Author Topic: Best Of Techniques for Sound in N-scale Diesels - with Index (Updated 3/12/2019)  (Read 11396 times)

0 Members and 1 Guest are viewing this topic.

jdcolombo

  • Crew
  • *
  • Posts: 1684
  • Respect: +516
[Added by the mods: see the end of this post for an index to the various sound install threads that John has posted. -gfh]

Hi everyone.

After now doing over a dozen sound installations in N scale diesels (including Atlas GP7/9, RS3, RS11, SD9, GP30, VO-1000, and a FVM GEVO), I've refined my techniques a bit.  Here's my current thinking on some things that differ slightly from what I've described in my various installation threads.

1.  I no longer think drilling holes in the shell for sound egress is necessary.   As an experiment, I didn't drill any holes in my GP30's fan detail, and it sounded fine.  So when I did the RS-3's, I didn't drill any holes in them either, and they sounded great, too.  My theory here is that there is plenty of air leakage between the shell and frame, so no holes are really necessary.  YMMV, as they say, but if you are doing an installation yourself, you might skip this step and see how things sound to you.  You can always pop off the shell and drill holes later.

2.  Speakers: UPDATED (January 4, 2017)  The speaker market has changed a lot since my first installations.  The cell-phone type speakers available today are made by Soberton, PUI, and CUI.  Of these, only Soberton makes an 8x12mm; all of them make the larger (9x16, 11x15, 13x18) sizes.  The Soberton speakers are available from Digikey; the CUI and PUI speakers are available from both Digikey and Mouser.  John LeMerise has designed several different 3D-printed custom speaker enclosures for different speakers and different applications that are listed at the end of this post.

The key to best performance from any speaker is (1) it MUST be mounted in its own, sealed enclosure (do NOT try to use a tender shell or an engine shell as a baffle; it just doesn't work) and (2) the internal volume of the speaker enclosure must be large enough to allow the air trapped inside to act as a dampening "spring" for the speaker diaphragm.  My own experiments indicate that a minimum enclosure volume of 500 cu.mm is necessary for an 8x12mm speaker; 750 for a 9x16 or 11x15; and 1000 for a 13x18.  More is nearly always better.  Also remember that the speaker enclosure does not need to be a plain box.  You can "offset" the speaker in a larger box; make an "L" shaped box where one end is thinner than the other; or anything else that makes sense.  The shape is irrelevant; the key is to make sure the enclosure is fully sealed and of sufficient internal volume.

3.  Keep Alives.  I haven't noticed sound drop-out problems in my "road" diesels - units that I use in a three-unit consist to run trains over the mainline from a staging yard to the main classification yard and then back to the staging yard.  But I did notice  a few dropouts when trying to use those units for slow-speed switching.  So I've started adding a single 220uf "stay alive" capacitor to my installations (again, this is easier in short 4-axle diesels if you use the 8x12mm speaker) in any case where I think I may use the unit for switching duties.  Since most of us find that our operations evolve over time, I'd probably now recommend a stay-alive cap if at all possible in the space available.  My recommended capacitor for these purposes is a 220uf, 16v Tantalum "chip" (surface mount) capacitor.  These are usually available in a package that is about 7.3mm long, 4.3mm wide, and 3-4mm high.  I try to get ones that are 3mm high, so that I have maximum options in mounting them in confined spaces (for example, glued to the front of the speaker, as in my RS3 installation).  A 16v cap should play well with any DCC system that puts out a reasonable voltage for N or HO scale trains; but if you've decided to set up your DCC system to crank out 20v for O scale, you're in trouble.  I use a Digitrax Chief system set for N scale output; you can change this via a switch to HO or O, which ups the voltage to the track.  I've measure the voltage in HO mode at 14.7v; in "O" mode, it jumps to about 18v.  18v will cause your cap to explode in a fiery, shell-melting display (I had one do this, because the cap was defective.  It was spectacular; fortunately, I had a replacement shell, and the underlying decoder wasn't harmed; don't expect to be so lucky).  So before you add a 16v stay-alive cap, you might want to measure the output voltage of your DCC system . . .

UPDATE (3/1/2019).  Today I think most of us doing sound installs agree that keep alive caps are a necessity, and the more the merrier.  I now use two 220uf caps unless absolutely impossible to do so (e.g., the Atlas RS3), and in steam installations, I've used as many as a half-dozen.  In larger steam tenders, I've even gone to using a TCS KA-3, which will keep an N scale steamer moving with the sound on for several seconds.  We also have migrated to using tantalum polymer caps, instead of the standard tantalum caps, since these have better voltage tolerance and are less likely to explode in an over-voltage situation.  The available values and packages vary, but they are available from Mouser and Digikey.

