MRC 4-6-2 and 2-8-2 Sound Decoder Selection and RS-11 Install

[jdcolombo]

The 12 x 8mm had a specific number recommended, the SP-1208, but for the 11 x 15mm, from this list I was considering the second line item.  The specs seemed similar enough but the thickness was the least.  Does that look OK or would you choose a different one?

https://www.digikey.com/products/en/audio-products/speakers/156?k=&pkeyword=11+x+15+soberton&pv589=3&FV=ffe0009c%2Cfffc01b1%2Cb8a819%2C1f140000&quantity=0&ColumnSort=0&page=1&pageSize=25

I probably wouldn't use that one for two reasons.  First, it has an adhesive on the front of the speaker, which is pretty much useless for our applications and will just make things messy.  Second, the top end response is listed at 3.4khz.  That's really not enough - a steam whistle has overtones that extend well beyond that.

For speakers bigger than an 8x12mm, I tend to favor CUI.  Here's their current 11x15mm

https://www.digikey.com/product-detail/en/cui-inc/CMS-15118E-SP/102-4340-ND/7404548

Digikey also still has stock of a couple of others that are now discontinued parts, but also work fine:

https://www.digikey.com/product-detail/en/cui-inc/CDS-15118B/102-3537-ND/5355532

In general, I look for an efficiency rating of 86db or better and a top end that extends to 10khz or more.  PUI (as opposed to CUI) also has a line of speakers that would work well.

John C.

[narrowminded]

Thanks John.  I feel like I'm learning things.   And as is often the case at this stage in my life, it's typically stuff I never really wanted to know .... but I have to if I'm going to do this.

Summarizing speaker selection criteria this far:
1) Rectangular for efficient use of space
2) 8 ohms or less (down to 4 ohms)
3) 10 khz +
4) 86 db +
5) ?

[jdcolombo]

Thanks John.  I feel like I'm learning things.   And as is often the case at this stage in my life, it's typically stuff I never really wanted to know .... but I have to if I'm going to do this.

Summarizing speaker selection criteria this far:
1) Rectangular for efficient use of space
2) 8 ohms or less (down to 4 ohms)
3) 10 khz +
4) 86 db +
5) ?

Well, not sure there is a 5.  Manufacturers often make a waterproof version of a speaker that costs more, but really isn't necessary for our use.  They also sometimes have the same speaker with different connection terminals. Most of these speakers use "spring" terminals so that they can be pressed down on a circuit board trace and be powered that way.  We just solder wires to the spring terminals.  But sometimes they produce a version with solder pads, instead of spring terminals.  I tend to prefer the spring terminals, because it's harder to overheat the connection with those when soldering wires and the length of the spring terminal gives a bit more room for soldering error.  Soldering a 32-gauge wire to a 2mm rectangular solder terminal is harder than it sounds .

Beyond that, not much.  I try to buy speakers that have smooth sides, instead of a "key" protrusion, which again is something you'd just have to cut off to use in a standard rectangular box enclosure.  But I actually don't see many of the keyed speakers any more anyway, so that's usually not a problem.

Finally, don't overthink it.  You can spend hours comparing stats of various speakers, but the basic truth is that these tiny speakers don't give much, if any, output below 500hz.  There is no bass output - even if a speaker's specifications say that it goes down to 100 or 200 hz, there really won't be any useable output at that point.  So I don't worry about the low end frequency response spec; I do look at the high-end FR, because steam whistles and bells have high-frequency overtones that are important to their sound.  But 10khz is plenty high enough to capture this.  So I tend to buy whatever is available at Digikey that falls within these general parameters.

John

[narrowminded]

Thanks again, John C and everybody who posts here.   And a moment to give kudos to SBSDCC.COM Brian.  Very helpful.  No wonder so many folks sing his praises.

Based on this thread but also extensive researching over the last few weeks I think I've gathered enough understanding to take the next step and get some materials in hand to start playing, gaining the knowledge that only experience dispenses.  So, in the last couple of days I have placed orders for some components and supplies.  Digi-key for speakers and capacitors and SBSDCC (Brian) for several ESU decoders and some fine and flexible color coded hookup wire.

