Topics

DCC Bus distribution and snubbers question

Blair & Rasa
 

Hi
I'm planning my DCC layout for my Algoma Central at this time.  I have located three boosters, each at the root of the three peninsulas.  Each Booster will feed quad current limiters, either PM42s or something more recent.  However, each limiter's output will feed multiple bus segments(e.g. one bus on the upper peninsula, one on the lower, each fed by the same limiter).  As each of those runs will exceed 30', I am wondering if I need an RC snubber at the end of each, or should I put one at the output of the limiter, before we fan out to both runs?
In the grand schema, we have 34 sections of bus to be fed; in some cases a limiter will feed up to four segments of track, so if I have to put a snubber on each long run, the current drain will become significant, I think.  
Is it correct that I need to snub each long run, or is it effective to put a single snubber at the base of the branches?
Thanks
Blair

Paul O
 

Blair, I’m a believer in the saying “if it ain’t broke, don’t fix it.“
My suggestion would be to build the layout, test it and if you have a problem with the DCC signal, then look into installing snubbers where needed.

Paul O

Blair & Rasa
 

Paul
We already have sporadic loss of control at the far end of one run. Passes quarter test. Not enough experience yet, but we added a snubber to that run.
Blair

On 2019-09-20 22:22, Paul O wrote:
Blair, I’m a believer in the saying “if it ain’t broke, don’t fix it.“
My suggestion would be to build the layout, test it and if you have a problem with the DCC signal, then look into installing snubbers where needed.

Paul O




Ross Kudlick
 

Blair,

I have 3 thoughts:

"each limiter's output will feed multiple bus segments" my not be convenient for operations.  A short in one segment will stop trains in all other segments which will be frustrating to your train crews..  I suggest having each 'circuit breaker' power a contiguous section so a track short creates minimal train operation disruption.

"each of those runs will exceed 30' " --  Can you relocate your boosters to avoid exceeding 30' main bus runs?  A 60' length can be powered as 2 30' sections by lacating the booster with a main bus configured as a 'T."  You could power a 60' section on both levels (120' total) with a single booster without exceeding 30' main bus length.   This will mimiize (and possibly eliminate) any need for 'snubbers.'

Circuit breaker selection - consider using 'single' breakers rather than a 'quad' breaker.  Individual breakers can be located along the main bus creating a 'sub-bus' feeding  each section.  This mimimizes the amount of large gauge bus wire needed (and mimizes the 'spaghetti bowl' of wiring under the railroad).  I prefer the PSX breakers; others may offer other suggestions.  (Note - a PSX4 is actually 4 indiviual PSX breakers that can be snapped apart for individual installation.)

 Designing to 'best wiring practices' now will be benficial in the long run.  You may have to think 'outside the box' for the best solution for your railroad.

Regards,
Ross

Blair & Rasa
 

Ross
(minor responses interspersed below, then read on here:)
The layout room is 52 x 28; the layout is HO.  Attached is the overall DCC plan. The room is basically a double deck, with three locations where it becomes triple.  Some context: the oval at the bottom is a three tier staging room, 10x28; The three peninsulas are fed from their bases(points A, B, and C), as are the outside walls of the room.  Be aware, there are more subtleties, but this gives you the flavor.  We have built about 30% of the upper deck, as well as the branchline on the third level, upper left (i.e. busses C2 and C3 are in place and in use).  We have been operating with a single DCS100 located at 'C'.  The time has come to build the rest of the upper deck, which requires boosters at A and B, hence my questions.  Once built, A will be the location of the DCS100, B will be a DB210, and C will be a DB150.  Each will have a quad circuit breaker; yes, this could change, but the basic topology doesn't change if I add or remove a breaker or two.

In short,
- There's around 600' of main line, plus two 120' interchange RRs, and a 70' branchline. 
- the topology of the RR lends itself to the chosen booster locations.  There's just a lot of track in excess of 30' from the chosen locations.
- most bus runs go in two directions from the boosters already. 
- Most of those runs are > 30' each way.  All busses shown are planned to be 12 ga. twisted.
- everything installed to date passes the 'quarter' test, and runs fine, with the exception of two reported instances of loss of control at the far ends of busses C3a and C3b.  These could have been operator error (i.e. using a DT400, you can end up with the wrong knob active, and only think you're changing your locomotive's speed...)
- I'm trying to avoid more boosters, as for the operations I plan, more current is unnecessary based on loco and passenger count; two 5A boosters and an 8A booster are enough.  That will service the three yards, and the long mainline country runs.  If I eventually need more, it's likely that the staging yard room will be the reason, and it will be easy to separate it out and add a booster or two.

