Trains.com

Subscriber & Member Login

Login, or register today to interact in our online community, comment on articles, receive our newsletter, manage your account online and more!

Snubber purpose

7049 views
11 replies
1 rating 2 rating 3 rating 4 rating 5 rating
  • Member since
    September 2002
  • 7,486 posts
Snubber purpose
Posted by ndbprr on Wednesday, April 6, 2016 1:46 PM

NCE say's on long blocks to add a snubber consisting of a resister and capacitor across the bus wires. What is the purpose and where in the block should it be placed?

  • Member since
    July 2006
  • From: Pittsburgh, PA
  • 1,796 posts
Posted by JoeinPA on Wednesday, April 6, 2016 2:13 PM

Here's a short explanation: http://www.members.optusnet.com.au/nswmn2/DCC.htm#Terminators

Note that they call them Terminators.

Joe

  • Member since
    July 2009
  • From: lavale, md
  • 4,678 posts
Posted by gregc on Wednesday, April 6, 2016 4:59 PM

ndbprr
What is the purpose and where in the block should it be placed?

unterminated transmission lines reflect signals.   This is normally only a problem for high frequency signals where the signals along the transmission lines can be cancelled due to the reflection at half and multiples of the wavelength of the signal (10 MHz ~ 30m).   But DCC and cab-bus are ~10kHz.

a transmission line terminator would be added at the end of the transmission line (where an antenna might be).   It would provide a specific impedance to minimize the reflection.

NCE's Mark Gurries' Snubber/RC-Filter  page talks about an RC-Filter and spiky noise most likely from motors or locomotives momentarily loosing contact with the track over dirt spots or turnouts.

spiky noise has high frequency components and can interfere with DCC signals.   It sounds like the resistor-capacitor filter suppresses the noise spikes from being reflected.

placing the resistor-capacitor filter in the middle of the line would suppress the signal on the side opposite the source (DCC system), so it has to be at the end of the line.

greg - Philadelphia & Reading / Reading

  • Member since
    October 2005
  • 1,047 posts
Posted by betamax on Wednesday, April 6, 2016 6:49 PM

The idea is that it provides a path for the signal when the phases switch. With a long bus, and a lot of inductance, the charge/discharge cycle can distort the waveform when that rail is held to "ground" and the other one is energized. It may produce a spike.

The terminator provides a path for that energy to be dissipated.

https://www.dccwiki.com/Bus_Termination

Note that you should only install them if you are having signal problems, and if you can, verify the integrity of the waveform with an oscilloscope first. Placing the bus wires close together will reduce the inductance as well.

  • Member since
    August 2013
  • From: Richmond, VA
  • 1,890 posts
Posted by carl425 on Thursday, April 7, 2016 8:49 AM

betamax
verify the integrity of the waveform with an oscilloscope first.

Big Smile I'd bet that anyone that has an oscilloscope and/or knows how to use one would already know how to deal with the snubber question.

betamax
Note that you should only install them if you are having signal problems

What is the reason for this caveat?  I've always heard that they won't always help but they'll never hurt.  Have I been mislead? (it's happened before)

I have the right to remain silent.  By posting here I have given up that right and accept that anything I say can and will be used as evidence to critique me.

  • Member since
    February 2002
  • From: Reading, PA
  • 30,002 posts
Posted by rrinker on Thursday, April 7, 2016 1:49 PM

 Now that I have a scope I'm curious to check it out. Except now I don't have a layout.

 The amount of ringing you get at the end of a transmission line also depends on the way the signal is sourced. The output drive section of different brands of boosters is different. NCE always recommends snubbers, Digitrax does not. We do not use them on our club layout, which when assembled fully is cirrently somewhere around 28x160 (giant donut - so there is some 376 linear feet of railroad, not counting that the ends are curved, and that there are at least 2 tracks everywhere). And there are no signal problems, at least none that cause any issues with anyone's locos, even at the very ends of each booster district.

