Skip to main content

Hi guys, not sure what forum to put this on, so I went here....

There is a company on eBay selling an HO scale locomotive that includes "simulated brake spark".

The referenced item can be found here.

There used to be a video of this train, showing the various included systems, including the brake spark during production testing. I couldn't find it and have asked the owner of the store for the link if available. I will post if I get it.

I have seen postings on the OGR forums about simulating the arcing of subways and will be working on a simulation of that for my own layout.

Wondering if anybody has tried simulating brake spark and if so, how they did it.

-Todd

Last edited by Rich Melvin
Original Post

Replies sorted oldest to newest

Generating the brake spark can be as simple and inexpensive as a flickering LED... for example from an LED candle you get at DollarTree for, uhh, $1.  Or there are many pre-built flickering/random blinking circuits used in modeling to simulate campfires, welding, etc.

More interesting is knowing WHEN to trigger or turn on the sparking.  In the example you give, the engine's electronics KNOWS WHEN it is slowing down and hence KNOWS WHEN to turn on the lighting effect.  If attempting to add this effect after-the-fact, I'd be surprised to find any engine electronics that has an output or trigger signal indicating when the engine is braking or rapidly decelerating.  For example, I know that in MTH PS2/3 you can get a brake squeal sound when rapidly decelerating....but I'm fairly certain this event is not available as an electrical trigger signal that can turn on a flickering/sparking LED circuit.

In which case I'd think you need to concoct a method to detect rapid deceleration by somehow monitoring the speed of the engine.  There are many ways to detect deceleration which can be elaborated upon if there's interest.

Rick- Most videos on YouTube are showing brakes sparking from slipping, not braking, but there are a few. I appreciate your comment from real world experience. Nothing from a movie, I just like pushing the envelope on what we can do. Spark from a pantograph and 3rd rail is in the works on my RR, so when I saw the model for sale on eBay with brake spark it got my interest.

John- that's exactly what I was thinking, something simple (or as simple as possible

Stan, I am not there yet, but wanted to see the crowds reaction and if somebody had already done it. John is on the same track as me, to keep it simple, but you bring up the crucial point as to how to trigger it. Any deeper thoughts are certainly appreciated if/when you have time. It's a project I will be looking at after the 3 or 4 right in front of me are done I have seen a handful of guys do spark/arcing from 3rd rail subways and pantographs and that is in the works currently on my RR. I have personally only seen 1 attempt at brake spark, that's why I posted. Thanks.

Wheels brake locked, sparking on the rail is a no-no too; flat spots.

How about a nice hotbox? There are cars that simulate a hotbox bearing.

Sounds andoand smoke too 

Warning: this will fast become the most annoying thing you own. About once or twice a year I fire mine up for the smoke feature; sound off.

You can find "hotbox refers", etc. on Youtube and a few old threads to see what it's all about.

@Adriatic posted:
Warning: this will fast become the most annoying thing you own. About once or twice a year I fire mine up for the smoke feature; sound off.

You can find "hotbox refers", etc. on Youtube and a few old threads to see what it's all about.

I have two of those, and one of my first mods was to add TMCC control.  That allows me to run around and randomly trigger it ones in a while for amusement, but not have it constantly triggering.  I'm going to expand the controls one of these days and add a relay to enable/disable sound by command control as well.  Since I used the ERR Mini-Commander ACC, I have extra outputs, so it's a simple addition.

@jockey31 posted:
..

Stan, I am not there yet, but wanted to see the crowds reaction and if somebody had already done it. John is on the same track as me, to keep it simple, but you bring up the crucial point as to how to trigger it. Any deeper thoughts are certainly appreciated if/when you have time. It's a project I will be looking at after the 3 or 4 right in front of me are done I have seen a handful of guys do spark/arcing from 3rd rail subways and pantographs and that is in the works currently on my RR. I have personally only seen 1 attempt at brake spark, that's why I posted. Thanks.

As you say the crucial point is how to trigger it.  I find it odd that even though the engine in your video has "access" to the trigger (e.g., the brake squealing sound), the sparking continues after the train/squeal stops.   Seems they could have done better.  Also, if going through the effort, it's odd they did not randomize the sparking of the 4 brakes - they all spark in unison which detracts from the effect - in my opinion of course; yes, 4 independent LEDs means 4 independent circuits but they did seem to do a nice job with some of the other lighting effects.

brake spark

Motor voltage can be a simple source of a trigger but can be problematic.  For example, it's not the applied motor voltage from the electronics that you want to measure; it's the back-emf or motor generator voltage which is a far superior indicator of motor speed.  But I'm getting ahead of myself!  When you get closer to actually experimenting with the idea, I'll be interested in continuing the dialog.