From what I've read, adding snubbers improves performance to the point where I'll be okay with the longer runs, but how many snubbers to add is the question.  I'm well aware that each one presents an added current load, though not a lot; most recommendations list a 1/2 watt 100 ohm resistor.  If we assume 1/4 watt dissipation, each resistor draws about 50 ma.  (P=I*I*R, R=100, P=0.25, therefore I=0.05).  If I had to install 34 snubbers, I'd be looking at 1.7 amperes just for them.

Sorry for the length of this, but I'm trying to give you all the info you might need, as you seem to want to help.
Thanks
Blair

On 2019-09-21 8:39, Ross Kudlick wrote:
Blair,

I have 3 thoughts:

"each limiter's output will feed multiple bus segments" my not be convenient for operations.  A short in one segment will stop trains in all other segments which will be frustrating to your train crews..  I suggest having each 'circuit breaker' power a contiguous section so a track short creates minimal train operation disruption.
Already sort of done.  See plan.

"each of those runs will exceed 30' " --  Can you relocate your boosters to avoid exceeding 30' main bus runs?  A 60' length can be powered as 2 30' sections by lacating the booster with a main bus configured as a 'T."  You could power a 60' section on both levels (120' total) with a single booster without exceeding 30' main bus length.   This will mimiize (and possibly eliminate) any need for 'snubbers.'
Already done. 

Circuit breaker selection - consider using 'single' breakers rather than a 'quad' breaker.  Individual breakers can be located along the main bus creating a 'sub-bus' feeding  each section.  This mimimizes the amount of large gauge bus wire needed (and mimizes the 'spaghetti bowl' of wiring under the railroad).  I prefer the PSX breakers; others may offer other suggestions.  (Note - a PSX4 is actually 4 indiviual PSX breakers that can be snapped apart for individual installation.)
Noted.  I have the PM42s, and intend to use them until they annoy me enough, or are inadequate, at which point PSX is likely where I'm going, unless there's a newer product available.  We'll see.

 Designing to 'best wiring practices' now will be benficial in the long run.  You may have to think 'outside the box' for the best solution for your railroad.
Yep.  Already trying to do that, that's why I asked before I proceeded.  Short of spreading boosters all over the place, I would have thought "best wiring practices" would include snubbers, where length dictates.  Not sure which "best wiring practice' you think I'm violating, so please elaborate.

But thinking outside the box requires that one actually know where the box boundaries are...
8-)

Regards,
Ross

Jerry Michels
 

This is good advice. We operate a large two-level laout, about 10 scale miles, and use no snubbers.  All works well regarding the DCC signal.

Jerry Michels
Amarillo Railroad Museum

Blair & Rasa
 

What is, Jerry? 
Also, what are your bus lengths, what's your scale?  These are pertinent inputs.
Thanks
Blair

On 2019-09-21 11:40, Jerry Michels wrote:
This is good advice. We operate a large two-level laout, about 10 scale miles, and use no snubbers.  All works well regarding the DCC signal.

Jerry Michels
Amarillo Railroad Museum

Blair & Rasa
 

A note on the shorthand used on the drawing might be in order.
A1a - indicates that the buss segment depicted is coming from Booster A, circuit breaker 1, and is the first segment so connected.  So B3d is on booster B, circuit breaker 3, and is the fourth wire segment attached to that breaker.
Blair

On 2019-09-21 11:19, Blair & Rasa wrote:
Ross
(minor responses interspersed below, then read on here:)
The layout room is 52 x 28; the layout is HO.  Attached is the overall DCC plan. The room is basically a double deck, with three locations where it becomes triple.  Some context: the oval at the bottom is a three tier staging room, 10x28; The three peninsulas are fed from their bases(points A, B, and C), as are the outside walls of the room.  Be aware, there are more subtleties, but this gives you the flavor.  We have built about 30% of the upper deck, as well as the branchline on the third level, upper left (i.e. busses C2 and C3 are in place and in use).  We have been operating with a single DCS100 located at 'C'.  The time has come to build the rest of the upper deck, which requires boosters at A and B, hence my questions.  Once built, A will be the location of the DCS100, B will be a DB210, and C will be a DB150.  Each will have a quad circuit breaker; yes, this could change, but the basic topology doesn't change if I add or remove a breaker or two.

