Doubtful circuit department

I am guessing, because I don't have context to understand what problem is trying to be solved here, but are the motors being fed by opposite polarity?  Wouldn't they run in different directions?  I don't really understand how they are powering the center rail with two separate transformers this way.  Wouldn't the two diodes create an out of phase condition, since they are installed opposite for each?

Interested in hearing the correct technical answer...

George

Am I wrong in saying that diodes only work with DC current ?   In my simple mind, the only way to control two trains on the same track without DCC or something similar, would be to isolate the outer rails.  Then the wheels of the loco would have to be isolated from one another.  After all of that, there are other problems to overcome.  I would think turnouts would be an issue with the method I described.  

The simplest solution would be to run electrics under catenary and diesels by the center rail.  But not everyone wants or likes electrics.  Oh well.  

My understanding is that the diode will convert the AC current to DC in a single phase. The direction of the diode will determine which half of the AC wave is chopped. Since the diodes are applied to two transformers connected with the same common, I assume they are in phase. Therefore, each leg will have DC but I believe they will have different polarity. I would think this would run the motors in different directions if they were on separate tracks, but on the same track, I would think you would have a short, and the breakers would throw. Also, with diodes on the motors, you couldn't reverse direction. Hope my High School level electronics knowledge isn't making my post into mumbo jumbo.

Bob Nelson:

"I'll describe the oddball scheme that I use (in a simplified form).

Install a rectifier diode in each of your two locomotives, between the pickup and everything else.  Make the two diodes point in opposite directions between the two locomotives.  Put a 5000 microfarad electrolytic capacitor in parallel with all the locomotive circuitry downstream of the rectifier.

Use a single transformer, preferably one with a fairly high output voltage, like a type Z.  Connect two Lionel rheostats, each with a rectifier diode in series with it, in parallel and put that combination in series with the single transformer output, turned up all the way.  Control the two trains with the rheostats.

Any whistles or horns will sound continuously with this scheme; so you'll have to disable them."

Yes, Bob Nelson's scheme will work, but the one in the Popular Science article will not.

Why?

Is it the lack of capacitors?  Without the capacitors, I would guess the motors would not maintain current during the phase of the chopped wave.

 I would think you would have a short, and the breakers would throw.

Maybe I am missing something.
It looks to me like the two diodes connected to the "A" and "D" terminals of a ZW transformer would create a short between the two posts anytime the handles were set to different on positions. The internal circuit breaker would not trip because it does not protect any circuits that do not involve the "U" (common) terminals.
This sort of "accidental" circuit is one of the reasons I put a circuit breaker as the first item on each terminal, "A", "B", "C" and "D".

C W Burfle posted:

 I would think you would have a short, and the breakers would throw.

Maybe I am missing something.
It looks to me like the two diodes connected to the "A" and "D" terminals of a ZW transformer would create a short between the two posts anytime the handles were set to different on positions. The internal circuit breaker would not trip because it does not protect any circuits that do not involve the "U" (common) terminals.

Bingo!

Yes, the circuit as shown will cause a short-circuit on one half of the AC waveform if two outputs of the transformer are not equal. And, of course, if the two outputs ARE equal, the circuit serves no purpose, since both motors will see the same voltage anyway.

Silver Star for C. W. Burfle!

Please educate me a little further.  My understanding is that a 'short' is a circuit where there is little to no resistance, and both outputs of the power supply are connected directly.  The current will increase rapidly until the power supply fails or the circuit is broken (usually through heat and melting).  So what causes it in this situation and why doesn't it happen if the handles are in the same position? 

The handles at the same position create the same voltage. So in this scenario, the common will receive a sine wave of power with the same magnitude and it would be in phase.  Correct?  If the handles are in different positions, the voltages are different and the magnitudes of the sine wave would be different for different ends of the wave.  What happens here?

The handles at the same position create the same voltage. So in this scenario, the common will receive a sine wave of power with the same magnitude and it would be in phase.  Correct?  If the handles are in different positions, the voltages are different and the magnitudes of the sine wave would be different for different ends of the wave.  What happens here?

Having the handles at two different positions means that a portion of the secondary coil is in a circuit from post a, through the diodes and back through post b. The further the handles are apart, the more secondary coil windings will be in the circuit, and the higher the voltage.

