Trains.com

Subscriber & Member Login

Login, or register today to interact in our online community, comment on articles, receive our newsletter, manage your account online and more!

Directional Constant Brightness Using Incandescent Lamps

2365 views
67 replies
1 rating 2 rating 3 rating 4 rating 5 rating
  • Member since
    September 2003
  • 17,140 posts
Posted by Overmod on Thursday, July 22, 2021 2:11 PM

gregc
or do you mean any approach that puts full voltage for a fraction of each cycle, a pulse.   if the pulse is 10% of the cycle, the average voltage is 10% of full voltage.  such a pulse is not likely to look like a square wave except at 50%.

Take the recent Revolution Trackside TE as an example.  This uses a nominal 22.96kHz frequency, and of course regulates 'output voltage' with pulse-width modulation (they refer to the output functionality as pulse-width control but it's essentially the same thing).  

You can realize very quickly that at the necessary pulse duration the risetime (and decay) rates must be quick in order to achieve a stable low voltage output with peak voltage (for charging capacitors and the like) nominally high.  So I don't find it surprising that any high-frequency source, whether in a switching power supply or a regulator, has very clean near-square-wave pulse shape.up to whatever DC driving voltage is being supplied to the switching unit.

DCC is a somewhat anomalous case, as the pulsetrains are alternately switched in polarity to reduce charge carried across the wire (as I first saw in the constellations of V32bis and faster modems back in the Bronze Age of Internet connectivity and if you scoped it would look like a 28V rail-to-rail swing, although in practice it isn't.  I think that is at a much higher clock frequency than any hobby-power PWM design, though.

  • Member since
    January 2009
  • From: Maryland
  • 10,959 posts
Posted by ATLANTIC CENTRAL on Thursday, July 22, 2021 2:13 PM

Greg, see Overmod's info about the Train Engineer Throttle, the version I use is the 27MHz radio, ten channel trackside model working at a frequency of 22.96kHz.

Here is another photo for those not paying attention.

 

The handheld talks to the base station at 27MHz, the base station and power supply act like a power pack and are connected to the track with that old fashioned stuff called wire. Or in my case, they feed a moderately complex system of buttons and relays that assign the correct throttle to the correct track sections with very minimal user input - typically less input than throwing turnouts with a DCC throttle.

At that frequency the on time of each pulse increases as the throttle speed is increased. On and off spikes are nearly vertical on a scope and the top is flat at full voltage even at the lowest speed setting. It is truely switched on and switched off. It is pure DC, there is no phase shift. It is on, or off.

At the highest speed setting the "off" time is virtually zero, less than 10% of the frequency rate.

BUT there is never any filter to smooth out the pulses, that would defeat the purpose.

Aristo included a switch, as Overmod described, which introduced a filter on the output side, for those with coreless motors or who had other concerns about PWM - most such concerns are unfounded.

Aristo recommended the cleanest posible power. The Train Engineer is just a throttle, it did not come with a power supply. The version I use is rated for up to 10 amps at minimum input voltage of 12 volts and a max of 24 volts.

I use 13.8 volt filtered and regulated power supplies designed to simulate battery power for automotive electronics.

This page has some good info on understanding PWM.

http://www3.sympatico.ca/kstapleton3/851.HTM

I was installing these kinds of motor controllers on assembly line equipment 40 years ago, nothing new. In fact, they used DC motors because they could get this kind of fine motor control like we want for our trains.

I have never put a scope on a DCC decoder, but my understanding is the motor control output is similar - full voltage square wave pulse width modulated.

So, I guess the little lamps, or LED's, are flashing at 22.96kHz, but they look like the are steady on, to everyone who has ever watched a train powered by an Aristo Train Engineer throttle.

Sheldon

 

    

  • Member since
    January 2009
  • From: Maryland
  • 10,959 posts
Posted by ATLANTIC CENTRAL on Thursday, July 22, 2021 2:54 PM

Here is a link to a web page with considerable info on the Aristo TE products.