In short,
- There's around 600' of main line, plus two 120' interchange RRs, and a 70' branchline. 
- the topology of the RR lends itself to the chosen booster locations.  There's just a lot of track in excess of 30' from the chosen locations.
- most bus runs go in two directions from the boosters already. 
- Most of those runs are > 30' each way.  All busses shown are planned to be 12 ga. twisted.
- everything installed to date passes the 'quarter' test, and runs fine, with the exception of two reported instances of loss of control at the far ends of busses C3a and C3b.  These could have been operator error (i.e. using a DT400, you can end up with the wrong knob active, and only think you're changing your locomotive's speed...)
- I'm trying to avoid more boosters, as for the operations I plan, more current is unnecessary based on loco and passenger count; two 5A boosters and an 8A booster are enough.  That will service the three yards, and the long mainline country runs.  If I eventually need more, it's likely that the staging yard room will be the reason, and it will be easy to separate it out and add a booster or two.

From what I've read, adding snubbers improves performance to the point where I'll be okay with the longer runs, but how many snubbers to add is the question.  I'm well aware that each one presents an added current load, though not a lot; most recommendations list a 1/2 watt 100 ohm resistor.  If we assume 1/4 watt dissipation, each resistor draws about 50 ma.  (P=I*I*R, R=100, P=0.25, therefore I=0.05).  If I had to install 34 snubbers, I'd be looking at 1.7 amperes just for them.

Sorry for the length of this, but I'm trying to give you all the info you might need, as you seem to want to help.
Thanks
Blair

On 2019-09-21 8:39, Ross Kudlick wrote:
Blair,

I have 3 thoughts:

"each limiter's output will feed multiple bus segments" my not be convenient for operations.  A short in one segment will stop trains in all other segments which will be frustrating to your train crews..  I suggest having each 'circuit breaker' power a contiguous section so a track short creates minimal train operation disruption.
Already sort of done.  See plan.

"each of those runs will exceed 30' " --  Can you relocate your boosters to avoid exceeding 30' main bus runs?  A 60' length can be powered as 2 30' sections by lacating the booster with a main bus configured as a 'T."  You could power a 60' section on both levels (120' total) with a single booster without exceeding 30' main bus length.   This will mimiize (and possibly eliminate) any need for 'snubbers.'
Already done. 

Circuit breaker selection - consider using 'single' breakers rather than a 'quad' breaker.  Individual breakers can be located along the main bus creating a 'sub-bus' feeding  each section.  This mimimizes the amount of large gauge bus wire needed (and mimizes the 'spaghetti bowl' of wiring under the railroad).  I prefer the PSX breakers; others may offer other suggestions.  (Note - a PSX4 is actually 4 indiviual PSX breakers that can be snapped apart for individual installation.)
Noted.  I have the PM42s, and intend to use them until they annoy me enough, or are inadequate, at which point PSX is likely where I'm going, unless there's a newer product available.  We'll see.

 Designing to 'best wiring practices' now will be benficial in the long run.  You may have to think 'outside the box' for the best solution for your railroad.
Yep.  Already trying to do that, that's why I asked before I proceeded.  Short of spreading boosters all over the place, I would have thought "best wiring practices" would include snubbers, where length dictates.  Not sure which "best wiring practice' you think I'm violating, so please elaborate.

But thinking outside the box requires that one actually know where the box boundaries are...
8-)

Regards,
Ross


dcesharkman
 

The snubbers are only anectoatal solutions based on many factors. Mostly parasitic reactance from using too large of a bus wire over an extended distance. The best advice is to use the smaller gauge with higher quality oxygen free copper. It is a little more expensive but with going to a smaller gauge, it is not quite so bad.

I am in N Scale and I have used 16 gauge wire for my DCC buss to over 50 feet with no signal degradation. My voltage drop over the distance was less than .5 volts. So it is more about using quality materials and you do not have to worry about snubbers and other wild solutions.

Respectfully,

Blair & Rasa
 

Well, so far, no answer from Ross.
From others, an unknown scale 10 miles in length reports no problems, and an N scaler tells me to reduce my gauge and use OFHC copper, and I won't have problems.