 

I think it's pretty straightforward: each motor gets half of a sine wave of AC power, which essentially looks like pulsed DC power to each motor. One motor gets the positive half and the other motor gets the negative half of each waveform. The throttles are not "shorting" to each other through the diodes because they are in phase. The limitation is that the motors do not get full power on half waves.

Lionel-type universal motors work on either AC or DC. The e-units will still cycle when power is interrupted. If this was used for DC trains there would be no reversing capability.

I think there used to be other ideas of 3-rail dual-train control that used insulated running rails, which would not allow reverse loops. Some early modelers combined outside third rail or overhead wire for independent train control.

 The throttles are not "shorting" to each other through the diodes because they are in phase.

The bar of diode (CR-1) is connected to the triangle of diode (CR-1). That means that current will flow. There is a short. In order for the current to be blocked, the connection would have to be from triangle to triangle, or bar to bar.

They are in effect wired in series with matching polarity.

C W Burfle posted:

 The throttles are not "shorting" to each other through the diodes because they are in phase.

The bar of diode (CR-1) is connected to the triangle of diode (CR-1). That means that current will flow. There is a short. In order for the current to be blocked, the connection would have to be from triangle to triangle, or bar to bar.

They are in effect wired in series with matching polarity.

I see that, but there is no current flow because the throttle outputs are in phase. I think Popular Science would have done their homework before publishing this. Someone should actually try it if they want to prove me wrong.

Draw out a sine wave diagram and you will see if it works or not.

 

I see that, but there is no current flow because the throttle outputs are in phase. I think Popular Science would have done their homework before publishing this. Someone should actually try it if they want to prove me wrong.

If that was so, then connecting a plain wire across the "A" and "D" terminals would not create a short circuit either.

Been there, done that, you get a nice fat spark.

While it would not work on one transformer, this idea seems plausible if you use two separate transformers.  Of course you still have the problem with the whistle. If both throttles are not set the same the whistle/bell will sound.  

JGL

"If that was so, then connecting a plain wire across the "A" and "D" terminals would not create a short circuit either. Been there, done that, you get a nice fat spark."

OK. This is just not true on a post-war ZW (if the handles are set the same). On a ZW-C, there was an issue with some PH 180 bricks being out of phase. You could check by putting a bulb across the A and D terminals. If the bulb lit, then the bricks were out of phase. If the bulb did not light, then the bricks were in phase. If the bulb won't light, you can put a piece of wire across them and there will be no big spark. This is the same as a pickup roller bridging two blocks.

Ah, I just read about voltage differential. If the voltage of A and D are the same, there is no voltage differential. If the two handles are set at different voltages, then there would be a voltage differential and current would flow.

C W Burfle posted:

The handles at the same position create the same voltage. So in this scenario, the common will receive a sine wave of power with the same magnitude and it would be in phase.  Correct?  If the handles are in different positions, the voltages are different and the magnitudes of the sine wave would be different for different ends of the wave.  What happens here?

Having the handles at two different positions means that a portion of the secondary coil is in a circuit from post a, through the diodes and back through post b. The further the handles are apart, the more secondary coil windings will be in the circuit, and the higher the voltage.

 

Now I understand what this means.  

So, if one handle is set to 18 volts and the other handle is set to 8 volts, is that a 10 volt differential?  If the breaker or fuse is on the common, what part of the circuit would fail first (due to heat) and how long would it take to fail?  I ask this to understand if this circuit could work for a while until something overheated.

George S posted:

...what part of the circuit would fail first (due to heat) and how long would it take to fail? 

The secondary windings between the two taps/rollers.

ADCX Rob posted:

George S posted:

...what part of the circuit would fail first (due to heat) and how long would it take to fail? 

The secondary windings between the two taps/rollers.

OK, this is fascinating.  I hope you guys don't think I am hijacking this thread...