Separately, I seem to recall an OGR thread where the idea was to detect when a train is going around a curve...I think for the purpose of triggering flange squeal sound(?).  IIRC when you do the math, it might be possible to use inexpensive solid-state accelerometers to detect change in direction.  That is, an accelerometer or similar inertial sensor could also detect rapid deceleration.

Attachments

Images (1)
  • brake spark

Unless you're providing the motor control, I think measuring the back-EMF may be somewhat difficult, you have to time it for when the motor is not getting drive voltage.  FWIW, while the motor voltage is not an iron-clad indicator, it's a ton easier to use.  Also, if the locomotive has cruise, voltage on the motor will certainly indicate it's moving.  It really depends on the actual locomotive you're trying to add the feature to.

I agree with Stan, the four locations should be independent and random.

As for curves, the Lionel Legacy Vision Line Tank Cars from 2009 had an angle sensor to change the sounds on curves, it was a simple arrangement.  It had a hanging pendulum that swung out on curves and a sensor board under it that triggered the sound board to recognize the curves. It works great on the tank cars, I love the effect and the reaction to curves, really adds a dimension to the sound generation.

Here's some of the pieces used in the tank car.

   

Attachments

Images (3)
  • mceclip0
  • mceclip1
  • mceclip2

It comes down to defining "simple".  Agreed measuring back-emf is more complicated than simply measuring applied motor voltage.  But like many things in life you get what you pay for.  And of course the objective is to detect deceleration...not  speed.  I suggest waiting until the OP gets closer to implementation when we can hear exactly the configuration, command vs. conventional, whether soldering and component-level assembly is on/off the table, Arduino on/off the table, and so on.

As for the hanging pendulum inertial sensor, this would be like those cattle cars that would trigger the cow or chicken sounds upon movement... IIRC, QSI came up with this sometime in the last century.

Untitled

Clearly this would be a simple and inexpensive method to detect deceleration.  The trick would be to only detect deceleration and not acceleration.  In a cattle car trigger one expects sounds upon starting or stopping.  Likewise, if used for curve detection it doesn't matter if "turning" left or right.  So it would take some thought to "tune" the pendulum mechanism or insert an electrical filter circuit to only generate a trigger when it swings in one direction so to speak.  That's the advantage of chip accelerometers as the polarity of the voltage output indicates both the sign and magnitude so that acceleration can be ignored.

 

Attachments

Images (1)
  • Untitled

Well, with DCS I don't see how back-EMF is going to work out, I'll let someone else try to get that working.  If you're measuring motor voltage, you could logically wait to trigger for the voltage to start decreasing and then trigger the effect.

I'd run the motor voltage detection using an isolated power module and power the whole system within that framework.  Probably an A/D input on that Arduino would be the ticket to detect when you were decelerating.

What triggers brake/flange squeal on my TMCC E-33 or railsounds Berk Jr tender?  

Could a sound frequency be a trigger with the aid of filtering so every sound doesn't trigger it.  (I know it could, I'm curious if a Chinese board might exist already. Like the old music detection "dancing" mood lights, with colored lights triggered by various frequencies and volume level changes varying color& intensity.)

( 🤔..RIP Dale M. 🤧)

 

@Adriatic posted:

What triggers brake/flange squeal on my TMCC E-33 or railsounds Berk Jr tender? 

The R2LC serial data comes over and the RS board decodes if you're braking based in the serial data stream.

@Adriatic posted:

Could a sound frequency be a trigger with the aid of filtering so every sound doesn't trigger it.

Could it?  Maybe.  However, that filter would likely change for each type of horn and/or whistle.  Also, I have serious reservations about how reliable that triggering would be.

@gunrunnerjohn posted: ...snip...

As for curves, the Lionel Legacy Vision Line Tank Cars from 2009 had an angle sensor to change the sounds on curves, it was a simple arrangement.  It had a hanging pendulum that swung out on curves and a sensor board under it that triggered the sound board to recognize the curves. It works great on the tank cars, I love the effect and the reaction to curves, really adds a dimension to the sound generation.

