Installing Just a Speaker in a Non-Sound Loco for Lashup with Sound Loco?

[tehachapifan]

I've wondered how this might work but haven't tried it yet. Seems like it might be a way to address that sometimes-conspicuous silence emitting from a non-sound loco in a two-loco lashup. Obviously it would have its shortcomings, such as perfectly synchronized prime mover sounds from both locos, horns and bells sounding from both units, etc., plus the logistics and issues of how to extend wires between locos and accounting for output/volume changes when going from one speaker to two and then back again. Anyone ever ponder or play around with this? It seems like it could function sort of similarly to when a non-sound loco is positioned between two sound locos and it's difficult to tell that it doesn't have sound. OK, maybe not the same, but you know what I mean?

Sorry if this has been discussed before.

[tehachapifan]

….making the wires between locos something that can be periodically connected and disconnected, without becoming a source of a potential short with, say, the tracks, seems like one of the toughest challenges with this concept....especially since running the wires in the area of the MU hoses seems like a good (visual) choice.

[jdcolombo]

A major downside of this is the potential for "phasing effects" between the speakers.  You can hear the result of phasing if you have two sound-equipped locos, each with the exact same sound decoder and sound file.  If you MU these locomotives, you will hear the effects of "phasing" - the sounds will fade in and out as the various sounds either come into phase (amplifying each other) or go out of phase (cancelling each other) - it's sort of a weird doppler effect, but it is unmistakable when you hear it.  This is also why using multiple speakers in an installation (instead of one larger one) is often a problem.  Without very careful engineering, you end up with oddball phasing effects and weird sound.

This is why the best route to take for multiple sound units is either to have the second (third, fourth, etc.) unit use a slightly different sound file (this is partly why ESU offers three different versions of their EMD 645 file and a couple of different versions of the 16-cyl 567 file), or to use the "sound speed" settings of an ESU decoder (Zimo has such settings, too - don't know about others) to slightly vary the pitch and timing of the various sounds - mostly the prime mover sound.

All in all, I don't think the wiring complication and other downsides of this approach (phasing, having the bell and whistle coming from both units, etc.) outweigh the benefits of just having some sound come from the second unit.  If having a non-sound unit MU-ed to a sound unit is really bothersome, your best bet IMHO is to simply put sound in the second unit.  More expensive, I know, but in the long run it will be a better solution.  I finally took this route - if I have a two-unit lashup, both have sound, with one decoder's sound speed adjusted; if I have a three-unit lashup, the first and third units have sound, with the third unit adjusted.  In this case, having a "quiet" middle unit is less noticeable than having one "dead" unit in a two-unit MU, or having two dead units in a 3-unit MU.  Or at least, that's where I ended up.

John C.

[tehachapifan]

Thanks for the reply and insight, John.

Regarding that weird, phasing, doppler-like effect, I've definitely heard that occur with two locos running the exact same sound file at the same sound speed setting, perhaps because the sounds are not actually perfectly synchronized(?), but I've never noticed it occurring in my one and only loco (so far) that has two speakers installed instead of one. I've heard you can wire speakers together wrong, causing some kind of cancellation as you mentioned (another logistical issue to solve with this concept), but the one loco with two speakers installed sounds fine (although not noticeably any better than my other locos with only one speaker).

[peteski]

If both speakers are driven from the same decoder (and wired in-phase), there will not be any phasing.  If you have a home stereo, put it in the monophonic mode to see how it will sound.  It will just sound normal (with sound seemingly emanating in the center point between the speakers.  With our tiny speakers and small distance between them, they will sound fine. But If 2 8-ohm speakers are hooked up in parallel, the total impedance will 4-ohms. Some amplifiers cannot handle that.  If the amp can handle 4-ohms then will now drive the speaker set with double of the power that a single 8-ohm speaker would receive. So, the 2-speaker set will sound louder and the same volume setting.
If the speakers are hooked up in series, the total impedance will be 16-ohms.  The amplifier will not have problem with that but the sound level will be lower (only half the power of the amplifier will be utilized).