4.  And for the decoder manufacturers (ESU; Soundtraxx; Digitrax; TCS; QSI; etc) - a sound decoder that came in a slightly narrower package than the ESU LokSound Select Micro would be a big help.  9.5mm wide by 25mm long would be perfect for nearly anything; 10mm wide would be fine for 95% of the potential uses, and much easier to deal with than ESU's 10.6mm width.  Rumors are flying that Soundtraxx is at work on successor to its current Tsunami line and of course ESU is doing the sound installation for Atlas' S-2 switcher, which has to have a smaller PC board footprint than the current Select Micro.  I think there is a market for N scale sound in diesels that isn't being tapped because doing it well is still pretty intimidating.   The decoder manufacturers and diesel manufacturers could make this easier by working together, or at least considering individually the engineering required for sound installations.  We are now at the point where all diesels provide some sort of "drop in" DCC option; how about thinking along the lines of a "drop in" sound option?  For example, Atlas, how hard would it be to design new frames that have a "notch" at the rear or simply a lower shelf for someone to add a speaker?  Then one could imagine a decoder manufacturer designing a speaker/enclosure that would drop into that notch, and a sound decoder that would drop into the space reserved for a regular DCC decoder.  Yes, you'd still have to do some work to add LED's for lights, but this would make things a whole lot easier.

UPDATE (2/28/2019).  The arrival of the ESU board-style decoders (73100, 73199) and the Zimo MX660 have answered the desire for a narrower decoder.  And Atlas has started shipping "sound ready" locomotives with a speaker already installed.  So maybe the manufacturers DO listen . . .

John C.


Sound Installation Threads:

RS-11 Tutorial
RS-11 initial thread, just the completed installation YouTube video
VO-1000 tutorial
GP7, not a full tutorial
SD9, not a full tutorial
GP30, not a full tutorial
RS3 Tutorial
Atlas C420 Tutorial
FVM GEVO Tutorial
Loksound in a Bachmann Berkshire Tutorial (not a diesel, but still . . . )
Loksound in a Bachmann 2-8-0 Consolidation
Loksound in a BLI PA-1
Loksound in a Kato PA-1
Loksound in a LifeLike FA
Speaker swap in an IM SD40-2
Zimo MX660 in a Kato NW2
Loksound in a Walthers 0-8-0

Since I posted this original message, many other folks have done additional LokSound installations in a lot of different engines.  I thought it would be useful to gather them in this post as well, so here goes.

Rick Brodzinsky's installations:
Loksound in a Kato F3
Loksound in a Kato SD70ACe
Loksound in a Kato MP36PH
Loksound in a Kato F40PH
Loksound in a Kato E8 with marker lights
Loksound in a Kato P42 with 6 function outputs
Loksound in a Kato SD45 with ditch lights
Loksound in a Kato Mikado

AKNscale's installations:
Loksound in a Kato C30=7
Loksound in an Atlas B30-7

Bill Denton's installations:
Loksound 73100 board in Atlas GP9
Loksound in IM F7A

Mark W.'s installations:
Kato F40PH
Loksound Board installs in Kato SD70ACe and ES44AC, and a FVM ES44AC 

SteveRuger45's installations:
Loksound 73100 board installation in FVM GP60
Loksound 73100 board installation in non-DCC-ready GP60
Loksound 73100 board installation in Kato Dash 9
Loksound in Atlas SD50
Loksound in LifeLike GP60
Loksound in Kato C30-7
Loksound in Atlas B23-7

Jim Starbuck's Installations:
Loksound in Arnold SW1
Loksound installation and re-motoring a LifeLike SW9
Loksound 73100 in an Atlas MP15DC

Another Atlas MP15 install using a wired 73800 by Chris Broughton: https://www.therailwire.net/forum/index.php?topic=46662.msg611205#msg611205

Kato SD9 LokSound installation by Carl Sowell: https://www.therailwire.net/forum/index.php?topic=39805.0

Kato U30C LokSound install by John F.: https://www.therailwire.net/forum/index.php?topic=35661.0

Kato Dash 9 installation by GM50 4164 using 73100 board: https://www.therailwire.net/forum/index.php?topic=45250.0

LokSound Select Micro 73100 board installation in IM SD45T-2 by Tehachapifan: https://www.therailwire.net/forum/index.php?topic=43396.0


Threads on generally applicable techniques:
Peteski's dissection of the LokSound Select Direct Micro Boards (73100 and 73199) for installing keep-alive: https://www.therailwire.net/forum/index.php?topic=44324.0

