteristics of the DCC signal (low frequency, high power), 1 and 2 are important, 3 is not so much, but is still worth mentioning
Twisting of track bus pairs does not reduce noise emissions. Running the two bus wires in parallel with very little separation does reduce noise emissions. The track bus pairs will naturally happen to be in parallel as a consequence of twisting.
The transmission line characteristics of a typical DCC bus are such that ringing/overshoot typically has a characteristic frequency above 200kHz. A 150ohm + 0.1uF capacitor has a corner frequency of about 10.6kHz. This seems a bit low to me. I need to look into this further. The capacitor needs to be rated for 50V or higher. If a surface mount ceramic capacitor is used, it should be a C0G/NP0 type so as not to cause an audible piezoelectric effect with the Bi-Directional cutout.