[jdcolombo]

Hi Russ.

The phasing problem is, in fact, because the sounds ARE perfectly synchronized at the speaker.  That's why adjusting the sound speed of one of the decoders slightly will cause it to go away.  What happens with phasing is that two otherwise-perfectly-synchronized sounds actually arrive at your ears at a slightly different time because of the distance between the two speakers (we're talking milliseconds here).  Depending on frequency, where you are standing in relation to the engines and other factors, this tiny time differential can cause sounds of the same frequency and amplitude to go in and out of phase, creating the weird doppler-like effect.  That's why you don't hear this all the time - the phasing happens intermittently as the engines move in relation to your ears.   If the time differential of the sounds is INCREASED, however, (e.g., by altering the sound speed, or by separating the speakers by a longer distance), then the sounds won't arrive out of phase, but simply be heard as different sounds - that is, the timing difference (which is still in milliseconds, but more of them!) is large enough that the sounds don't emphasis or cancel each other - they simply are perceived by your brain as separate sounds.

Multiple speaker installs are usually fine if the speakers are placed very close to each other because this minimizes the possibility of slight timing differences as the engine moves in relation to your ears causing the phasing effect.  The more distance there is between the speakers, the more likely you will get phasing effects until you get the speakers far enough apart that instead of phasing you simply hear them as separate sounds.  That's why you might not hear phasing if you have a 4-loco lashup with the first and last loco having the same sound file.  In that case, the speakers are separated by so much distance that they don't create phasing effects - your ears literally hear them as separate sounds.  But close up the distance (e.g., with two engines, or even three where the rear engine is facing backward, bringing the rear speaker - if it is located in the rear of the engine, as it usually is - closer to the front speaker), and the sounds arrive at your ears nearly, but not quite, at the same time, causing the effect.  But I've heard phasing caused by installing two speakers in the same loco, one at the front and one at the rear.  And even if the speakers are close enough together to eliminate obvious phasing, you can still get frequency-response oddities from using multiple speakers, which is why multiple-speaker installs rarely sound better than a single speaker.  Multiple-driver speaker arrays require very precise engineering to get the benefit of the larger diaphragm area represented by the multiple speakers, which precise engineering is virtually impossible for us given the space limitations we work with.

The "phasing" that I'm talking about is different from what Peteski is talking about.  If you hook up two speakers to the same amplifier out-of-phase, then you will get very serious frequency-response anomalies.  But you won't hear the "doppler-like" phasing effect that I'm talking about.  That effect is caused by two sounds at the same frequency and amplitude arriving at your ears at slightly different times - like when you have two locos with the same sound file playing at the same time, but with the speakers separated just enough that your ears get slight timing differences as the locos pass by.

John C.

[peteski]

But I've heard phasing caused by installing two speakers in the same loco, one at the front and one at the rear.  And even if the speakers are close enough together to eliminate obvious phasing, you can still get frequency-response oddities from using multiple speakers, which is why multiple-speaker installs rarely sound better than a single speaker.  Multiple-driver speaker arrays require very precise engineering to get the benefit of the larger diaphragm area represented by the multiple speakers, which precise engineering is virtually impossible for us given the space limitations we work with.

The "phasing" that I'm talking about is different from what Peteski is talking about.  If you hook up two speakers to the same amplifier out-of-phase, then you will get very serious frequency-response anomalies.  But you won't hear the "doppler-like" phasing effect that I'm talking about.  That effect is caused by two sounds at the same frequency and amplitude arriving at your ears at slightly different times - like when you have two locos with the same sound file playing at the same time, but with the speakers separated just enough that your ears get slight timing differences as the locos pass by.

John C.

Wow John, you have very sensitive ears.  For having 2 tiny speakers hooked up together (and in phase) in 2 N scale locos (or in the front and back of a single loco) separated by just few inches, you can hear the phasing effect?  I'm really surprised and impressed.  To me 2 in-phase speakers hardwired together, driven from a single amplifier would produce sound which is quite acceptable and without any objectionable artifacts.

[jdcolombo]

Hi Peteski.