I agree that the secondary windings would get hot and fail.  However, let's say the voltage differential was only 2 volts with the trains running at 9v and 7v. (BTW, I believe the math is much more complicated than what I am posing on an AC circuit.)  That coil is built to handle an 18 volt load at 8 to 10 amps. Yes, it will heat up, but how long before it fails?  My assumption is that you could run this setup for a fair amount of time before the secondary winding failed and ruined your transformer (and maybe started a fire).  Maybe that is why people think it works?  Thoughts?

On a test with a ZW, with no other load and a difference of 5 volts between throttle settings, the current between A & D(or any two throttles) is only 5 amps, but by the time the voltage setting got to 6 volts difference, the needle was jumping off the scale at well over 20 amps, a death knell pending... something's got to give.

But on the original question as to whether it will work, you need to draw a sine wave diagram. 

What will throttle 1, alone, look like if viewed on a scope?

Then throttle 2 alone?

Then if both are raised at the same time? What does the sine wave on the track look like as you increase the voltage settings?

Now leave 2 at 16 volts and lower 1 to 8 volts... What does the scope show? Do the two sine waves(upper for one, lower for the other) buck each other at any point?

The secondary windings between the two taps/rollers.

Years back, a fellow in our train club had the secondary windings in his ZW ruined because the "A" and "D" terminals were accidentally connected to the same block of a display layout. Only one control was using during the day, but the other handle was left in an "on" position.

But on the original question as to whether it will work, you need to draw a sine wave diagram. 

The "spark test" says it won't. You'll just create a short.

 

C W Burfle posted:

But on the original question as to whether it will work, you need to draw a sine wave diagram. 

The "spark test" says it won't. You'll just create a short.

 

But the breaker won't blow, so the circuit could run for a while.  I believe, as Rob says, it depends.  If the trains are running at close to the same speed, the circuit may not get hot.  If one train is running and the other is not, it could get very hot, very quickly.  

You are right in the end, but someone could probably run a quick test of this to prove to themselves it would work and not catch the flaw. Also, the load will use some of the over current, and the trains may run faster than expected (like my light bulb example).  I wouldn't want to test this on my ZW.

As long as there were trains on the track and neither control set to zero, it would run almost indefinitely.

Remember that diodes have voltage drops. Thus, for there to be flow through both diodes the voltage difference must be greater than roughly 1.4 volts. Thus if the peak voltages are within 1.4 volts of each other, there would never be any flow through both diodes simultaneously causing large current flow. Even if the peak voltages differed by more than 1.4 volts, but were only 2-4 volts apart, the  current would flow for only part of the cycle.

For example, if the difference of peak voltages was 3 volts, current would flow for about 2/3 of the time and it would only be the current generated from a max voltage of 1.6V. One could calculate the rms (heating) value using some knowledge of calculus.

If someone would like, I would be happy to draw some diagrams and show the math.

-Lad

 

Lad Nagurney posted:

As long as there were trains on the track and neither control set to zero, it would run almost indefinitely.

Remember that diodes have voltage drops. Thus, for there to be flow through both diodes the voltage difference must be greater than roughly 1.4 volts. Thus if the peak voltages are within 1.4 volts of each other, there would never be any flow through both diodes simultaneously causing large current flow. Even if the peak voltages differed by more than 1.4 volts, but were only 2-4 volts apart, the  current would flow for only part of the cycle.

For example, if the difference of peak voltages was 3 volts, current would flow for about 2/3 of the time and it would only be the current generated from a max voltage of 1.6V. One could calculate the rms (heating) value using some knowledge of calculus.

 

 

Yes, of course, it would indeed be possible to assign circuit values and design a transformer that would not destroy itself in this circuit, but if you go back to my original post it is not about destroying the transformer, it is about working correctly!

Even if you select circuit values that permit continued operation, the "lower voltage" handle will always affect the "higher voltage" handle. During the half-cycle in which both diodes are forward-biased, the instantaneous output voltage (with respect to the common end of the transformer) is clamped at one diode drop above the position of the lower voltage tap. During the half cycle in which only one diode is forward biased, the instantaneous output voltage is one diode drop less than the voltage at the lower tap. The whole object of this scheme is to permit independent control of the two halves of the output waveform, but this arrangement does not appear to be a satisfactory solution.

If instead of two taps on a single core, two transformers were substituted, the results would obviously depend on the relative impedances and X/R ratios of the two transformers, but I doubt that satisfactory operation would be achievable.