Using a choke along with keep alive caps to avoid LokProgrammer programming errors: https://www.therailwire.net/forum/index.php?topic=45354.msg587453#msg587453

Making a speaker enclosure

Making a small PC board with surface-mount resistors for LED lighting: https://www.therailwire.net/forum/index.php?topic=41340.msg515416#msg515416


Keystone Customs custom speaker enclosures:
For 8x12mm Soberton, 6mm depth, general application: http://www.shapeways.com/product/5L4ZXWA28/8mm-x-12mm-soberton-speaker-enclosure-4pk?optionId=60853776

For 8x12mm Soberton, 4mm depth but extended length, offset, for general application: https://www.shapeways.com/product/HRECPQVSB/9-62x16x4mm-enclosure-for-8x12-soberton-speaker?optionId=66848954

Offset enclosure for using 8x12mm speaker in an IM SD40-2: https://www.shapeways.com/product/N8M7HBL6Y/2pk-8x12-soberton-offset-speaker-enclosure?optionId=62934409

Offset L-shaped enclosure for 9x16mm speaker in an Atlas GP38/40 with LokSound board-style decoder: https://www.shapeways.com/product/47AN33V3Y/gp-38-40-9x16-speaker-enclosure-2pk?optionId=85016610

Offset enclosure for using a 9x16mm speaker in an IM SD40-2: https://www.shapeways.com/product/W64FT84NN/16x9-spkr-enclos-2pk?optionId=62916866


« Last Edit: March 12, 2019, 10:23:30 AM by jdcolombo »

atsf_arizona

  • Crew
  • *
  • Posts: 693
  • Respect: +31
    • My PBase Photo album where my Model RR pictures are
Re: Updated Techniques for Sound in N-scale Diesels
« Reply #1 on: February 04, 2014, 10:33:29 AM »
0
John, thx for the great summary info and your continuing inspiration and experience in N scale sound for all of us to see.  :-)   Really appreciate what you're doing for us in sharing this info.
John Sing
Sarasota, FL
http://pbase.com/atsf_arizona
https://web.archive.org/web/20151002184727/home.comcast.net/~j.sing/
========
Modeling the Santa Fe's Peavine Line (Ash Fork -> Phoenix, Arizona) during the 50s and 60s

sp org div

  • Crew
  • *
  • Posts: 190
  • Respect: +21
Re: Updated Techniques for Sound in N-scale Diesels
« Reply #2 on: February 04, 2014, 11:26:15 AM »
0
Ya, thanks for putting that up.
The N scale sound market must be a lot worse than I think it is, because it seems like advancements have stalled for years...
I havent added any more sound units after installing multiples of Digitrax sound drop ins... and most of them required some mods (and with only fair sound files).
Im burned out on all the hardwire decoder installs that have taken place, and would jump on more user friendly designs from Tsunami or Loksound. 
Maybe when the economy starts rolling a bit more?

Jeff
http://espeeoregondivision.blogspot.com/

peteski

  • Crew
  • *
  • Posts: 19461
  • Gender: Male
  • Respect: +1615
    • Coming (not so) soon...
Re: Updated Techniques for Sound in N-scale Diesels
« Reply #3 on: February 04, 2014, 03:21:36 PM »
0
Nice summary John.

I also want to add my warning about voltage ratings for tantalum caps.  I prefer to err on the side of caution, and use 20V or 25V rated caps. For me 16V is too close for comfort in N scale DCC applications.

As John mentioned (and experienced), tantalum caps have spectacularly violent failure mode.  Unlike LEDs or other semiconductor devices which simply let the smoke out when they fail, tantalum caps are more like fireworks!  Back in the day when I used to troubleshoot and repair comuter circuit boards, I had a tantalum cap actually fly off the board like rocket, missing my head by few inches.  It happened rather quickly. I saw a brief bright glow on the board I was repairing and next thing I knew was that the red-hot cap body took off the board and flew past my head.  I suspect that this cap was installed backwards (it was so burned that it was impossible to confirm that). But that incident made me treat tantalum caps with respect. Electrolytic caps can also fail, but they usually go with a small pop as the safety vent simply relieves the built up pressure.

Of course the tradeoff is that caps rated at higher voltages are physically larger than equivalent value caps rated at a lower voltage. In N scale space is at a premium, so this makes a difference. But still, I prefer caps rated at 20V.