As you look at this info, the trackside unit I use is the one refered to as ART-5471.

http://www.trainweb.org/girr/tips/tips3/pwc_tips.html

http://www.girr.org/girr/tips/tips1/te_programming.html

Sheldon

    

  • Member since
    July 2009
  • From: somerset, nj
  • 3,641 posts
Posted by gregc on Thursday, July 22, 2021 3:09 PM

ATLANTIC CENTRAL
So, I guess the little lamps, or LED's, are flashing at 7.9kHz, but they look like the are steady on, to everyone who has ever watched a train powered by an Aristo Train Engineer throttle

guess ?   

i don't know what the Aristo throttle waveform looks like.   A common approach mentioned by Overmod is to put an AC voltage on the track.

DCC supports DC locomotives by pulse stretching, extending the time a pulse is positive or negative so that the average voltage is not zero and enough to power a DC locomotive.   maybe the Aristo throttle is doing something.

 

a common approach to control the brightness of LEDs is to vary the duty cycle (% on) of a PWM signal.  

there are a variety of ways of implementing PWM.  the most common is a constant frequency that varies the width of the on time from 0-100% of the period.

another approach i've worked with is to either generate a single on pulse for one cycle and N off pulses or a single off pulse and N on pulses.   this sounds similar to the method you describe by varying the frequency of the pulses.

what matters is the area under the on pulse.   by varying the start time of the AC cycle a dimmer varies the area under the AC waveform.    the waveform doesn't have to be "square.

the various PWM approaches i've seen are used to vary the average DC voltage.

the output power from a power supply turning the voltage on for 40% of the time is only 40%.  how could a lamp look fully bright getting < half the power?

 

guess?

greg - Philadelphia & Reading / Reading

  • Member since
    January 2009
  • From: Maryland
  • 10,959 posts
Posted by ATLANTIC CENTRAL on Thursday, July 22, 2021 3:41 PM

Did you look at the wave form graphs on any of the links I posted?

The frequency of the pulse does not vary, the on time varies. The voltage does not vary, every pulse is full voltage.

Why do you keep talking about AC? There is nothing about this that relates to AC at all.

It is pure DC that is switched on and off. When it is on, it is full voltage, when it is off, it is ZERO voltage.

Inductive loads like motors see the average power, resistive loads like lamps simply come on full brightness during the on time.

Can you see 23 thousand flashes a second? We can't even see the 60 cycle phase shift in a 75 watt bulb.

But the advantage for the motor is that the full voltage spike breaks any armature resistance, and provides stall free stable speed control and highest possible torque.

In industry this is 40 plus year old technology. The Aristo Train Engineer was developed in the 90's.

So I get it, because you don't understand, I must be wrong. I have stuff to do.

Drive down here, and I will set up a test track..........

Sheldon

    

  • Member since
    February 2007
  • From: Christiana, TN
  • 1,750 posts
Posted by CSX Robert on Thursday, July 22, 2021 3:49 PM

gregc
the output power from a power supply turning the voltage on for 40% of the time is only 40%.  how could a lamp look fully bright getting < half the power?

Well, for one, the lamp is not using all of the available power.

  • Member since
    July 2009
  • From: somerset, nj
  • 3,641 posts
Posted by gregc on Thursday, July 22, 2021 3:59 PM

ATLANTIC CENTRAL
So I get it, becuase you don't understand, I must be wrong.

you're drawing the wrong conclusion.

i'm looking for an explanation and trying to understand your perspective.

 

gregc
however, the resistance of incandescent bulbs increases as the bulb gets hot and i assume temperature and resistance will be  lower (i.e. more current) even with partial voltage

In the case of incandescent lamps, the effect can be even more pronounced. At low filament temperatures, the resistance of the filament is lower than when the lamp is full bright. Therefore it draws even more current until it heats up. This causes a lamp to brighten up even more when pulsed at less than 100% duty factor.

incandescent lamps have different characteristics than LEDs.

so an incandescent lamp will appear brighter than you would expect at a lower effective voltage because it's not as hot and draws more current during the times the voltage is on.

greg - Philadelphia & Reading / Reading

  • Member since
    January 2009
  • From: Maryland
  • 10,959 posts
Posted by ATLANTIC CENTRAL on Thursday, July 22, 2021 4:07 PM

gregc

 

 
ATLANTIC CENTRAL
So I get it, becuase you don't understand, I must be wrong.