The phasing effect I'm talking about is very obvious and yes, it occurs mostly when there are two decoders, each with the same sound file, in two engines MU'ed.  It is so common (and annoying) that Matt Hermann specifically pointed out that one can use the "sound speed" adjustment in the ESU LokSound decoders to avoid it.  There's a whole thread on it on the ESU LokSound Yahoo group.

Here is Matt Hermann's complete response to a question about phasing from the Yahoo list:

This is a little known “secret” with ESU decoders. One I’d like to get out there as it can be really helpful.
On my layout we randomly adjust the “Sound Speed” of the prime mover sound slot to get rid of phasing.
If the prime mover sound is on F8 you can adjust the following:

With CV32 to 1 (do this first)
CV261 is the minimum sound speed
This is the slowest it will play.

The range is 0-255. The default (normal sound speed) is 128.
CV262 is the maximum sound speed
This is the fastest it will play.
The ranges are the same as above.

If you slow these down or speed these up ever so slightly it will offset from other locos. A LITTLE GOES A LONG WAY. Though I suggest you experiment with it to see what it does, you will notice for the prime mover you can get some strange sounds if you go too far one way or the other!

BTW this can be done for all the sound slots. I even occasionally use it on the horns and bells. I can’t stress enough though that minor adjustments will give you the best results.

Have fun!
Matt Herman
General Manager
ESU LLC

You don't have to have golden ears to hear this.  And yes, it is caused by minor timing differences in the same sounds reaching your ears.  Most people don't know just how sensitive normal human ears are to sound timing - in fact, processing timing differences of mere milliseconds is what allows us to localize sounds.  In a stereo system, very slight timing differences can smear transient response and cause irregular frequency response, which is why recording studios have their mastering rooms treated with sound absorbent material and choose very-high-end time-aligned speakers for monitoring recordings.

Yes, it is important to hook speakers up "in phase."  What that does is make sure that the speaker diaphragms are moving the same way at any given frequency.  But simply hooking up speakers "in phase" doesn't account for timing differences that result from speaker separation.  That's why you will often see high-end speaker designs that use sloped baffles - to better time-align the sounds coming from the separate speakers.  It is also why many people prefer the sound of electrostatic or planar magnetic speakers, which have a single diaphragm that covers the entire audible sound spectrum - these speakers inherently have no timing errors.

If you'd like to read some articles/books on the subject, I can recommend some.  Let me know.

John C.

[RBrodzinsky]

I have definitely experienced this effect as well. It is extremely annoying when it happens, but as John says, also easy to avoid by adjusting the sound speeds.

[peteski]

Guys, either I'm missing something here or you are missing the fact that the original question specifically addressed feeding both speakers from a single decoder.

I don't see how in that setup any phasing will be audible. Like my earlier example of using your home stereo system in monophonic mode (same audio feed through left and right speaker).  In that scenario you can walk around the room and not hear any phasing. Same will apply to 2 speakers mounted couple inches apart.

[jdcolombo]

Well, the empirical method works.  Just try it.  If it sounds OK, then no problem! 

John C.

[tehachapifan]

Well, the empirical method works.  Just try it.  If it sounds OK, then no problem! 

John C.

It may come to that....if some of the other logistics can be worked out. I may go a different route to start with that's sort of similar, but uses a lashed caboose to carry the decoder with the speaker mounted in a loco that a sound decoder won't also fit into. I already use the lashed caboose concept...thankfully the SP hauled a caboose around with the loco on a lot of local jobs...but both the sound decoder and the speaker are currently located in the caboose (there's a non-sound decoder in the locos I use this with).

Guys, either I'm missing something here or you are missing the fact that the original question specifically addressed feeding both speakers from a single decoder.

I don't see how in that setup any phasing will be audible. Like my earlier example of using your home stereo system in monophonic mode (same audio feed through left and right speaker).  In that scenario you can walk around the room and not hear any phasing. Same will apply to 2 speakers mounted couple inches apart.