I ordered several of each size speaker, 12x8, 15x11, 16x9, and 18x8.  I figure all of those should be useful at some point and enough variety to demonstrate the performance effects of changes.  I will play with those and various sizes of enclosures to become familiar with the effects of those variables.  For decoders I ordered for this steam project but also diesels.  One each of the 73800 Select Micro and Direct 73199 as well as two of the 73100's  for older diesels.  This will let me see first hand all of the units I'm likely to want to use and start to plan how to use them.  Sight unseen but after reading and looking at some of the conversion candidates I'm thinking that the Direct units will have some merit for the narrow hood diesels and even diesels in general, whether DCC ready or not.  I am going to try to standardize the components, the general adaption method, and the machining as much as possible for simplicity and reliability and this will be the way to see if some of my ideas will work. 

I'm having all three of the Direct decoders shipped with an Alco 244 sound file programmed and the Micro with the Soo 2-8-2 Mikado file as recommended by John C as well as Brian.  The point in that is that I will be looking at doing the MRC steam units mentioned in the title of this thread but at the same time, some RS-11 DCC ready Atlas units (and others) as well as investigating the feasibility of some non DCC ready units like the old LL FA-1's and some RS units.  This should give me enough variety and experience doing the installations and running startup procedures with the full intention that I will get a programmer (out of stock until next month) and be able to change any and all files at will.  Depending on how far I go with this I will at some point be doing several dozen locos of every description.  I'm also thinking this effort will help my posture as my wallet won't be nearly as heavy nor bulky, applying that constant uneven force on my hips.

Thanks all.  I think it's going to be fun.

[narrowminded]

Shipment from SBSDCC arrived as well as a partial from DigiKey.  The 12x 8's should arrive tomorrow, hopefully. 

In anticipation of the speaker arrival I was able to install one of the 73100's in an Atlas RS-11 and fit the leds to it so that they sit high in the body and directly in the light pipe.  The factory light board didn't fit as nice, being well fitted at one end but pretty far back on the other.  Now they're equal and high enough to help with speaker clearance.  This should make for a very nice and very clean install.  I will get pictures and description of the work for those who care.

I ran it (without sound) and am pretty excited to see the performance and now anxious to see the whole package running.  This loco (I have three of them) is super quiet and smooth and it runs really nice on this decoder.  Also haven't fitted any caps for keep alive but probably will.  We'll see how it goes, hopefully tomorrow.

And thanks to John and all who have helped with info this far.  I feel like this is going to be a success first time out, something that couldn't have happened so painlessly without the expert guidance.  I feel like I slept in a Holiday Inn Express last night!

[narrowminded]

Haven't gotten much time to work on this but did get the chassis milled for the speaker.  Not much room but I think it's going to be alright and should get in the vicinity of that 500 cubic mm goal. 

I've seen it asked how folks are gripping the chassis for machining and being my first one, here's what I did.  I put some styrene shim pieces between all of the flat faces, at the screw faces as well.  The gap was about .026" measured with the screws in place so I used .030" styrene and screwed the halves together with the shims supporting the halves at all of the flat opposing faces.  I then clamped it in the machine vise at the fuel tank casting which was the most robust part with a square face that was also fully supported with a shim directly backing up the halves.  Many of the chassis parts would not have a directly clamped shim and tightening on those faces would invite distorting the chassis if not losing the grip and spitting the part. 

Because this setup was a one time (or two or three) and the shape didn't lend itself to a rigid setup without playing around, especially out at the end of the hollowed out spindly chassis, where the cuts were being made, I decided to go ahead with the less than optimum grip but just took light cuts.  The shims added some needed stability by bridging the two halves together but it's still not the greatest.  It's also not a huge cut so not worth the time to set it up "right".  I suspect this will be a pretty common approach to most of these different chassis modifications.

The cutter I used was a 1/8" carbide, no surface treatments, and was chosen to keep the cutter load down in the spindly setup.  A 3/32" cutter might have been an even better choice for the same reasons.  Because the cut is so small there is no real economy in trying to rush and save a pass or two and a literal minute or two, especially with the chance of wrecking the chassis.  This one went fine, the part is dimensionally perfect and we're ready for the next step, figure out the final details of the speaker enclosure. 

The general approach for the speaker is figured but the final detail confirming shell fit and clearance of the light pipe and board need to be checked and may need some minor shifting.  There's no extra space anywhere if the speaker enclosure is to be as big as desired.  To be continued...