Were those the ethanol tankers? If so, I thought the sound was more like someone on the inside of the tank beating on it with a hammer! It did not really bother me as I had (may still have) plans to two-rail them and ditch the sounds. The only drawback that I remember was the EOT flasher circuit board may have been integral to the sound board.

Incidentally, I just got finished working on a PW whistle tender, and the spark effect when the whistle is triggered is almost exactly what that video shows, but with real sparks! I don't know how much it would damage the track over time, but it would be cool to have a ground circuit like that, but skip the whistle and trigger it in one of the ways you guys are mentioning.

I'll try to get a video this evening, but I suspect any PW guys will know exactly what I'm talking about.

I would rather have sparks and cinders flying out of the stack.

A simple proxy might be an upward pointing LED hidden in the smoke stack.  Here is a photo from a previous OGR thread about BLUE smoke (i.e., diesel combustion).

ogr%2520LED%2520blue-smoke

So the idea is to replace the blue LED with flickering orange LED(s).  The fan-driven smoke stream reflects the light providing the illusion that sparks, cinders or whatever detritus are actually spewing/emitting from the stack.

For any determined modelers with time on their hands, I encourage you experiment with illuminating a fan-driven smoke unit's stream with strategically placed tiny surface-mount LED(s) near the output.  For example, here's a 1:48 scale model F-14 Tomcat jet engine LED-illuminated after-burner with and without smoke:

ogr%2520smoke%2520enhances%2520led%2520dramatically

And here's an orange LED hidden in the tip of the 88 gun on a King Tiger tank.  Flame burst effect is much more dramatic in the dark, but the idea is to use the exhausting smoke puff to reflect the LED flash which greatly multiplies the illusion.

led in tank turret

 

 

Attachments

Images (3)
  • ogr%2520LED%2520blue-smoke
  • ogr%2520smoke%2520enhances%2520led%2520dramatically
  • led in tank turret
@jockey31 posted:

You guys are awesome! I love the enthusiasm about new ideas. I am a ways off from implementation, clearly, but the ideas are great.

To answer the posed questions;

-Control is DCS

-Soldering Yes

-Arduino (would need help)

Again, since this is not today's project, I think there's value in letting the ideas stew and perhaps others will add to the discussion.

Knowing we're dealing with command-control DCS is important.  Specifically, DCS uses pulse-width modulated voltage to drive a DC-can motor.  The motor pulses are ~20V, thousands of times per second.  If you're not already intimately familiar with the Arduino or a similar microcontroller platform, I don't think that would be the easy path forward. 

Likewise, as GRJ implies, extracting speed by measuring the back-emf motor voltage (NOT the applied motor voltage) is not the easy path forward.  To summarize, measuring the back-emf voltage requires intermittently cutting the path between the motor controller voltage and the motor itself.  When so isolated, the two motor terminals will generate a DC voltage proportional to the motor speed.  This can all be performed in, say, 1/100th of a second for a typical DC-can motor that might be used in O-gauge.  So it could be done but not an easy path forward.

That said, as GRJ suggested early-on, using applied motor voltage is probably the easiest-to-implement proxy for engine speed.  Two end-point examples illustrate why applied voltage is only a proxy for what the engine is actually doing.  1) If the engine is stalled, then applying 20V DC or whatever is clearly a bad proxy for how fast the engine is going.  Conversely, 2) If the engine is cruising along at some healthy clip, then cutting the applied voltage to 0 is also a bad proxy for how fast the engine is going as it coasts down to 0 speed over several seconds.

Detecting drops in applied motor voltage to trigger braking is also problematic especially with command-cruise control.  For example, say the engine is climbing a grade and requires 12V.  Then it reaches the summit and now requires only 10V to maintain the command-control speed.  So the applied motor voltage suddenly drops but clearly this should not trigger the braking effect.  Likewise an command-control engine going around an O-22 curve then reaching a straightaway will see a sudden drop in applied motor voltage.  Again, this event should not trigger the braking effect.

Sidebar.  In re-watching the video of the HO engine sparking, it seems to me that you should not have sparking when the engine decelerates to below, say, 10 sMPH.  That is, do brakes really spark at slow speeds?  Now that I think about it, I believe MTH PS2/3 engines do NOT play the brake squealing sound when the engine is going really slowly and stops.

