[fabben]

I just read Richard Johnson's post on the "Capacitors - Do we need them thread?".

I have a power bus ring and I see that Richard recommended that this be cut to form a "Tee" and that both ends be terminated with a capacitor and resistor, in series. I took a look at Richard's web site and he recommends the same termination but the value of the capacitor is different.

So I suppose my question is addressed to Richard, should I use a 100 ohm/01mfd (as in the thread) or use a 100 ohm/0.01mfd (as on the web site)?

Is the termination recommended for all protocols, ie LocoNet, XpressNET, etc? I've got a Hornby Elite.

Also (I've just discovered protocols) so can decoders for say LocoNet be used on XpressNET layout?

I'm finding that DCC is quite a voyage of discovery, the more I learn, the more there is to learn.
I'm quite fascinated by it all really.

Fabben

 

[Makemineadouble]

QUOTE Also (I've just discovered protocols) so can decoders for say LocoNet be used on XpressNET layout?

yes no problems at all just fit um in and let um rip

 

[Richard Johnson]

QUOTE (fabben @ 19 Dec 2007, 23:38) <{POST_SNAPBACK}>I just read Richard Johnson's post on the "Capacitors - Do we need them thread?".

I have a power bus ring and I see that Richard recommended that this be cut to form a "Tee" and that both ends be terminated with a capacitor and resistor, in series. I took a look at Richard's web site and he recommends the same termination but the value of the capacitor is different.

So I suppose my question is addressed to Richard, should I use a 100 ohm/01mfd (as in the thread) or use a 100 ohm/0.01mfd (as on the web site)?

Is the termination recommended for all protocols, ie LocoNet, XpressNET, etc? I've got a Hornby Elite.

Also (I've just discovered protocols) so can decoders for say LocoNet be used on XpressNET layout?

I'm finding that DCC is quite a voyage of discovery, the more I learn, the more there is to learn.
I'm quite fascinated by it all really.

Fabben

Hi Fabben

The difference is a typo - missed the "dot" on the post

Use a 100 ohm 1/2 watt resistor and a 0`.1microFarad ceramic capacitor (sometimes also called a monolithic capacitor)

Regards

Richard
DCCconcepts

 

[Sol]

OK, this may have been posted before or at least some answers but in trying to search for it, I find getting lost reading 6 pages of replies just on one topic.
As Richard J knows, I have now taken the change to DCC & thus madly reading.

I was under the impression that the Cab/Net wires for the controllers are not looped & has a terminating network on the end of the run but was assumed that the power bus from which all the tracks are connected to, could be in a loop.

Now reading the above posts, etc, I am now confused.

The Power Bus wires - are they looped or stop & be terminated like the Cab/Net controller wires?

 

[Sol]

Further to the above question which after much searching of the Net, indicates that looping of the Power Bus is not recommended & that the ends are terminated, I also see that it is recommended that the power bus be a twisted pair but again I was under the impression that the two bus wires are run so they are separate by about 3" or more. In fact I have seen DCC layouts that have the power bus not twisted but separated. Any "hard & fast" reasons for either please?

 

[Richard Johnson]

QUOTE (Sol @ 20 Dec 2007, 14:31) <{POST_SNAPBACK}>Further to the above question which after much searching of the Net, indicates that looping of the Power Bus is not recommended & that the ends are terminated, I also see that it is recommended that the power bus be a twisted pair but again I was under the impression that the two bus wires are run so they are separate by about 3" or more. In fact I have seen DCC layouts that have the power bus not twisted but separated. Any "hard & fast" reasons for either please?

***Sol

You need to read my website again, there's nothing to be confused about.

Its not necessary to even think about terminating the control bus unless you are approaching the bus design limits - in the case of your system that is in excess of 300 feet, and if it gets that long adding power to it (12v regulated DC wall-warts) also becomes recommended. Your systems bus is like many actually RS485 based - very forgiving and stable.

The DCC power bus carries a non symmetrical high power square wave. All forms of alternating current have a field around the wire. If two wires are close together, the fields inter-react and crate inductance, which can affect the signal quality and also translates to impedance and therefore voltage loss.

Its therefore better to twist a bus made from twin cable several turns per foot. The twisting cancels the problem so to speak. The option as you noted is to run the two bus wires at least several inches apart. either is just fine.

Both the above are relatively not important on a small layout, with importance and benefits of doing it right growing with layout size.

On large installations just having two wires alongside each other has an effect.... for example on my own very large layout the static impedance of the bus wiring with absolutely nothing connected to it is 1500 ohms. If it wasn't twisted the cost of this artificially generated impedance would be a loss of about 15% of voltage under load, so it IS a factor as size grows. (and my cable is a mixture of huge 10 and 12 gauge pure oxygen free copper!)

Looping... Its not recommended to loop either the power bus or the control bus. Some argue that the track os a loop anyway but it is not - it has breaks at every trunout and branching..... The world won't stop if the power bus is a ring but again, the larger the layout the greater the reason not to do it.

The termination on the power bus helps clean up the waveform but more importantly clamps and suppresses sharp voltage peaks that are generated by momentary short circuits. These peaks are about 3x track voltage so the suppression is really helpful in keeping things running reliably by virtue of power stability and not stressing decoder components (momentary shorts are the most common cause of decoder death in a properly installed loco.

Richard

DCCconcepts

 

[Doug]

As Richard says, these issues really only affect you if you have a massive layout. In that case you are better to divide it up into power districts.

My previous table-top layout had a ring power BUS and I didn't have any problems. The DCC BUS was twisted and was about 12 meters long overall.

The new layout has a 25 meter T-structured DCC BUS that is not twisted or terminated. Works fine with no loss of volts or signal quality along the wire. I am using my 6mm² chunky wire.

 

[Sol]

Thanks for that information Richard & Doug.
I now know that as I alter my partially started layout with an existing common return which becomes one of the power bus runs , I do not have to twist another wire around it but will keep them apart but not looped & provide a termination.

 

[Richard Johnson]

QUOTE (Doug @ 20 Dec 2007, 16:52) <{POST_SNAPBACK}>As Richard says, these issues really only affect you if you have a massive layout. In that case you are better to divide it up into power districts.

My previous table-top layout had a ring power BUS and I didn't have any problems. The DCC BUS was twisted and was about 12 meters long overall.

The new layout has a 25 meter T-structured DCC BUS that is not twisted or terminated. Works fine with no loss of volts or signal quality along the wire. I am using my 6mm² chunky wire.

***Not so much massive as big - and they CAN affect smaller layouts if the wiring isn't best practice, so its worth following the same rules.

The arrangement sounds good. T bus is ideal and wire size is perfect.... and thats only 12.5 metres per arm of the tee so I'd say "medium" sized in global average terms.

Add the terminators though Doug - the suppression of those spikes after any minor back to back or similar short will pay dividends in live decoders long term, and the parts are so cheap its almost free.

Richard
DCCconcepts