 

you're drawing the wrong conclusion.

i'm looking for an explanation and trying to understand your perspective.

 

 

 
gregc
however, the resistance of incandescent bulbs increases as the bulb gets hot and i assume temperature and resistance will be  lower (i.e. more current) even with partial voltage

 

In the case of incandescent lamps, the effect can be even more pronounced. At low filament temperatures, the resistance of the filament is lower than when the lamp is full bright. Therefore it draws even more current until it heats up. This causes a lamp to brighten up even more when pulsed at less than 100% duty factor.

incandescent lamps have different characteristics than LEDs.

so an incandescent lamp will appear brighter than you would expect at a lower effective voltage because it's not as hot and draws more current during the times the voltage is on.

 

Not disagreeing with any of that.

And remember all these lights, incandescent or LED, are in some sort of directional/constant brightness circuit. But not all such circuits are same, yet the results are amazingly similar.

Bachmann, Proto2000 (with different circuits from different eras), Intermountain, Genesis, others, and few home built ones, all respond more or less the same. 

Pulse on, but not enough to start motor, lamps come on, at or very near full brightness. 

Pulse rate increases, motor starts, light still same brightness. 

Maybe by half throttle there is some very small increase in brightness.

Sheldon

    

  • Member since
    February 2007
  • From: Christiana, TN
  • 1,750 posts
Posted by CSX Robert on Thursday, July 22, 2021 4:28 PM

gregc
so an incandescent lamp will appear brighter than you would expect at a lower effective voltage because it's not as hot and draws more current during the times the voltage is on.

You can actually see this effect when you turn an incandescent lamp on.  It will come on brighter than normal and then quickly dim slightly as the filament heats up.  This is also why when an incandescent bulb blows it is almost always right as you turn it on.

  • Member since
    July 2009
  • From: somerset, nj
  • 3,641 posts
Posted by gregc on Thursday, July 22, 2021 4:43 PM

yes, recently helped someone figure out why their frog juicer didn't seem to work correctly when they checked the polarity with an incandescent voltage probe.   the probe drew too much current before the lamp brightened causing the juicer to reverser polarity.

greg - Philadelphia & Reading / Reading

  • Member since
    September 2003
  • 17,140 posts
Posted by Overmod on Thursday, July 22, 2021 5:04 PM

gregc
i don't know what the Aristo throttle waveform looks like.

Graphs of it are on the Sumida Crossing site's discussion of the Train Engineer system, into a number of different types of load.  As I recall there are some other sites that show it.

One issue that hasn't come up for discussion is that BEMF was not something that the Train Engineer system did the last time I read about it.  There are two potential ways that a high-frequency PWM system can read armature back EMF: one of which is to try to read between pulses and account for any 'smoothing' voltage present in the interval, another is to pause the pulsetrain for some interval and read the back EMF when the forward voltage has decayed.

The last time I looked at decoder PWM, the high frequencies could range up to 43kHz; I believe Sheldon has experience with considerably higher but probably not with particularly small motors and certainly not with inexpensive ones.  In my opinion the high frequencies are for use with brushed coreless motors, where eddy current losses are minimized and resistive losses in the copper reduced as the actual voltage delta becomes very small between pulses.

  • Member since
    December 2008
  • From: Heart of Georgia
  • 4,446 posts
Posted by Doughless on Thursday, July 22, 2021 5:09 PM

ATLANTIC CENTRAL
And remember all these lights, incandescent or LED, are in some sort of directional/constant brightness circuit. But not all such circuits are same, yet the results are amazingly similar. Bachmann, Proto2000 (with different circuits from different eras), Intermountain, Genesis, others, and few home built ones, all respond more or less the same.  Pulse on, but not enough to start motor, lamps come on, at or very near full brightness.  Pulse rate increases, motor starts, light still same brightness.  Maybe by half throttle there is some very small increase in brightness. Sheldon

Your explanations are consistent with my observations and explain to me clearly what was behind what I was seeing.