 

For a PW ZW remember

1.  Zero volts output isn't the low end of the coil, it is the roller disconnected from the secondary coil by parking on an isolated pad, and

2.  The least voltage coming out of a ZW is a jump to 6-8V so that an E-unit will cycle.

The worst-case situation is about 10V difference when the 2 diode drops are included, and that is only for the half of the wave for which the diodes are forward biased.  Your would need some fairly large diodes.

Will the original circuit work in a practical situation, assuming moderate speeds for both trains to avoid collisions (if they are really on the same track)?  Probably, but it is dangerous without any circuit protection on the individual outputs.

Dale, thanks very much for your expert input. I suspected it would work to some extent, although it certainly has limitations and can't really be recommended. I'll give it a try myself when I get a chance, with separate phased transformers.

PLCProf posted:

... I am not talking about running the motors on half-wave rectified DC, there is a bigger problem. There are ways to apply the concept, but the example here is not correct.

Seems even today, some 50 years later, there's still the application to operate 2 trains independently.  Some would say command control and move on...so be it. 

But this could be a baby bath-water situation.  I'd think modern components could make this concept more viable.  Specifically some form of synchronous or active rectification where the two rectifiers respectively turn-on when it's their turn (60 times per second) and then shut off when the other is active.   This would prevent the shorting, back-feeding, or whatever you want to call it when the two voltages vary by more than 2 diode drops.  The synchronous rectification would likely be implemented with power transistors synchronized to the line cycle.  I'd think the cost would be just a few dollars and less expensive and more power efficient than using high-power rheostats.

stan2004 posted:

PLCProf posted:

... I am not talking about running the motors on half-wave rectified DC, there is a bigger problem. There are ways to apply the concept, but the example here is not correct.

Seems even today, some 50 years later, there's still the application to operate 2 trains independently.  Some would say command control and move on...so be it. 

But this could be a baby bath-water situation.  I'd think modern components could make this concept more viable.  Specifically some form of synchronous or active rectification where the two rectifiers respectively turn-on when it's their turn (60 times per second) and then shut off when the other is active.   This would prevent the shorting, back-feeding, or whatever you want to call it when the two voltages vary by more than 2 diode drops.  The synchronous rectification would likely be implemented with power transistors synchronized to the line cycle.  I'd think the cost would be just a few dollars and less expensive and more power efficient than using high-power rheostats.

I fiddled around with this stuff in junior high school over 50 years ago, including the circuit in question. It was pretty sad. If you search back through the auto-radio archives, you will discover that early radios used a mechanical vibrator, basically a glorified doorbell buzzer to generate AC that could be fed to a step-up transformer to generate the high voltage necessary to run vacuum tubes. Ordinarily, the output of the step-up transformer then went through a full-wave gas rectifier tube like an 0Z4 to produce DC.

I modified one of those so the coil ran on 60 cycle ac and hooked it up to a couple of train transformers to get the effect we are discussing. Basically I had an spdt relay that followed the 60 cycle line. It sort of worked, but in that era large low voltage capacitors were bulky, rare and expensive so I really couldn't filter the DC at the loco, and the vibrator contacts bounced and chattered and all that. Of course, due to operating time of the mechanicals and the inductance of the coil the synchronization with the power line was less than perfect so there was bad sparking at the contacts. 

I spent a whole Christmas holiday fiddling with that and finally gave up. I had some C. P. Clare mercury-wetted relays that I tried to use but those did not give a break-before-make when run at high speed. Also tried the transformer method under discussion but burned out all my dad's epoxy rectifiers, and it didn't work anyway. Semiconductors were expensive and not very forgiving in that era! 

Boy, I had a lot of fun back then!

PLCProf,

Tubes, aren't those semiconductor parts that glow orange when voltage is applied?

Boy I feel old. when in HS, I fiddled with those vibrator devices (mind out of the gutter, please) in car radios. I could take them apart and burnish the contacts then reinstall.  Usually had about 90% success rate.  Way I earned pocket train money was doing radio repairs especially the 5 tube gutless wonders.  Even helped the principal earn his PhD by repairing electronic typing gizmos used in dissertation study.