Four engine consists in HO meeting four engine consists in HO means I won't be reducing my primary bus gauge.

As for OFHC, point me at some objective measurements and reasoned arguments, please, because I'm not believing yet.

I reviewed my blocking, and rearranged a few to reduce the number of transitions between circuit breaker sections a bit. I also reviewed the mainline, and my loading estimates aren't changing. So I'll proceed with my plans and insert snubbers when I can prove electrical disturbance using an oscilloscope, which I fortunately have.

Ross, when you've had a chance, please get back to me with some feedback.
Thanks
Blair

Michael Rozeboom
 

On 2019-09-21 4:34 p.m., Blair & Rasa wrote:
Well, so far, no answer from Ross.
From others, an unknown scale 10 miles in length reports no problems, and an N scaler tells me to reduce my gauge and use OFHC copper, and I won't have problems.

Four engine consists in HO meeting four engine consists in HO means I won't be reducing my primary bus gauge.

As for OFHC, point me at some objective measurements and reasoned arguments, please, because I'm not believing yet.

I'll answer that: It makes no difference.  All copper wire is "oxygen free", because if there is too much oxygen, you can't draw it into wire as it is too brittle.

I regularly make measurements with an uncertainty of 20ppm.  Using regular wire, be it welding cable or 12AWG, even with off the shelf coax and instrument wire in the mix.  If using an exotic wire would make any difference, that would have been discovered long ago.

When making current measurements, I can magically make my secondary leads have zero impedance.  But the wire itself isn't magical.

The oxygen free idea is really designed to lighten the wallets of silly audiophools, who believe that it will improve their experience.  They also believe that 4AWG welding cable is better and will improve their audio.




--




Michael Rozeboom


Electrical Metrology

Blair & Rasa
 

Thanks, Michael.  I have seen exactly one instance in my past electronics/physics life where OFHC demonstrably made a difference to something, and that was in a 600 keV proton linac fed with a 500 kW 267 MHz RF source.    Not quite what I am building in my basement.  I agree, there's lots of "alternative fact" on the internet.

Blair


On 2019-09-21 17:07, Michael Rozeboom wrote:
On 2019-09-21 4:34 p.m., Blair & Rasa wrote:
Well, so far, no answer from Ross.
From others, an unknown scale 10 miles in length reports no problems, and an N scaler tells me to reduce my gauge and use OFHC copper, and I won't have problems.

Four engine consists in HO meeting four engine consists in HO means I won't be reducing my primary bus gauge.

As for OFHC, point me at some objective measurements and reasoned arguments, please, because I'm not believing yet.

I'll answer that: It makes no difference.  All copper wire is "oxygen free", because if there is too much oxygen, you can't draw it into wire as it is too brittle.

I regularly make measurements with an uncertainty of 20ppm.  Using regular wire, be it welding cable or 12AWG, even with off the shelf coax and instrument wire in the mix.  If using an exotic wire would make any difference, that would have been discovered long ago.

When making current measurements, I can magically make my secondary leads have zero impedance.  But the wire itself isn't magical.

The oxygen free idea is really designed to lighten the wallets of silly audiophools, who believe that it will improve their experience.  They also believe that 4AWG welding cable is better and will improve their audio.




--




Michael Rozeboom


Electrical Metrology

Dale Gloer
 

I had success using a snubber on one occasion.  When I started building my layout I had a DCC bus length approaching 120 feet in length.  The bus was 14 gauge solid copper wire (stripped out of Loomex).  Near the end of the bus I had control problem with some decoders.  Installing a snubber at the end solved the control problems but the permanent fix was to divide the bus up and provide another booster in the middle of two sections of the bus.

Dale Gloer

Blair & Rasa
 

Dale,
If the snubber fixed it, why then add the booster?  Were you also facing a load problem (i.e. too many locos/lighted cars/accessories)?

On 2019-09-22 10:15, Dale Gloer wrote:
I had success using a snubber on one occasion.  When I started building my layout I had a DCC bus length approaching 120 feet in length.  The bus was 14 gauge solid copper wire (stripped out of Loomex).  Near the end of the bus I had control problem with some decoders.  Installing a snubber at the end solved the control problems but the permanent fix was to divide the bus up and provide another booster in the middle of two sections of the bus.

Dale Gloer