Different light boards do have slight nuances depending upon brand or vintage but do not substantially change the performance of the lighting when using the TE, to a degree that I would call them "different"

Two that stood out in my experience.  The ATLAS RS32 with LED lighting actually displayed at half brightness on the lowest speed tap before illuminating brightly therafter.  I had never seen an LED illuminate halfway, and thought they could only be on or off.

And a loco equipped with the old 2 function Atlas yellow-green decoder, where you manually place the plug into the DC position, would flash or strobe its iincandecent headlights on the low speed taps before constantly illuminating as the throttle progressed.  Full on bright, then off, not dim, at regular flashing intervals.  An interesting effect.  Obviously, the circuitry in the DC side of that basic early dual mode decoder was a bit different than most light boards.

Something for the technicians to think about.

- Douglas

  • Member since
    September 2003
  • 17,140 posts
Posted by Overmod on Thursday, July 22, 2021 5:39 PM

Keep in mind that any incandescent bulb, once it starts glowing, isn't going to be 'turning on and off' as the power pulses.  It will glow more or less brightly for slow pulses or short pulses at longer spacing (neither of which are characteristic of modern PWM).  Compare the behavior of an incandescent bulb with some of the cheap commercial dimmers, where you have to turn the device 'up' to relatively bright level to get it to illuminate, but you can turn it down to an orange glow... which doesn't visibly flicker on and off.  If it were to shut off, even for a short time, I suspect the resistance would increase and you'd have to turn it up again to get it to illuminate...

Some LEDs are made with secondary phosphors to reduce irritating flicker, on the same principle as early computer monitors with longer-persistence screen phosphors, or to change the emitted color spectrum.  Some of these may visibly 'shine' a bit after you've turned them off.  These too shouldn't show flicker at modern PWM modulations.

I'm nearsighted and suffer terrible flicker from short-phosphor computer CRTs driven at 60Hz, but have little bad effect even as low as 65Hz and none visible at 72 (which is where the 1024P24 initiative with motion-vector-steered frame tripling for movie rendition came from).  Perception of light flicker is probably in the same frequency range...

Just to keep the discussion on point, the OP specifically mentioned that this is for DC, which makes the Train Engineer relevant but DCC decoders less so.  I don't think he has mentioned what his power equipment is, but that might govern the specific circuit(s) he uses. 

  • Member since
    February 2007
  • From: Christiana, TN
  • 1,750 posts
Posted by CSX Robert on Thursday, July 22, 2021 9:07 PM

ATLANTIC CENTRAL
And remember all these lights, incandescent or LED, are in some sort of directional/constant brightness circuit.

Well, not all of them, at least not in N-scale.  I've got several lcoomotives from the 90's and possibly into the early 2000's where the light is simply wired directly to the pickups in parallel with the motor.  That may be the difference in the one I mentioned earlier where there is a significant change in brightness throughout the speed range of the motor when wired to the decoder output.  The light does come on before the motor starts turning, and it's output is not linear - it gets brighter at a faster rate than the motor speeds up - but it's still a significant difference from one end of the range to the other.  Adding a directinal diode and a capacitor, which all of my most modern non-DCC locos have, would help a lot.  The lamps (or LEDs on most of the more modern ones) do not draw much current and would allow the cap to reach close to full voltage during the "on" time without discharging it too much during the "off" time.

  • Member since
    February 2007
  • From: Christiana, TN
  • 1,750 posts
Posted by CSX Robert on Thursday, July 22, 2021 9:25 PM

Doughless
And a loco equipped with the old 2 function Atlas yellow-green decoder, where you manually place the plug into the DC position, would flash or strobe its iincandecent headlights on the low speed taps before constantly illuminating as the throttle progressed.  Full on bright, then off, not dim, at regular flashing intervals.  An interesting effect.  Obviously, the circuitry in the DC side of that basic early dual mode decoder was a bit different than most light boards. Something for the technicians to think about.

