Hey all, Wanted to see if anyone has attempted to make a DIY sound decoder? I have built a few decoders for motor control but was interested in advancing to a diy sound all in one decoder. Thanks! Sent from my SM-G920P using Tapatalk
You'd have to get all the board components, sound board circuits, etc. Probably cheaper to go a commercial route. Better sound too.
All the components are cheap. I may have just found a way. If successful, ill share. Sent from my SM-G920P using Tapatalk
I'd be interested in your progress on this. Not sure it'd be any cheaper than what the major manufacturers can get/obtain, because they can get components much cheaper as they buy millions of parts on a very large scale.
True, im not getting the components for cents but at around $6 in total price still beats $100 Sent from my SM-G920P using Tapatalk
Late to the party here. I was wondering where you are going to get your sound files from? Looks like a lots of work just to save a $? I have very little time to invest into something I can buy off the shelf and is backed by some kind of warranty. Where are you going to get your PC made? Also I am into N scale and getting small items stuffed into an N scale unit is very trying. So I will invest my time in something else for now. Best of luck and keep us posted has you progress forward.
Sorry for the late response, for HO scale. I've been playing around with certain aspects. Haven't had much time to really sit down and do it completely. The sound files will be coming from certain open sources for now. Mostly its to toy around with. I get the whole warranty thing etc and spending the $70 on a sound decoder but this is something that interest me. If a 14 year old can do it on YouTube (I have the link somewhere) than awesome would love to try it. Sent from my SM-G920P using Tapatalk
I want to see more of this. If you can do it for $6, why should I pay $70-100? Soundtraxx appears to have some sound downloads or recordings you may be able to use.
EXACTLY! Essentially all decoders really are, are glorified mp3 players. The board is the most expensive part ($6) its a matter of mapping the wiring. The chip itself that stores all the sounds and codes is 30 to 60 cents. All I would really have to do is either write my own code to respond to functions or copy the coding on any decoder and just change the sound files. You are correct, there are free sound librarys such as Massoth etc that you just download and transfer the sounds over. It's a really inflated market. The hardwares cheap, the programming is a set standard the only cost I could really see would be due to the sounds. (Paying a crew to go out once and record an engine) all of the programming of a decoder can be done from a laptop. Sent from my SM-G920P using Tapatalk
No one wants you to dampen your enthusiasm to hobby on something. If you want to make a motor decoder, you can make one cheap, but not really small. If you want to make a sound decoder, you can make something cheap, but it won't approach either the quality and fidelity of the mainstream decoders, and you will never find the sound file repertoire to make it compete. So, it's an exercise, and probably fun, but without a significant investment in a custom asic and sound library files, you won't be able to equal even a $50 decoder. Calling a sophisticated DSP and other hardware a glorified mp3 player shows you don't fully fathom what is going on inside. Just tell me how an MP3 player can do doppler... Again, not meant to dampen Greg
It seems that I am being shot down lol seeing as ive seen decoders about the same size of the market decoders. I'm not looking to market these. If successful I will share. I have seen a few young kids with self made sound decoders that matched the same size of that of the market decoders. On doplar I do know exactly what your talking about. It has been explained to me. Thanks for your concern lol Sent from my SM-G920P using Tapatalk
Not shooting you down, just reacting to your statement: "Essentially all decoders really are, are glorified mp3 players." Very untrue, and reflects a fundamental misunderstanding. Think about what it will take to do doppler shift of the horn/whistle and you will see it's not a simple player. Also, think about what it takes to change the chuff on a steam loco to make it "bark" or drift", and not just volume change. The point is you CAN make a "glorified" MP3 player, but it WON'T be the same a quality sound decoder. If just different loops of the same sounds is fine, you are good to go, but there IS a difference. Greg
[QUOTE="... All I would really have to do is either write my own code to respond to functions or copy the coding on any decoder and just change the sound files. [/QUOTE] The code in the manufacturers decoders are protected by patents, and as such you are walking on a slippery slope. While the parts to make the decoders may be cheap, it is the programming that is the intellectual property of the manufacturer. With commercial decoders, that is what the bulk of the cost is, the programming, not the parts. They all have a pretty big investment in the software development. It is doable, but there is a reason why ESU LokSound decoders sound and move so well. It is the thousands of lines of event driven micro-code they developed to run on the $6 parts.
Another BIG difference between sound decoders and mp3 players is polyphonic sound. An mp3 player can only play one sound at a time (well actually two with a stereo track, but the two are combined into one track and can not be started and stopped individually). A sound decoder needs to be able to play more than one sound at a time - you don't want the prime mover to suddenly go quite when the bell is rung or the horn is sounded. Also, you probably can't read code off another decoder anyway. Most programmable microcontrollers have the ability to set a bit when programming so that the code can not be read back off of it. Even if you could get the code off, an experienced programmer would have a hard time using any of it unless he is used to dealing with machine language, because the code's not going to be in a high level language like C.
Oh for heaven's sake, all these nay-sayers! The hobby is all about having fun. Go for it J9! It sounds interesting and you will certainly learn quite a bit about DCC control and sound. The idea of using sound samples instead of proprietary file formats is intriguing.
Concerning inexpensive decoders for a loco, it seems to me that the vendors inflate the menu of available sounds to inflate the price of what they offer. Sure, the sounds are 'authentic', but after a while they're just noise. In general, most sounds are really superflous and mostly just background noise, IMHO. The first level sound is simply a bell and horn/whistle, and the rest is just excess that isn't used 99% of the time. At my club, all the extraneous sound is just distracting.
I'm cooking something up. With what little time I have. I've actually come close to finishing a solid project. The cost? close to $10. Sent from my SM-G920P using Tapatalk