Back to the matter at hand.  I'm thinking maybe a dozen inexpensive components (less than $5 total) could make a suitable trigger.  A 10-cent bridge rectifier converts the applied motor voltage to DC.  You need this since the train can be going forward or backward so the bridge converts the motor voltage to DC.  The DC motor voltage is pulsing so you need to smooth it out since dealing with PS2/3 20V motor pulses is electrically challenging.  My idea is to smooth out these pulses with a "fast" filter and a "slow" filter.  A filter simply converts the motor pulses to a smooth DC voltage that is easier to process.  A filter can be as simple as a 10-cent resistor and capacitor.  Here's the trick.  And this may mean absolutely nothing to you...but will serve as a reminder to me next year or whenever you re-visit this project.  The idea is to detect when the short-term motor voltage drops by some settable number of volts below the long-term motor voltage.  There are many simple methods to compare two voltage and detect when the difference exceeds some level...maybe 50 cents in parts.  This voltage drop is the trigger to the flickering LED circuit(s).  

flicker led

That you can get a flickering LED on eBay for 30 cents each from a US seller tells me this part of the project should not be a problem! 

Again, a lot of techno-babble above but serves as a thought-diary that I can refer to later...

 

 

Attachments

Images (1)
  • flicker led

There is another possibility, likely the one I'd pick if I wanted to do this.  I'd repurpose my Chuff-Generator to output the signals.  It reads the tach strip, thus certainly knowing exactly what speed the motor is going, and it has two FET outputs that can support plenty of current to manage the lights, and a couple more PIC processor pins that could drive a flickering LED with no problem.  It's so much easier to do stuff like this if you have a little intelligence in the circuit.

Attachments

Images (1)
  • mceclip0

... and it has two FET outputs that can support plenty of current to manage the lights, and a couple more PIC processor pins that could drive a flickering LED with no problem.  It's so much easier to do stuff like this if you have a little intelligence in the circuit.

Only 2 FET outputs?!   How about 4 FET outputs! 

4 FET outputs

But seriously though; GRJ has it right.  IF you can find the self-described "needed help" to go down the Arduino or microcontroller path, many doors open. 

Like his Chuff Generator, above custom-built board uses a custom-programmed 8-pin microcontroller IC chip (less than $1).  I figure the brake sparker could use up to 4 independent flickering LEDs; those would take 4 of the 8 pins.  You need 2 pins for DC power.  So that leaves 2 pins, one of which can sense the flywheel striping on MTH PS2/3 engines like the Chuff Generator.

As GRJ says, this gets you an accurate and real-time measurement of actual engine speed rather than some proxy (e.g., speed derived from applied motor voltage).  Then literally do-the-math to calculate changes in speed...and when the deceleration exceeds some threshold then independently flicker the 4 LEDS.

Here's the 4 LED board programmed to independently flicker 4 LEDs that simulate the 4 50-cal guns on a 1:48 scale model of a P-38 Lightning.  Like the brake sparking application, the LED action is synchronized to an external event...in this case when Major Bong's Marge comes in for a strafing run.

If just getting started into microcontrollers I suppose the Arduino might be the way to go.  Not sure what to suggest.  That is, something like the Arduino-Nano ought to fit in a O-gauge diesel shell.  Might still want to add FET output buffers to insure adequate current drive to flickering LEDs...for example you might be driving multiple LEDs on each of the 4 outputs if sparking is seen on either side of the engine.

Attachments

Images (1)
  • 4 FET outputs
@stan2004 posted:

Only 2 FET outputs?!   How about 4 FET outputs! 

If I were building a custom board, then I'd have four. However, I have an existing board that already knows how to read the stripes on the flywheel, so I figured the path of least resistance would be to "repurpose" that board. Obviously, a new layout could have four FET's, I wouldn't need my cal jumper or the indicator LED that I use to prompt calibration and tape testing.

With the Arduino Nano, you could drive separate LED's from outputs, you have a ton of outputs to use.  That way you could have one LED for each output and not exceed limits.  However, I don't know about total power dissipation, never tried to drive a ton of stuff from the Arduino, especially the Nano with the smaller footprint uP.

Last edited by gunrunnerjohn

Add Reply

Post

OGR Publishing, Inc., 1310 Eastside Centre Ct, Suite 6, Mountain Home, AR 72653
800-980-OGRR (6477)
www.ogaugerr.com

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