On those locos, all the jumper did was switch the motor from the decoder output to a direct connection to the rails.  The lighting was still driven by the decoder, and since decoders in general don't like pulse power it's no wonder the lighting acted funny.  I suspect one of two things were happening, and it could be a combination of both.  If the decoder shuts down during the "off" cylcle, the lighting will naturally turn off as well, so the decoder could just be powering up and down, turning the light on and off as it does.  The other possibility could happen if the capacitor on the decoder is enough to keep the decoder powered up during the "off" cycles.  The decoder has to sample the rail voltage to know which direction the loco is traveling and hence which light to illuminate.  It's sampling could be catching the "on" cycle at times and the "off" cycle at other times, causing the decoder to swicth the light on and off.

  • Member since
    December 2008
  • From: Heart of Georgia
  • 4,446 posts
Posted by Doughless on Thursday, July 22, 2021 9:32 PM

CSX Robert

 

 
Doughless
And a loco equipped with the old 2 function Atlas yellow-green decoder, where you manually place the plug into the DC position, would flash or strobe its iincandecent headlights on the low speed taps before constantly illuminating as the throttle progressed.  Full on bright, then off, not dim, at regular flashing intervals.  An interesting effect.  Obviously, the circuitry in the DC side of that basic early dual mode decoder was a bit different than most light boards. Something for the technicians to think about.

 

On those locos, all the jumper did was switch the motor from the decoder output to a direct connection to the rails.  The lighting was still driven by the decoder, and since decoders in general don't like pulse power it's no wonder the lighting acted funny.  I suspect one of two things were happening, and it could be a combination of both.  If the decoder shuts down during the "off" cylcle, the lighting will naturally turn off as well, so the decoder could just be powering up and down, turning the light on and off as it does.  The other possibility could happen if the capacitor on the decoder is enough to keep the decoder powered up during the "off" cycles.  The decoder has to sample the rail voltage to know which direction the loco is traveling and hence which light to illuminate.  It's sampling could be catching the "on" cycle at times and the "off" cycle at other times, causing the decoder to swicth the light on and off.

 

Thanks.  Those explanations make sense to me.  This happened years ago anyway.  Those decoders are long gone.

- Douglas

  • Member since
    April 2004
  • From: Ontario Canada
  • 3,434 posts
Posted by Mark R. on Thursday, July 22, 2021 11:16 PM

I think the OP threw his hands up in the air and left somewhere back on page one while the rest continue to carry on about something he has no interest in !  Laugh

Mark.

¡ uʍop ǝpısdn sı ǝɹnʇɐuƃıs ʎɯ 'dlǝɥ

  • Member since
    July 2009
  • From: somerset, nj
  • 3,641 posts
Posted by gregc on Friday, July 23, 2021 5:25 AM

sheldon has pointed out in PMs that I like to understand "why" things works.   

so for me, the lengthy discussion helped me learn something about the difference between incadescent lamps and LEDs.   maybe others learned something as well, including the OP

greg - Philadelphia & Reading / Reading

  • Member since
    February 2007
  • From: Christiana, TN
  • 1,750 posts
Posted by CSX Robert on Friday, July 23, 2021 9:10 AM

Mark R.

I think the OP threw his up in the air and left somewhere back on page one while the rest continue to carry on about something he has no interest in !  Laugh

Mark.

 

Yeah, probably, but that happens a lot on this forum (as well as many others), but some good information often comes out of such discussions, and at least this time it stayed related to the original question.

  • Member since
    January 2009
  • From: Maryland
  • 10,959 posts
Posted by ATLANTIC CENTRAL on Friday, July 23, 2021 9:11 AM

OK Greg, here is my total engineering take on what happens.

All these traditional DC lighting circuits, incandesent or LED, have threshold voltages of 1.5 to 2.5 volts.

There is no simple formula to extrapolate the effective average voltage of the PWM signal, but rather it is load dependent. 

A largely resistive load will see a different effective voltage than an inductive load.

So, lets talk about the motor first. One primary feature of high frequency full voltage PWM (without any BEMF) is that while the motor will start at a lower speed, the starting threshold average voltage of the PWM signal is higher than that same motor at the same RPM on pure DC.

The advantage of this translates to higher torque and less chance of stalling from small changes in the initial load.

Once my trains move, they virtually never stall.

So, lets just call the "motor starting voltage" 3-4 volts if we "averaged" the PWM signal.