[narrowminded]

The Soberton 12 x 8 speakers arrived so I was able to get a look at one and make the speaker enclosure plan.  There is so little room that I'm doing an odd shape, cut to fit the space available.  Due to the shape the only span on the enclosure with any flat area to speak of is the bottom so I decided to go with thinner than .040" walls and am using ABS instead of styrene to help with rigidity in the thinner walls.  The wall thicknesses are .015" to .020", depending on their location.  This has afforded internal volume between 520 and 550 cubic mm depending on the speaker allowance.  I'm hoping it works well and feel it will with my vast inexperience hard at work!  

Started the actual machining tonight.  This setup will make 10 parts, 5 each top and bottom, 5 complete enclosures.  I need three for RS-11's and I wouldn't be surprised if they could be used in other small installs.

[ Guests cannot view attachments ]

[jdcolombo]

Looks like you are making excellent progress.  Keep the posts coming - I'm really interested in how the RS11 project turns out, since the RS11 was the very first loco I put a sound decoder in, nearly five years ago now:

https://www.therailwire.net/forum/index.php?topic=31391.0

It's absolutely amazing to me how far we've come in N scale on the sound front!

John C.

[narrowminded]

Thanks, @jdcolombo .  It wouldn't be this far without the good guidance I've received on this board.   I think you'll like the cleanness of the installation that the 73100 Direct decoder affords.  No grinding of the shell required.   But nothing is perfect.  The decoder overhang at each end consumes some of the space that would be nice to be able to use for the speaker but with some finagling, what's in these posts, I think I got that covered.  Time will tell.

Meanwhile, finished the pockets in the enclosure backs tonight.  All of the machining went well.  Tedious but well.   The enclosure material was secured to the fixture block with a glue applied at .002" thick utilizing a rubber glue type product (not sure what it is, just acts that way) that goes on like a double sided tape but is actually just the glue once the backing is removed.  The thickness of the glue apply is very accurate and can be predictably allowed for in the setup.  Still have to release the parts from the fixture block, clean them up, and then setup and finish one cut on each part from the opposite side.  Not sure I can get back to it for a few days but may.  Good news is, they look like they are going to work as planned. 

And here's a picture of those parts fresh from surgery.  This picture's only worth about 50 words.

[ Guests cannot view attachments ]

[narrowminded]

I finally got the speaker enclosures finished for the RS-11 and have one assembled.  The fit is good and it makes for a clean install.  The chassis end was chosen for the end with the least metal removed as well as the end of the board with the least dimensional loss underneath due to components placed there.  It afforded the most space for the enclosure.

The enclosure is mounted to the flat spot on the bottom of the board right behind the headlight pads using the same .002" thick, tape applied glue that I used to hold the material for machining.  The step at the front is to clear the headlight pipe in the shell.  The furnished led's were used, bent and trimmed to place them as high as the shell would allow and directly aligned with the light pipes.  This whole assembly makes servicing very easy, no different than a standard loco with a standard light board. 

The milling of the frame is minimal, basically one lump off the end, and as such the net weight loss after machining but then adding back the sound components was only .9 grams.  So far I haven't added any capacitors and probably will so that .9 grams will be even closer to no net loss.  And at .9 grams I'm not the least bit worried about it.

It's about as neat and clean as my inexperienced self could come up with.  There's decent room for capacitors at the opposite end, under the board, and they should be able to mount to the board in much the same fashion as the speaker end.  That's not been fully studied.  As far as the sound quality, it works, is plenty loud, but it's the wrong sound file and I literally sat it on the track, fired it up, all of the functions were there, two minutes, didn't even set the address, and that's it.   Next step, get the Lokprogrammer (when back in stock) and learn what size hammer it takes to make it all work.

Here are some photos. 

[ Guests cannot view attachments ]

[ Guests cannot view attachments ]

[ Guests cannot view attachments ]

[jdcolombo]

That's really superb work.  The speaker enclosure is ingenious; I wonder if it would fit other hood diesels besides the RS11.

The RS11 used the 12-cylinder Alco 251B prime mover in its 1800hp incarnation.  The correct sound file for it is #73841.

