Skip to main content

Per @gunrunnerjohn's suggestion, I decided to take on building my own whistle box.

I use a KW transformer with the original rectifier disc.  I also have mostly PS1 stuff from MTH.  As I started commissioning my fleet, I was not able to consistently blow the horn on all locos, and on the ones I *could* get the horn to blow with the KW, the voltage would **** near drop out and the engines would stall as the horn blew. I did try using some lighted cars to put a load on the transformer, but it really didn't do much.

Also, as many of you know (but I didn't), with the rectifier disc you get *either* the bell *or* the whistle, but you can't have both.  The polarity of the track connection determined which one was going to sound when you wiggle the KW's horn handle *just so*.

But the biggest issue for me was that I couldn't consistently use the KW to program my PS1 boards (or, to be more precise, return them all to factory settings using reset 18).  That was an issue, because the smattering of boards that I had would have various features turned off or on and it made doing things like uncoupling a bit tedious.  

So, after researching the issue, and with some of John's advice, I cobbled together about 12 bucks worth of parts from the local electrical surplus store and got busy.  (I originally planned to do it next week while I watched basketball, but we all know how that ended up >

Took me about 2 hours altogether, using some really crude soldering, some electrical tape and frequent visits back to John's pictures from his other thread about building a box.  And wouldn't you know it, this non-electrician mechanical guy was able to get it to work on the FIRST TRY!

I did do a couple things slightly different than John's.  First, I used six diodes instead of 5 in each direction.  Just basic, run-of-the-mill 400V 6A diodes.  I did run some tests with 5 diodes each direction vs. 6, but sometimes with 5 the horn would just blow randomly for several seconds.  Interesting, but not the effect I Was looking for.

Second, I used Normally Open switches instead of Normally closed.  So instead of running constant voltage and then slightly dropping the voltage across the diodes by opening the circuit when the switch is pressed (resulting in a corresponding slowdown of the loco when the horn blows), the voltage is slightly boosted when the horn switch is closed (biasing the DC positive instead of negative I guess).  This was a bit more desirable to me, since I tend to run things pretty slow anyway, and the very slight speed boost isn't enough to bother me.  There's a video posted so you can see what I mean.

All in all, Im super excited that I'm learning some new tricks with the help of this forum.  Next up:  Gonna make me a bunch of BCR's!

Photos and video of the box in action!

The Guts

Had to wrap the diode series in a bunch of tape to prevent accidental shorts .  It's a mess, but I'm not a pro and I'm actually kind of proud.  Too bad you can't see my bang-up soldering job! (It's bad).

The lidReady to closeThe finished Setup

 

 

Attachments

Images (4)
  • The Guts: Had to wrap all the diodes in tape so there weren't any accidental shorts. Sorry you can't see my bang-up soldering job!
  • The lid: Used Normally Open Switches
  • Ready to close: Ready to close
  • The finished Setup: The finished Setup!
Videos (1)
20200312_215205_Trim
Last edited by Jeff_the_Coaster_Guy
Original Post

Replies sorted oldest to newest

Nice and neat.

The loco motor(s) might like the DC bias. Some just do.

The N.O. should cause the slowdown as once closed, only then is power is leached to create bias. Leached being the key word; a power loss on half the ac wave.

Not too too important if it works, but "theory". (Got a meter? See the actual vs design expectations of AC & DC biases yourself with some quick voltage readings and add/subtractions

 The number of diodes needed depends on their efficiency, some have more voltage drop than others. Cheap general purpose usually eat more.

I'm sure it'll all hold; looks like you could be heavier handed on solder, but far from sloppy or "need" to add.   In a factory the penny pinchers would love ya!

Next you'll be dialing in constant voltages with them (opposing pairs in parallel for an even ac drop )

 

Looking good, and solves the problem at hand.

FWIW, my circuit uses N/O switches and does what you describe.  That is, it drops the voltage all the time, and when you press the whistle or bell button, the DC offset gets added to the AC waveform. 

I've never seen the issue of odd behavior with five diodes, but several people did mention it.  That's why I suggested six diodes, that seems to be the "magic" number.

The diode I used was just a normal 400V 6A rectifier diode (I think that's what they're called). I grabbed a bunch from a drawer at the surplus shop, they didn't have part numbers per se.  Just a voltage and an amp rating.  I think any diode between 200V and 600V rating, and a minimum 4A or anything above would be fine.  I agree, it's overwhelming at first because there's a million kinds.  I just bought the cheapest "inline" type I could find (i.e. a tiny little black cylinder with straight leads poking out each end).

Maybe John can link back to his original post.

My only caution: make sure you solder the diodes close enough together that the string of them doesn't exceed the inside length of your box minus an inch or so.  More room to work in the box is going to make your life much easier, especially if you have fat fingers like me.

I also used binding posts that poke through the case and have a long stud on the inside.  I mounted a little crimped eyelt to one end of the diode strings and was able to slide it over the post.  But the other side needed a piece of wire soldered to it so I had maneuvering room to slide a crimped eyelet to the other end of the wires and slip it over the post after the diode array was seated all the way don in the box.

Keep the diode array as low as you can, because the buttons stick further below the lid top than you first think. Or you can do what John did and actually mount the buttons to the "bottom" of the box so the diode array and the buttons are more or less in the same plane.  But My box was not wide enough to do that, so I had to put them on the lid and then use wire nuts to connect them to the array before closing up the box.

Good luck!

Last edited by Jeff_the_Coaster_Guy

Well, this has me stumped. Box just stopped working.  Now, whenever I go to start any of my engines with the box connected, I get the 2 dings like normal, I hear the startup sound of the engine, then after another second or so, another ding and all sound stops.  I can advance into forward but the horn blows constantly with no way to stop it.  No reaction to the bell button at all.

I thought maybe I burned out a diode or something, so I took apart the box and rebuilt it with all new diodes...still no luck.

Nothing's changed...what could it be?  Dirty track (cleaned with alcohol, and it's a tiny little 027 loop so I'm not losing much voltage around the track).  Different behavior of my 110AC source in my house ?

I tried backing off to 5 diodes in the series, but that won't work.  I have to go all the way to only 1 diode before the engine will even start properly.

So odd...

 

Add Reply

Post
×
×
×
×
Link copied to your clipboard.
×
×