That is above the threshold voltage of the lighting circuit. So the lighting circuit "works" well before the motor will start in nearly every case.

All that said, I'm not buying a scope, and not doing a bunch of research to write a series of equations......... 

My lights come on before my trains move, without having any little brains in my trains......

My signals work without any "software", my throttles only have five buttons and I don't push any more buttons than the average DCC operator to run a train.

Sheldon

    

  • Member since
    September 2003
  • 17,140 posts
Posted by Overmod on Friday, July 23, 2021 9:35 AM

Let me premise this by noting that the KISS solution for the OP, who is using DC and has not yet responded if he uses pulse or PWM power, was discussed very early in the thread.  In my opinion he should use a combination of clamp diode and bypass resistor to give clamped 1.5V to his light strings: the lights will come on when adequate DC voltage is present but no matter how much the supply voltage increases to, the voltage to the lights never materially exceeds what the strings are designed to use.  Solutions that merely divide down the voltage shared with the motor seem to me to run the risk of sending overvoltage to what may be voltage-sensitive devices.

The discussion of PWM practice in modrl-railroad throttles ought to include the idea that momentary higher-voltage pulses have been long seen as desirable for overcoming 'stiction' in various designs of DC motor or helping apparent slow-speed performance of locomotives that may have balky drivetrains.  I would not be surprised to see both pulse and PWM from 'powerpacks' optimized to produce these... and the resulting "DC" being more or less dirty to devices expecting clean, filtered power.

I have read some discussions on the historical ways some manufacturers used to 'make the magic happen' -- some of which smack to me of expedient or cheap design that happens to assist with slow-speed motor operation.  But I am not surprised to find high peak voltage chosen intentionally for PWM drives, including in pulse shaping for constant-frequency constant-pulse-length systems.

  • Member since
    July 2009
  • From: somerset, nj
  • 3,641 posts
Posted by gregc on Friday, July 23, 2021 11:04 AM

gregc
... but i don't agree with the explanation.

the Aristo PWC Tips page provided the explanation

 

ATLANTIC CENTRAL
All these traditional DC lighting circuits, incandesent or LED, have threshold voltages of 1.5 to 2.5 volts.

diode (LEDs are diode) have a minimum voltage below which they pass very little current.

incandescent bulbs actually have mimimum resistance at the lowest voltage and will pass maximum current when voltage is first applied

 

ATLANTIC CENTRAL
There is no simple formula to extrapolate the effective average voltage of the PWM signal, but rather it is load dependent.

Duty Cycle x High Voltage Level = Average Voltage

 

ATLANTIC CENTRAL
A largely resistive load will see a different effective voltage than an inductive load.

not true, but the current characteristics are different.

for resistive loads,  V = IR

for inductive loads, V = L di/dt, which means that an inductive load will resist a change in current and is the reason for those reverse biased diodes across the switch transistors of motor circuits

the effective voltage is also affected by BEMF which depends on RPM, meaning that they draw maximum current on startup and current increases as speed decreases under heavier load

 

ATLANTIC CENTRAL
So, lets talk about the motor first. One primary feature of high frequency full voltage PWM (without any BEMF) is that while the motor will start at a lower speed, the starting threshold average voltage of the PWM signal is higher than that same motor at the same RPM on pure DC.

The advantage of this translates to higher torque and less chance of stalling from small changes in the initial load.

Once my trains move, they virtually never stall.

high voltage pulses at any frequency help overcome motor stiction and improve low speed performance

Stiction can appear in the motor and drivetrain, as the amount of torque required to get moving is greater than that required to maintain motion. Digital Command Control's multifunction decoders use PWM or Pulse Wave Modulation to drive the motor. By application of full voltage to the motor in pulses at varying time periods, stiction can be overcome and low speed operations made realistic.

 

ATLANTIC CENTRAL
So, lets just call the "motor starting voltage" 3-4 volts if we "averaged" the PWM signal.