                          --Randy

 


Modeling the Reading Railroad in the 1950's

 

Visit my web site at www.readingeastpenn.com for construction updates, DCC Info, and more.

  • Member since
    July 2009
  • From: lavale, md
  • 4,678 posts
Posted by gregc on Thursday, April 7, 2016 3:02 PM

I agree with carl425 - why are they only needed if there are problems?

do snubbers fix the problem or only help in some cases?   It would be nice to see a scope trace of a signal that causes erratic behavior and the affect the snubber has on the signal, presumably correcting the problem.

I was surprised that my DCC loco started running when I was testing a DC loco with a PWM throttle.    The DCC decoder was recognizing its address and commands to run forward.   Looks like validation of a properly encoded packet of information is very week, making DCC susceptible to noise.

My guess is that longer length DCC busses are noisier and can cause the problem I describe above because DCC decoders can easily recognize noise as commands.

 

greg - Philadelphia & Reading / Reading

  • Member since
    October 2005
  • 1,047 posts
Posted by betamax on Thursday, April 7, 2016 4:59 PM
Your DCC locomotive was not recognizing the commands, it was switching to analog mode and operating as if it were a DC locomotive. A very distorted waveform can cause that to happen, which is why many modellers switch the analog mode off (DCC Only.) Prevents runaways.
  • Member since
    July 2009
  • From: lavale, md
  • 4,678 posts
Posted by gregc on Thursday, April 7, 2016 7:47 PM

betamax
it was switching to analog mode

that makes more sense.

I'm surprised that the S-9.2.4 section B says that a decoder can switch modes (i.e. DC) when a DCC signal is not detected for 30 ms.   This seems like a very short time for switch modes.

betamax
A very distorted waveform can cause that to happen

what causes this distortion and does a snubber correct it?

greg - Philadelphia & Reading / Reading

  • Member since
    October 2005
  • 1,047 posts
Posted by betamax on Thursday, April 7, 2016 8:01 PM

Thirty milliseconds isn't a long time when a DCC signal is in the range of 10 kHz. That time is less than two cycles at 60 Hz.

What causes distortion?  A lot of things. The design of the drivers, the impedance of the bus, electrical noise.

Since the DCC waveform is loaded with harmonics, the impedance can have a definite effect.

  • Member since
    March 2015
  • 1,358 posts
Posted by SouthPenn on Friday, April 8, 2016 5:38 AM

Wouldn't the snubber reduce the spike on the rails when shutting the power off? There could be a lot of capacitors discharging if you have sound in your locomotives. I would think this spike could cause decoders to get stupid for no apparent reason.

 

 

South Penn
  • Member since
    February 2002
  • From: Reading, PA
  • 30,002 posts
Posted by rrinker on Friday, April 8, 2016 7:10 AM

 Any keep-alive caps are downstream of the full wave bridge on the decoder, they don't suddenly discharge their power back into the rails. The DCC standards and RPs have specifications for noise immunity and signal characteristics in them, but some decoders seem more susceptible to a noisy signal than others. I have yet to see any decoder spontaneously scramble its programming, either on my home layouts or during week long club displays. That includes original BLI with QSI decoders on up to Paragon2, as well as Tsunami and Loksounds from 3.5 to 4.0 and Select. I've never had any of my locos run away when I power up the system, but I do turn off analog AND I've turned off DC operation in my command station.

 I don't twist my bus wires, either. But they aren't clamped in a perfectly parallel arrangement, either. Both wires run together through the same hooks and wire ties holding them in place.

                          --Randy

 


Modeling the Reading Railroad in the 1950's

 

Visit my web site at www.readingeastpenn.com for construction updates, DCC Info, and more.

Subscriber & Member Login

Login, or register today to interact in our online community, comment on articles, receive our newsletter, manage your account online and more!

Users Online

There are no community member online

Search the Community

ADVERTISEMENT
ADVERTISEMENT
ADVERTISEMENT
Model Railroader Newsletter See all
Sign up for our FREE e-newsletter and get model railroad news in your inbox!