As far as advancement goes, for me the next big step will be when (if ever) QSI will release a N scale version of their Titan sound decoder. I just love QSI sound decoders (for steam locos).
--- Peteski de Snarkski
--- Honorary Resident Curmudgeon

garethashenden

  • Crew
  • *
  • Posts: 1088
  • Respect: +259
Re: Updated Techniques for Sound in N-scale Diesels
« Reply #4 on: February 04, 2014, 03:47:00 PM »
0
Ya, thanks for putting that up.
The N scale sound market must be a lot worse than I think it is, because it seems like advancements have stalled for years...
I havent added any more sound units after installing multiples of Digitrax sound drop ins... and most of them required some mods (and with only fair sound files).
Im burned out on all the hardwire decoder installs that have taken place, and would jump on more user friendly designs from Tsunami or Loksound. 
Maybe when the economy starts rolling a bit more?

Jeff
http://espeeoregondivision.blogspot.com/

I suspect the S-2 may jump-start things a bit when it arrives. I preordered one (first thing I've ever preordered) and I went for sound because it was available and I knew I'd regret it later if I hadn't. Since then I've been following John's sound installations with interest and I'll do some at some point. Just yesterday I placed an order for the BLI E7 with sound, I suspect that once I have those two I won't be able to hold off on the rest.

Wutter

  • Crew
  • *
  • Posts: 283
  • Respect: +88
Re: Updated Techniques for Sound in N-scale Diesels
« Reply #5 on: February 04, 2014, 04:03:09 PM »
0
Just curious, between which pads on the decoder do you attach the capacitor to? And do you need a resistor on one of the sides to limit the charging current?
Alvin
Owner of Wutter Vehicles and Details
https://www.shapeways.com/shops/wuttervehicles

peteski

  • Crew
  • *
  • Posts: 19461
  • Gender: Male
  • Respect: +1615
    • Coming (not so) soon...
Re: Updated Techniques for Sound in N-scale Diesels
« Reply #6 on: February 04, 2014, 04:05:28 PM »
0
We do have several  models with factory-installed sound decoders.

BLI F7s
Walthers Y2 Mallets
Athearn Big Boys and Challengers
The latest is Bachmann DDA40X

I might have missed few. But at least it is a start.  :)
--- Peteski de Snarkski
--- Honorary Resident Curmudgeon

jdcolombo

  • Crew
  • *
  • Posts: 1684
  • Respect: +516
Re: Updated Techniques for Sound in N-scale Diesels
« Reply #7 on: February 04, 2014, 04:40:01 PM »
0
Just curious, between which pads on the decoder do you attach the capacitor to? And do you need a resistor on one of the sides to limit the charging current?

I don't use a resistor.  For caps of this (relatively) small size, the charging load isn't that high, and they don't induce a "fake short" mode in my command station.  The positive side of the cap is wired to the blue function common wire.  The negative side of the cap is wired to the pad that is directly opposite the blue function wire pad on the "back" side of the decoder.  If you look at this side of the decoder, you will see a bare gold pad, and then next to it at the very edge of the board, a pad with some solder on it.  The negative side of the cap gets wired to this pad (the one with the solder on it).  You'll have to cut away a bit of the plastic wrap to get to this pad.

John C.

jdcolombo

  • Crew
  • *
  • Posts: 1684
  • Respect: +516
Re: Updated Techniques for Sound in N-scale Diesels
« Reply #8 on: February 04, 2014, 04:50:42 PM »
0
Nice summary John.

I also want to add my warning about voltage ratings for tantalum caps.  I prefer to err on the side of caution, and use 20V or 25V rated caps. For me 16V is too close for comfort in N scale DCC applications.

As John mentioned (and experienced), tantalum caps have spectacularly violent failure mode.  Unlike LEDs or other semiconductor devices which simply let the smoke out when they fail, tantalum caps are more like fireworks!  Back in the day when I used to troubleshoot and repair comuter circuit boards, I had a tantalum cap actually fly off the board like rocket, missing my head by few inches.  It happened rather quickly. I saw a brief bright glow on the board I was repairing and next thing I knew was that the red-hot cap body took off the board and flew past my head.  I suspect that this cap was installed backwards (it was so burned that it was impossible to confirm that). But that incident made me treat tantalum caps with respect. Electrolytic caps can also fail, but they usually go with a small pop as the safety vent simply relieves the built up pressure.

Of course the tradeoff is that caps rated at higher voltages are physically larger than equivalent value caps rated at a lower voltage. In N scale space is at a premium, so this makes a difference. But still, I prefer caps rated at 20V.

As far as advancement goes, for me the next big step will be when (if ever) QSI will release a N scale version of their Titan sound decoder. I just love QSI sound decoders (for steam locos).

