Speaker recommendations

[Cajonpassfan]

My “go-to” favorite speaker has been the “Grand” 13x18 mm in tenders and wide body first generation diesels. Mated to a good size enclosure, they provide very good sound given their relatively small size, and their sound is SO much better than the smaller speakers in my narrow body diesels. Unfortunately, the “Grand” is no longer available, and my stash is running low...

Two questions then:
One, what is the next best thing in that size category?
Two, if space were not a big concern, as in a dummy wide-body diesel, what is the optimal speaker to use? I’m experimenting with larger “bass” speakers (see below) and frankly expected a much better performance than my old standby the “Grand”. Better, by not by much.
Thoughts and recommendations appreciated,
Otto K.

[jdcolombo]

The 13x18mm speakers sold by Streamlined Backshop I think are the last of the Knowles production; in any event, they are excellent.  But kind of pricey given that bare speakers are $2-$3.  For bare speakers, I've had really good results with the 13x18mm cell-phone type speaker made by CUI and available from Digikey (and probably Mouser).  Soberton, which makes the 12x8mm speakers I use, also makes a 13x18 (I think), but I've never tried it.

If you have lots of space, I'd try wiring two 13x18mm speakers in parallel.  Two 8-ohm speakers in parallel will give a 4-ohm load, well within the specs of modern sound decoders like ESU and Tsunami and WOW!.  It also doubles the effective diaphragm area.  SBS sells dual-speaker enclosures (3D printed) if you don't want to roll your own.

Never had much luck with the oval style speaker shown in your photo.  The flat cell-phone types are MUCH better.

John C.

[peteski]

Never had much luck with the oval style speaker shown in your photo.  The flat cell-phone types are MUCH better.

John C.

Yes, all those old-school oval speaker seem to have fairly stiff suspension, so they need quite a bit of power to drive them.  The softly-suspended cell-phone-type rectangular speakers have much softer suspension, so they are more efficient. At least that's how I understand it.

[RBrodzinsky]

Otto,

For the Erie Builts I did for you, be careful. The speaker wires you requested are already in parallel with the 8Ω speaker installed in the B unit.  A single 8Ω speaker attached to them will already take the total impedance down to 4ohms (unless you disconnect the B unit speaker).

[Cajonpassfan]

Thanks guys, I appreciate your feedback, but now I am confused.
John C., my sound hero, suggests the 4 Ohm impedance is well within the parameters, while Rick B., my other hero, tells me to be careful... so now I’m reluctant to connect the second speaker.
And yea, the oval is mildly unimpressive on its own... I expected a fuller sound.

So, if I use instead a double 13x18 in the dummy, with large enclosures, how do I wire them? In parallel to each other and with the little guy in the powered B unit? Should I worry about blowing the (pricey) ESU LS 5?

Oh, the challenges of N scale sound...
Otto

[nickelplate759]

Rick B is telling you that he thinks you already have a second 8Ω speaker in the circuit, in the B unit of your Erie-Builts.  In that case you would not want to add a third speaker to the circuit.

[RBrodzinsky]

Rick B is telling you that he thinks you already have a second 8Ω speaker in the circuit, in the B unit of your Erie-Builts.  In that case you would not want to add a third speaker to the circuit.

I know since I installed them for him

This is the B unit for the non-DCC ready LifeLike Erie Built, milled to within a mil of its life to get the decoder and speaker in (the leading A got a LokPilot Micro)

[ Guests cannot view attachments ]

On the left, coming off the decoder, one can see the brown speaker wires I installed to include hook up a second speaker, in the dummy A unit. With this frame, while there is a lot of lateral space, there is very little depth available for any speaker larger than the Soberton 1208. I was able to get a minimal amount of depth below the speaker, and then extend the enclosure laterally.  Here’s the frame as milled; the problem is the height of the bearing springs and the screws. You can see the small pocket I was able to create

[ Guests cannot view attachments ]


Otto, if you want to put two larger speakers in the dummy A, disconnect the speaker in the B.

[Cajonpassfan]

Thanks Rick for the advice. I think I understand now. Two wired in parallel ok, three not ok.
But that would move allthe sound to the trailing unit of the three unit set, not ideal. I think I want to keep the Soberton in the B, and use one speaker in the trailing A, wired in parallel. So it seems I came a full circle to my original question: if space was available in the dummy, what would be the best speaker choice? Is there anything bigger and better than the 13x18 that would fit? If the answer is no, I’ll just use one of the “Grande” 13x18’s...they are quite good.
Thank you all for the replies,
Otto

[RBrodzinsky]

I think the answer is "no", since we are limited by the width - really cannot go wider than 13mm, to leave room for an enclosure

[jdcolombo]

There may be another possibility.

You can have two sets of speakers wired in parallel, and then wire those sets together in series, and end up with the same impedance that each speaker has.

Example:  Two 8-ohm speakers wired in parallel = 4 ohms.  Two such sets of speakers wired in series = 8 ohms. 

What I don't know is if you could wire a single 8-ohm speaker in series with a parallel set of 8-ohm speakers and be OK.  I think electrical theory indicates that the resulting load would be 6 ohms.  But I'm not sure if the volume level coming from the single speaker vs. the parallel-wired pair would be comparable, or if one would be noticeably louder/softer than the other.  Maybe Peteski can weigh in on this.