This is what I would've thought as well. As I mentioned earlier, I've definitely heard the weird doppler-like phasing effect from two locos, each with its own decoder running the exact same sound file...where you get that periodic "jet flying over" sort of effect....but I wouldn't expect it with 2 speakers hooked to the same decoder either for the same reason as the home stereo example. Also, I would have thought that weird jet flying over sound, which is what I assume John means with the "doppler effect" reference, was caused by the two decoders playing at an ever-so-slightly different speed, being they have their own components to process the sound through, even if they were set to the same speed setting. All this said, this is one of the reasons why I posted the question.

[jdcolombo]

Yeah - "jet flying over" is a better description. 

I've never tried wiring a speaker in a separate loco to a decoder in another loco with its own speaker.  It might be fine.  And I suspect that you're right, Russ, that part of the "JFO" effect is a very slight timing difference caused by the second decoder - not a big enough difference to eliminate the effect, which you cure by adjusting the sound speed.  And it might be that if the two decoders were, in fact, absolutely perfectly synchronized (as would be the case with wiring two speakers to the same decoder) this effect would not appear.  I hope you try it, and report back!

And on logistics - the wiring would be the toughest part, but I think you could get away with some 32-gauge magnet wire to connect the second speaker.  I think the maximum current draw for 1 watt into 8 ohms is about 350ma (2.83v into 8 ohms), and that is well within 32-gauge wire for chassis wiring.  Paint it black, and it looks like MU hoses.  Finding a convenient way to disconnect the two engines might be more of an issue, but perhaps some sort of single-pole micro connector could be built into the front of the shell of the second loco at usual MU hose locations without it being too obvious - Peteski might have thoughts about this.

John C.

[tehachapifan]

Yeah - "jet flying over" is a better description. 

I've never tried wiring a speaker in a separate loco to a decoder in another loco with its own speaker.  It might be fine.  And I suspect that you're right, Russ, that part of the "JFO" effect is a very slight timing difference caused by the second decoder - not a big enough difference to eliminate the effect, which you cure by adjusting the sound speed.  And it might be that if the two decoders were, in fact, absolutely perfectly synchronized (as would be the case with wiring two speakers to the same decoder) this effect would not appear.  I hope you try it, and report back!

And on logistics - the wiring would be the toughest part, but I think you could get away with some 32-gauge magnet wire to connect the second speaker.  I think the maximum current draw for 1 watt into 8 ohms is about 350ma (2.83v into 8 ohms), and that is well within 32-gauge wire for chassis wiring.  Paint it black, and it looks like MU hoses.  Finding a convenient way to disconnect the two engines might be more of an issue, but perhaps some sort of single-pole micro connector could be built into the front of the shell of the second loco at usual MU hose locations without it being too obvious - Peteski might have thoughts about this.

John C.

Thanks, John. I'll keep pondering how to go about accomplishing this. I agree, the wire routing and connection methods are going to be a challenge.

….Off topic but sort of related story....went to a big concert recently a major league baseball stadium where the stage, with its own bank of speakers, was in center field at about the warning track and they had also placed some additional speakers in the infield diamond area, presumably for the benefit of the people located in the stands more or less behind home plate and the farthest from the stage. These additional speakers seemed to have a (programmed?) delay, which probably makes sense for those people straight back and so the sound is still in-time with the stage speakers (accounting for the speed of sound, I'm guessing), but we were seated on the side of this setup along the 3rd base line, about an equal distance from the stage speaker and the infield speakers, and the sound was an absolute mess where we were. Two separate drum beats and bad echoing all night!

[peteski]

The weird phasing effect everybody is describing here occurs when 2 locos with their own decoders are running the same sound file, but since the electronic components are not identical on both decoders, the sound played from one of the decoder is delayed by few milliseconds.  It is a reverb. Standard stadium PA systems (where all the speakers are fed from the same audio source), where the speakers are positioned hundreds of feet apart exhibit an extreme reverb where you can actually hear the distinct sounds from each of the speakers.  At that point it becomes an echo.

But if the speakers fed directly by the same source (which is what Russ was asking for), and placed just few inches apart should not produce any appreciable phasing or reverb.