For those of you who would feel better about using a 20v cap, you might be able to get away with a 100uf cap in a diesel install.  There are quite a few 20v 100uf tantalum caps that come in a package that is no bigger than a 16v 220uf (e.g., 7.3mm long; 4.3mm wide, and seated height of between 3-4mm).  I suspect that for modern diesels, which usually have really good electrical contact, a 100uf would cure 99% of the dropouts (assuming generally clean wheels and generally clean track).

John C.

garethashenden

  • Crew
  • *
  • Posts: 1088
  • Respect: +259
Re: Updated Techniques for Sound in N-scale Diesels
« Reply #9 on: February 04, 2014, 05:05:36 PM »
0
We do have several  models with factory-installed sound decoders.

BLI E7s
Walthers Y2 Mallets
Athearn Big Boys and Challengers
The latest is Bachmann DDA40X

I might have missed few. But at least it is a start.  :)
FTFY

milw156

  • Crew
  • *
  • Posts: 380
  • Gender: Male
  • Respect: +43
    • Modutrak
Re: Updated Techniques for Sound in N-scale Diesels
« Reply #10 on: February 04, 2014, 05:52:15 PM »
0
Athearn FP 45 and F 40
Rick

Kentuckian

  • Crew
  • *
  • Posts: 411
  • Gender: Male
  • "This all started with Romans 10:9!" -Apologetix
  • Respect: +45
Re: Updated Techniques for Sound in N-scale Diesels
« Reply #11 on: February 04, 2014, 07:10:19 PM »
0
Thanks, John, for all of the information and help.
Modeling the C&O in Eastern Kentucky.
C&O HS

peteski

  • Crew
  • *
  • Posts: 19461
  • Gender: Male
  • Respect: +1615
    • Coming (not so) soon...
Re: Updated Techniques for Sound in N-scale Diesels
« Reply #12 on: February 04, 2014, 07:23:15 PM »
0
I don't use a resistor.  For caps of this (relatively) small size, the charging load isn't that high, and they don't induce a "fake short" mode in my command station.  The positive side of the cap is wired to the blue function common wire.  The negative side of the cap is wired to the pad that is directly opposite the blue function wire pad on the "back" side of the decoder.  If you look at this side of the decoder, you will see a bare gold pad, and then next to it at the very edge of the board, a pad with some solder on it.  The negative side of the cap gets wired to this pad (the one with the solder on it).  You'll have to cut away a bit of the plastic wrap to get to this pad.

John C.

Just a word of caution to the novice sound decoder installers out there about hooking up the stay-alive cap. 

Every decoder is different. If John is talking about a specific brand and model of a decoder, then his description works. But different brand decoders (or even different models from the same manufacturers) might have different hookup pads for the cap.  Best way to find out how to properly hook up that cap is to read the decoder's manual.  If the manual does not mention adding the cap then the decoder is not designed for a stay-alive cap. In that case, someone who is electronically inclined could trace the circuit traces to find a suitable hookup point for the negative side of the cap.
--- Peteski de Snarkski
--- Honorary Resident Curmudgeon

jdcolombo

  • Crew
  • *
  • Posts: 1684
  • Respect: +516
Re: Updated Techniques for Sound in N-scale Diesels
« Reply #13 on: February 04, 2014, 07:35:46 PM »
0
Just a word of caution to the novice sound decoder installers out there about hooking up the stay-alive cap. 

Every decoder is different. If John is talking about a specific brand and model of a decoder, then his description works. But different brand decoders (or even different models from the same manufacturers) might have different hookup pads for the cap.  Best way to find out how to properly hook up that cap is to read the decoder's manual.  If the manual does not mention adding the cap then the decoder is not designed for a stay-alive cap. In that case, someone who is electronically inclined could trace the circuit traces to find a suitable hookup point for the negative side of the cap.

Yes - I was describing the wiring for an ESU LokSound Select Micro.  As Peteski notes, other decoders would have different wiring - for example, the micro-Tsunami (TSU750) has a specific wire dedicated to the negative side of the stay-alive cap (the positive still hooks to the blue wire). 

John C.

Kentuckian

  • Crew
  • *
  • Posts: 411
  • Gender: Male
  • "This all started with Romans 10:9!" -Apologetix
  • Respect: +45
Re: Updated Techniques for Sound in N-scale Diesels
« Reply #14 on: February 04, 2014, 07:43:42 PM »
0
QSI lists a Titan micro on their website for N scale, but the release date is TBD.  Maybe some emails to them would help?
Modeling the C&O in Eastern Kentucky.
C&O HS