On a 37' boat with a system as diagrammed in a previous thread in this Forum, would it seem acceptable for the trunk (backbone) to be micro or would there be a need to increase to mid?
Second, the N2K spec says no drop shall be more than 20', but it appears Airmar sells a cable to connect its PB200 to an N2K backbone, and that cable (for running in the mast) is sometimes 100'. Any problem running that drop cable 50' down my mast to the backbone?
Thanks,
Jon Kinne
Jon, the idea of the PB200 cable longer than 20' is that it becomes an end section of your backbone, not a drop. That's why Airmar builds a backbone terminator right into the PB200, which can be turned on or off by means of a removable pin in the cable assembly. All the N2K wind sensors can use some similar scheme, if just an inline backbone terminator.
Micro cable is fine, as long as you don't need too much 12v power too far from the source. I highly recommend that you read Maretron's NMEA 2000 installation guide, or similar works from other manufacturers. You need to understand the voltage drop issues, and apply the formula based on loads and backbone/drop lengths.
I have to assume that Airmar sizes the cable which connects the PB200 to the backbone to account for the PB200's current requirements, so the issue is the rest of the backbone. Even if I don't have a problem with voltage drop along the rest of the backbone, it seems to make sense to run "mid" cable, and take advantage of the increased conductor size, since it's no more trouble, access is fine and cost isn't that different.
I have the Maretron document you mentioned, and that's where I found the "20-foot drop length maximum". Thanks for clarifying that the PB200 in this instance is not at the end of a drop but at the end of the backbone.
On my boat (an Island Packet 38) I went through a few different network designs. The most limiting factor being that drop length, and the preference for single devices at the end of a T (no daisy chains).
I ended up resizing the network backbone to "MINI" as I thought it allowed the most forward looking capability. I installed the MINI/Micro drop T's back to back on a mounting board and ran my initial instruments all on a single branch out from the Power Dist. (8 drops in all, so far.)
I ended up finding out that a few things are fungable:
The wire run from my helm console to the backbone T is nearly 30', with no problems, even though it's created from an AT-10 going through Simnet cable, including a connector in the middle joining 2 lengths of cable, and 3 IS20's on the end. Clearly not a good idea given the wiring guidelines, but it works fine.
Regarding the trunk line up the mast: Much to my chagrin, when my friend went up the mast to install the Maretron wind sensor, and my repeated warning/instruction, he failed to install the inline resistor, and didn't mention the fact until a week later. It's worked fine from day 1, with no ill effects that I can tell at all..
I still plan on installing the resistor the next time up the mast however.
hope this helps,
Jon, Airmar has no control of the amount of voltage arriving at its cable. That has to do with the length of the backbone and number of other devices drawing juice off it. As mrfugu has found out, the NMEA 2000 spec is conservative and robust, and you can sometimes get away with liberties, but it's not pretty when it doesn't work right.
Ben,
right, I don't have the calculations in front of me, but the major change in thinking and network topology I made was making the backbone by only using MINI T's and the one length of mid cable for the mast in contrast to stretching a backbone from fore to aft, dropping a T at every sensor location.
TTTTT vs T-----T------T-----T--T
it made both the math and actual installation look (and I presume perform) better.
good luck,
Mrfugu, if by "calculations" you mean the simple voltage drop formula listed in most everyone's N2K manual, please understand that it is a simplification of the very complex calculation needed to truly analyze a network. I think it just separates 'no problem' configurations (most, currently) from ones that need a full bore analysis, or higher capacity cabling.
So I'm not sure your config performs better (it just may fit better in the simplified formula), but there's certainly nothing wrong with it. An N2K network does not have to be as linear as many tee and drop diagrams suggest. Check out the Furuno breakout box, for instance; it's possible to have the whole backbone, with terminators, in the box:
https://www.panbo.com/archives/2008/09/actisense_furuno_n2k_junction_boxes_impressions.html