If it DOES work, then you could wire a pair of speakers in parallel in the trailing A, and then wire the speaker in the B in series with the trailing A pair for an overall 6-ohm load.

John C.

[peteski]

There may be another possibility.

You can have two sets of speakers wired in parallel, and then wire those sets together in series, and end up with the same impedance that each speaker has.

Example:  Two 8-ohm speakers wired in parallel = 4 ohms.  Two such sets of speakers wired in series = 8 ohms. 

What I don't know is if you could wire a single 8-ohm speaker in series with a parallel set of 8-ohm speakers and be OK.  I think electrical theory indicates that the resulting load would be 6 ohms.  But I'm not sure if the volume level coming from the single speaker vs. the parallel-wired pair would be comparable, or if one would be noticeably louder/softer than the other.  Maybe Peteski can weigh in on this.

If it DOES work, then you could wire a pair of speakers in parallel in the trailing A, and then wire the speaker in the B in series with the trailing A pair for an overall 6-ohm load.

John C.

If I understand your description, the resulting impedance woudl be 12 ohms.  The current (thus the power) dissipated by each speaker would be as follows:  The 8 ohm speaker would dissipate half the power, and the other half of the power woudl be divided equally between the parallel-connected speakers. So each of the parallel-connected 8 ohm speakers woudl be dissipating 1/4 of the total power.  Power equals volume. So you likely aren't gaining much sound volume from the parallel connected pair.

However, if you had 4 identical speakers and connected 2 pairs in parallel (resulting in  4 ohm pairs), then connected the parallel pairs in series, you woudl end up with 8 ohm impedance for that set of 4 speakers.  Each speaker of that setup woudl dissipate 1/4 of the total power.

[jdcolombo]

Hi Peteski.  Thanks - that's the analysis I was looking for.  Yep, 12 ohm total load, not 6.  Did the math wrong.  12 ohms, though, would be OK from the standpoint of a modern decoder amplifier's power delivery.  ESU is good for 4-16 ohms, I think. 

The problem is the one you identified - the volume output of such an arrangement.  What I was thinking is that if the 8 ohm speaker dissipates half the power, and the other two dissipate the other half (1/4 each), wouldn't the volume between the single speaker and the pair be equal? 

Let's say the decoder is putting out 1 watt.  If 1/2 watt goes to the one 8-ohm speaker, and 1/2 watt goes to the pair (1/4 watt each), then the volume level between the single 8 and pair would be the same, IF the volume output curve of each speaker is the same and if that curve is linear - that is, if all 3 speakers are identical, and have identical power-to-spl curves, woudn't the volume be equal?  Or put another way, suppose all 3 speakers put out 60 db SPL at 1/2 watt, and 30 db SPL at 1/4 watt.  In this case, the single 8-ohm speaker would be pumping out 60db, and each of the pair would be pumping out 30db, which combined would be 60db.

That's probably a moot point, though, since in this case, the speakers would be from different manufacturers, probably with different power curves.  And it's probably the case that SPL is not a linear relationship to power anyway.

Hmmm.  I'm thinking an empirical experiment might be in order . . .

John C.

[peteski]

Hi Peteski.  Thanks - that's the analysis I was looking for.  Yep, 12 ohm total load, not 6.  Did the math wrong.  12 ohms, though, would be OK from the standpoint of a modern decoder amplifier's power delivery.  ESU is good for 4-16 ohms, I think. 

The problem is the one you identified - the volume output of such an arrangement.  What I was thinking is that if the 8 ohm speaker dissipates half the power, and the other two dissipate the other half (1/4 each), wouldn't the volume between the single speaker and the pair be equal? 

Theoretically, yes.

Let's say the decoder is putting out 1 watt.  If 1/2 watt goes to the one 8-ohm speaker, and 1/2 watt goes to the pair (1/4 watt each), then the volume level between the single 8 and pair would be the same, IF the volume output curve of each speaker is the same and if that curve is linear - that is, if all 3 speakers are identical, and have identical power-to-spl curves, woudn't the volume be equal?  Or put another way, suppose all 3 speakers put out 60 db SPL at 1/2 watt, and 30 db SPL at 1/4 watt.  In this case, the single 8-ohm speaker would be pumping out 60db, and each of the pair would be pumping out 30db, which combined would be 60db.

That's probably a moot point, though, since in this case, the speakers would be from different manufacturers, probably with different power curves.  And it's probably the case that SPL is not a linear relationship to power anyway.

That sounds logical.

Hmmm.  I'm thinking an empirical experiment might be in order . . .

And I agree again! 

[Cajonpassfan]

I did end up using the “bass” speaker in the dummy A. Coupled to the Soberton, it’s actually pretty impressive, as the sound emanates from two relatively distant locations (front of B, back of trailing A), and the two speakers have a distinctly different timber. They are also loud; I need to turn them down and tweak some sound effects. The screaming dynamics, though,  are awesome on the downgrade 8
More on my LA Division Layout Engineering thread...
Otto K.

[peteski]

Excellent!
I know that Soberton is one of the rectangular cell phone speakers, but what is a "bass speaker"?

Each speaker is 8 ohm? How are the speakers connected?