That is above the threshold voltage of the lighting circuit. So the lighting circuit "works" well before the motor will start in nearly every case.

again, explained on the Aristo PWC Tips page.   the lamps turn on faster than the motor does

In the case of incandescent lamps, the effect can be even more pronounced. At low filament temperatures, the resistance of the filament is lower than when the lamp is full bright. Therefore it draws even more current until it heats up. This causes a lamp to brighten up even more when pulsed at less than 100% duty factor.

 

ATLANTIC CENTRAL
All that said, I'm not buying a scope, and not doing a bunch of research to write a series of equations......... 

My lights come on before my trains move, without having any little brains in my trains......

My signals work without any "software", my throttles only have five buttons and I don't push any more buttons than the average DCC operator to run a train.

everything you've said about the Aristo Throttle generating pulses causing incandescent lamps to turn on before the motor should also be true for DCC decoders or PWM throttles driving motor with incandescent lamps wired across the motors

greg - Philadelphia & Reading / Reading

  • Member since
    January 2009
  • From: Maryland
  • 10,959 posts
Posted by ATLANTIC CENTRAL on Friday, July 23, 2021 11:28 AM

A few random last thoughts.

Duty cycle x voltage would seem to only define the no load voltage.

Yes, I should have said, VA, amps or watts, regarding the motor and its interaction with the PWM.

Point remains good PWM throttles provide very good control, and have this added feature, largely compatible with most DC constant directional lighting circuits.

Most incandescent lighting circuits rely on diodes, so same rules apply.

I only posted in this thread because the feature the OP desired was something I resolved 20 years ago.......

Sheldon

    

  • Member since
    December 2008
  • From: Heart of Georgia
  • 4,446 posts
Posted by Doughless on Friday, July 23, 2021 12:25 PM

Mark R.

I think the OP threw his up in the air and left somewhere back on page one while the rest continue to carry on about something he has no interest in !  Laugh

Mark.

 

Yes, the OPs question was about achieving constant lighting by building such directional lighting circuits into DC locos, if I summarized correctly.

- Douglas

  • Member since
    January 2009
  • From: Maryland
  • 10,959 posts
Posted by ATLANTIC CENTRAL on Saturday, July 24, 2021 7:44 AM

Mark R.

I think the OP threw his up in the air and left somewhere back on page one while the rest continue to carry on about something he has no interest in !  Laugh

Mark.

 

Maybe so, but the OP's specific question was not just about building a directional constant brightness DC lighting circuit. It was about having that circuit light the lights before the locomotive moves.

In my experiance that is not easily or consistantly achieved with conventional DC throttles.

Yet I have been enjoying the effect he desires for 20 years.

So I offered my solution/experiance/knowledge. Once again only to be challenged by those unfamiliar with the technology I use, just like my knowledge of relay circuits, equalized trucks, turnout geometry, passenger car diaphragms, curvature/easement design and more.

Layout construction begins soon........

Sheldon

    

  • Member since
    January 2017
  • From: Southern Florida Gulf Coast
  • 13,313 posts
Posted by SeeYou190 on Saturday, July 24, 2021 9:05 AM

ATLANTIC CENTRAL
Maybe so, but the OP's specific question was not just about building a directional constant brightness DC lighting circuit. It was about having that circuit light the lights before the locomotive moves.

Constant lighting with battery power that charges from the track was my solution, but that was with steam locomotives with room in the tenders.

Lately, lighting on locomotives means very little to me.

-Kevin

Wink Happily modeling my STRATTON & GILLETTE RAILROAD. A Class A line located in a personal fantasy world of semi-plausible nonsense on Tuesday, August 3rd, 1954.

  • Member since
    January 2009
  • From: Maryland
  • 10,959 posts
Posted by ATLANTIC CENTRAL on Saturday, July 24, 2021 10:21 AM

Well, that works too, but is time consuming and expensive for a large fleet.

I have always looked for solutions that simple, afordable, easy to operate and easy to implement.

Considering my layout goal of presenting the "big picture", I avoid things that are fussy to build, install or operate.

Others may not understand or imbrace these ideas, but that is why my CTC and signaling are simplified, why I still use DC, why I don't have onboard sound, why my throttles only require 5 control buttons.

When it comes to lights, I want them, but I don't want to have to think about them in daily operation.

DC, the Train Engineer, and basic factory DC lighting circuits fit the bill perfectly.