The LokProgrammer is really easy to use, but the documentation is . . . wanting.  Fortunately, an end-user of ESU's products re-wrote the LokProgrammer manual to bring it up to date; it is available in the files area of the Yahoo ESU group - or if you e-mail me directly  (jdcolombo at gmail), I can send you the PDF.  One caution is that when you reprogram a sound file, you will need to disconnect any keep-alive; for reasons I don't understand (but I'm sure an electronics engineer would), having a keep alive connected seems to interfere with the programming process (at least, the sound file replacement process).  Sometimes it works, but often it doesn't.  My solution is to not complete the wiring of a keep alive until I've programmed the right sound file and programmed everything on the decoder exactly as I want.  Then I hook up "the last wire."

John C.

[narrowminded]

That's really superb work.  The speaker enclosure is ingenious; I wonder if it would fit other hood diesels besides the RS11.

The RS11 used the 12-cylinder Alco 251B prime mover in its 1800hp incarnation.  The correct sound file for it is #73841.

The LokProgrammer is really easy to use, but the documentation is . . . wanting.  Fortunately, an end-user of ESU's products re-wrote the LokProgrammer manual to bring it up to date; it is available in the files area of the Yahoo ESU group - or if you e-mail me directly  (jdcolombo at gmail), I can send you the PDF.  One caution is that when you reprogram a sound file, you will need to disconnect any keep-alive; for reasons I don't understand (but I'm sure an electronics engineer would), having a keep alive connected seems to interfere with the programming process (at least, the sound file replacement process).  Sometimes it works, but often it doesn't.  My solution is to not complete the wiring of a keep alive until I've programmed the right sound file and programmed everything on the decoder exactly as I want.  Then I hook up "the last wire."

John C.

Thanks John.  I do remember you mentioning the capacitor hookup last but I had forgotten about that.  Good I didn't already install them.  I'm pretty sure I won't need them on my layout but if I ever do run elsewhere or ever sell something, it'll be best to have them based on prevailing wisdom.  And who knows, I may find I actually do need them.   Does 200uf sound adequate for decent backup, the minimum if a soul was going to bother at all?  I know it can vary but at some point is it the job of the decoder install to cover for atrocious track conditions?  I have some 20V 100uf  tantalum polymer caps that I picked up from Digikey that seem like a decent size, cost, and safety reserve to use as a (semi) standard.  They claim that they don't suffer as severly from the unreliable voltage rating and if they do fail, it's not supposed to be as catastrophic.  They are from Panasonic and I picked up on them from (I think) a post from @woodone.  Did I say I've gotten a lot of help from this board?

And hooking them up, if I understand Pete's excellent write-up correctly (thanks Pete ) the negative side could be attached just to the left of the speaker wire as seen in one of the photos and the positive at the +aux. pad underneath.  That seems pretty easy and if I understood Pete's write-up, could even be considered the preferred choice.  And for the anal part, any reason in decoder lingo not to use red positive and black negative?

I'm hoping to play with this some more tonight.  So far, so good.

[Steveruger45]

Very nice indeed. Would you post a video when done?, would love to hear and see it.

[narrowminded]

Very nice indeed. Would you post a video when done?, would love to hear and see it.

I'd be happy to.  It will be a little while as I don't have the correct sound file and the Lokprogrammer is backordered for now.  Maybe a few weeks?   Maybe for just the fact that it makes noise I could do one of the file that's in it.  I believe it's an Alco 244, what I asked Bryan to load on the three Directs I got as I have a number of RS-3's and FA-1's to do.

[peteski]

And hooking them up, if I understand Pete's excellent write-up correctly (thanks Pete ) the negative side could be attached just to the left of the speaker wire as seen in one of the photos and the positive at the +aux. pad underneath.  That seems pretty easy and if I understood Pete's write-up, could even be considered the preferred choice.  And for the anal part, any reason in decoder lingo not to use red positive and black negative?

I'm hoping to play with this some more tonight.  So far, so good.

For the negative side, any of the many green marked spots will work.  The green circles indicate large areas of copper on the circuit board which are the "ground" of the decoder.  If you want to attach your cap to those areas, you have to scrape off the green solder mask layer and expose the bare copper.

As far as red and black wires go, it is up to you, but in the DCC world red and green are colors are assigned to right and left track connections. You would not want to mix those up with the decoders internal ground and positive voltages.  That is why other color wires are used for the capacitors.  Blue is usually the internal positive voltage and black with white stripe is often used for internal ground.