Broadway Limited T1 Speaker Transplant

[jdcolombo]

Is there sufficient space (between the plastic posts and the circuit board) to accommodate a pair of those 13x18 speakers?

I think that would be pushing it.  If you did two 13x18's each in their own enclosure with .030 styrene, the total length would be 18x2 = 36, plus 3 (.75 x 4 for the enclosure sides).  That's 39mm, and with square enclosures, I think you'd be over the holes for the posts that hold the tender shell on.  Note that the posts you see in the photo at the far rear of the tender floor are NOT the posts I'm talking about.  Instead, it's the two holes slightly in front and off to the side of those posts.  That's where the screws from the underside of the tender floor go and screw into posts in the tender shell itself. 

However, you could pretty easily get two 11x15's in.  And there is also this, which I think would fit since the overall length is 36.3mm:

https://store.sbs4dcc.com/SBS4DCCSugarCubeSpeaker13x18mmw/TwinCoupledSoundChamber.aspx

The only other question I'd have is whether the Paragon 3 would be OK with either a 4 ohm (two 8-ohm speakers wired in parallel) or 16-ohm (wired in series) load.  Most audio amps that can handle an 8-ohm load can also deal with 4 and 16, but I don't know this to be true of the Paragon 3.  If you could find 4-ohm speakers, you could wire them in series for an 8-ohm total load (or if you have two 16-ohm speakers, wire in parallel and you get 8 ohms).  But the cell-phone type speakers sold at the main electronics suppliers (Digikey and Mouser) are mostly 8 ohm. 

John C.

[peteski]

Yes, quite! Thanks, John.
You would think BLI would note that somewhere. If they did, I can't find it.
I need to look into my diesel Paragon 3's...
Otto

Why would they advertise the fact that they now use 8 ohm speaker?  Or even include that info in their manuals?  They are selling a compete product (locomotive) with a built-in BLI sound decoder. They expect their customers to put them on the track and run them, not take them apart, reverse-engineer, and modify them!  Their target market is runners, not hackers. 

Now if BLI,  as a purveyer of DCC decoders, was selling decoders separately, I would then expect them mentioning the 8 ohm speaker in the technical specs. But they aren't in the market for that.

And to confirm, my BLI Centipede uses a pair of 100 ohm  speakers (probably hooked up in parallel, to make them 50 ohm).

[daniel_leavitt2000]

That just brings up another question. Why use a 50 ohm speaker to begin with? Wouldn't that mean they couldn't use any off the shelf digital amp or other circuit in addition to the speakers?

[peteski]

That just brings up another question. Why use a 50 ohm speaker to begin with? Wouldn't that mean they couldn't use any off the shelf digital amp or other circuit in addition to the speakers?

That is a very good question.  Higher impedance speakers work on higher voltages and require less current to produce the same amount of power than a low impedance speaker.

Just a quick and rough demonstration: To produce 1 Watt of acoustic power a 50 ohm speaker will need around 7V across its terminals (and will consume 140mA).  An 8 ohm speaker needs 2.8V (and consume 350mA) to produce 1 Watt of acoustic power. This is greatly simplified as I'm using strictly DC power calculations, but it does demonstrates the difference in voltage and current needed to produce the same output power.

The lower current consumption might be the reason that early decoders used high impedance speakers.  I suspect that the early sound decoders used analog sound amplifiers which by design have to waste (and dissipate as heat) lots of power.  New decoders use Class-D (digital) audio amps which are very efficient (very low amount of power has to be wasted as heat). Class-D amps are PWM devices (like the decoder's motor control). Very efficient and can easily supply higher currents to a low-impedance speaker.

Or maybe those were the only small speakers available at that time?  That was before smart phones and tiny speakers were likely just used for headphones (where higher impedance is usually the norm).

[jdcolombo]

That is a very good question.  Higher impedance speakers work on higher voltages and require less current to produce the same amount of power than a low impedance speaker.

Just a quick and rough demonstration: To produce 1 Watt of acoustic power a 50 ohm speaker will need around 7V across its terminals (and will consume 140mA).  An 8 ohm speaker needs 2.8V (and consume 350mA) to produce 1 Watt of acoustic power. This is greatly simplified as I'm using strictly DC power calculations, but it does demonstrates the difference in voltage and current needed to produce the same output power.