It's just like the Train Engineer itself, great speed control, wireless throttles, easy operation, no decoders to install, no programing, no button sequences to remember.

And before anyone says anything about my complex relay wiring, even with DCC, most of my wiring  (or some similar equivalent) would still be required for signals, CTC and turnout control.

All features that are important to me.

Sheldon

    

  • Member since
    February 2007
  • From: Christiana, TN
  • 1,750 posts
Posted by CSX Robert on Saturday, July 24, 2021 10:43 AM

ATLANTIC CENTRAL
Maybe so, but the OP's specific question was not just about building a directional constant brightness DC lighting circuit. It was about having that circuit light the lights before the locomotive moves. In my experiance that is not easily or consistantly achieved with conventional DC throttles.

Actually, that can be easily and consistantly achieved with a regular DC throttle.  The circuit linked to in the third post in this thread will do it: http://www.mrollins.com/constant.html

That circuit does have it's drawbacks, however.  You have to add it to every locomotive, even ones that you don't care about lighting in if you want them to behave the same way.  That circuit cuts about 2 volts from the motor, so the motor doesn't get anything untilt the throttle passes 2 volts, and the motor gets 2 volts less throuhgout the throttle range including the top end.  Nonetheless, it is a simple circuit that accomplishes what was aked for with any throttle.

  • Member since
    February 2007
  • From: Christiana, TN
  • 1,750 posts
Posted by CSX Robert on Saturday, July 24, 2021 11:04 AM

As I mentioned before, an inandescent lamp wired directly to the motor outptut of a DCC decoder (which outputs a square wave pwm signal) would change noticably in brightness throughout the motors speed range.  I also mentioned that having any directional or constant lighting circuitry in there, especially with capacitors, would probably change that.  I did some more testing with a voltmeter hooked up and confirmed it.

At speed step 1 it showed ~0.9 volts and the light was on but very dim.  As I went up through the speed steps, the voltage went up in a fairly linear fasion and the light got brighter.  It did get brighter quicker than the voltage ramp, as explained byt the "Aristo PWC Tips" article linked to earlier, but again, it was a significant difference.

I added a capacitor across the lamps leads and ran the test again.  At speed step 1 the voltage was over around 11 volts and the lamp was, naturally, at almost full brightness with very little variation left.

  • Member since
    January 2009
  • From: Maryland
  • 10,959 posts
Posted by ATLANTIC CENTRAL on Saturday, July 24, 2021 11:17 AM

CSX Robert

 

 
ATLANTIC CENTRAL
Maybe so, but the OP's specific question was not just about building a directional constant brightness DC lighting circuit. It was about having that circuit light the lights before the locomotive moves. In my experiance that is not easily or consistantly achieved with conventional DC throttles.

 

Actually, that can be easily and consistantly achieved with a regular DC throttle.  The circuit linked to in the third post in this thread will do it: http://www.mrollins.com/constant.html

That circuit does have it's drawbacks, however.  You have to add it to every locomotive, even ones that you don't care about lighting in if you want them to behave the same way.  That circuit cuts about 2 volts from the motor, so the motor doesn't get anything untilt the throttle passes 2 volts, and the motor gets 2 volts less throuhgout the throttle range including the top end.  Nonetheless, it is a simple circuit that accomplishes what was aked for with any throttle.

 

Actually, that is the same circuit used in early Proto locos. Lights do consistantly come right away at full brightness, but with more basic throttles like most MRC power packs, locomotive movement and lighting activation are less predictable and are often are nearly simultaneous.

And very hard to control the effect of turning on the light without moving the loco using the throttle knob.

But, that's just my experiance........

Bigger point is this, I'm getting tired of debating things I have known to be demonstrable fact for decades.

I'm going to the train room now.

Sheldon

 

    

Subscriber & Member Login

Login, or register today to interact in our online community, comment on articles, receive our newsletter, manage your account online and more!

Search the Community

ADVERTISEMENT
ADVERTISEMENT
ADVERTISEMENT
Model Railroader Newsletter See all
Sign up for our FREE e-newsletter and get model railroad news in your inbox!