The lower current consumption might be the reason that early decoders used high impedance speakers.  I suspect that the early sound decoders used analog sound amplifiers which by design have to waste (and dissipate as heat) lots of power.  New decoders use Class-D (digital) audio amps which are very efficient (very low amount of power has to be wasted as heat). Class-D amps are PWM devices (like the decoder's motor control). Very efficient and can easily supply higher currents to a low-impedance speaker.

Or maybe those were the only small speakers available at that time?  That was before smart phones and tiny speakers were likely just used for headphones (where higher impedance is usually the norm).

I suspect it was the power consumption issue.  8 ohms has always been a sort of standard impedance for speakers of virtually every size.  I remember taking apart a tiny transistor radio back in my childhood days - say circa 1965 - and it had an 8-ohm speaker (I remember this because the radio was . . . uh . . . my sister's, and she didn't really appreciate my disassembly of her radio.  Though I DID get it put back together!).  I think your observation about moving to Class-D amp circuits is right on the money.

John C.

[mmagliaro]

John, you did a beautiful job.   I am not even a fan of sound in model trains, but no doubt, what you did makes that T1 sound 10 times better than the stock engine.  I would never have expected that much low-end tone to come out of a small model.

[Cajonpassfan]

...
I remember taking apart a tiny transistor radio back in my childhood days - say circa 1965 - and it had an 8-ohm speaker (I remember this because the radio was . . . uh . . . my sister's, and she didn't really appreciate my disassembly of her radio.  Though I DID get it put back together!).
...
John C.

Lol, THAT explains everything John
That's why you're way ahead of the rest of us now...
Otto

[jdcolombo]

John, you did a beautiful job.   I am not even a fan of sound in model trains, but no doubt, what you did makes that T1 sound 10 times better than the stock engine.  I would never have expected that much low-end tone to come out of a small model.

Thanks, Max.

But my Berks sound even better with an ESU LokSound and their newest sound file for the SP GS4.  Yeah, not a Berk (I'm waiting for Matt Hermann to record NKP 765), but it's a similar Super Power design, and the chuffs sound pretty much like the NKP 765's.  I picked an alternate whistle that's also pretty close to the NKP's version, and well, bells are bells, and steam hiss is steam hiss, more or less. 

I have a more favorable view of the Paragon 3 in the T1 than the Paragon 2 that I had in my BLI PA-1's.  But I still don't think it is as good as a LokSound . . .  Unfortunately, there are no operating T1's that Matt can record (yet - the T1 Trust is building a brand new one, so maybe one day . . . ), so all of the sounds are sort of "best guess" anyway.  Except, presumably, for the whistle.

John

[propmeup1]

John, Thanks for the info. I am now on board the group here.  I have two of the speakers you used on order and will swap them out. I'm going to do one and run both together just i can hear the difference in person before i change the second one.

Thank you,
Keith

[jdcolombo]

John, Thanks for the info. I am now on board the group here.  I have two of the speakers you used on order and will swap them out. I'm going to do one and run both together just i can hear the difference in person before i change the second one.

Thank you,
Keith

Welcome to The Railwire, Keith.  If you have any questions, you can always e-mail me directly at jdcolombo at gmail.

John C.

[cbroughton67]

Thanks for the write-up, John! I'm sure the difference in sound quality is even more apparent in-person. I just ran mine for the first time yesterday and thought it sounded like a state-of-the-art sound decoder... from 10 years ago. I'm looking forward to making these changes to mine, now!

Chris

[Lemosteam]

@jdcolombo Does this speaker fit the bill? 13x18,  Lowest Hz at 200 compared to three others I could find.

https://www.digikey.com/product-detail/en/SP-1813S/433-1136-ND/6099110/?itemSeq=278604829

[chicken45]

Great write up! Thank you for sharing!

[jdcolombo]

@jdcolombo Does this speaker fit the bill? 13x18,  Lowest Hz at 200 compared to three others I could find.

https://www.digikey.com/product-detail/en/SP-1813S/433-1136-ND/6099110/?itemSeq=278604829

Hi John.

Yep, that will work (it's probably the one I'll use when I exhaust my supply of Knowles 13x18's, which are no longer made).  But you're not going to get any usable sound output from it at 200hz.  If you look at the data sheet, you'll see that it's sound output drops off a cliff below 800 hz.  For all practical purposes, 500hz is about the lower limit for any of these speakers, although the 13x18 will sound noticeably "richer" (more low-midrange output) than, say, an 8x12mm.

John C.

[propmeup1]

My new speakers arrived today and they are in. Took about and hour to do both. I'm glad I stumble over this conversion, for it has made a big difference in sound.  My two T1s sound happier now.
Thank you John

Keith