Mike Reagan on postwar transformers and legacy

I would be suspicious of the voltmeter readings. If you use a true RMS AC voltmeter then the power to the resistor should be the same even if the peak voltage on the chopped sine wave is greater. The heat output of the resistor is proportional to the average power applied, not the peak voltage. This applies to resistors connected across the rails. Those driven by electronic devices like AC regulators or R2LC triacs may produce different results.

Pete

Thanks.  That is interesting.  Our club runs AF 30B's with our Legacy and TMCC command bases at shows.  We get reasonably good smoke production out of our TMCC and Legacy AF engines.  At home, I use 30B's with TPCs and my Legacy Command Base.  It seems smoke production is more impressive, but then postwar engines with Gilbert Air Chime Whistles or Diesel Horns in them buzz all of the time and have to be disconnected.  Perhaps that is from the chopped sine wave?  

Dave,

As stated in the video the "peak" voltage is the DC voltage, RMS (Root Mean Square), which is the peak AC voltage times 1.414 = RMS. This RMS DC voltage is what the capacitors use to store a charge, so the stored voltage in the cap is the RMS DC voltage, not the AC voltage on the track. It is the chopped wave form that is providing the higher RMS reading, hence better performance of the newer electronics, such as the two steamers you questioned me about (they both have the newer RCMC platform inside them).

Thanks,

Mike

Why did you change the smoke units when they work so super on my fef legacy.  Mth engines smoke great and do not need chopped wave transformers.

"As stated in the video the "peak" voltage is the DC voltage, RMS (Root Mean Square), which is the peak AC voltage times 1.414 = RMS. This RMS DC voltage is what the capacitors use to store a charge, so the stored voltage in the cap is the RMS DC voltage, not the AC voltage on the track."

This is totally incorrect.  The RMS voltage is the equivalent amount of DC voltage that would produce the same amount of heating in a resistor.  For a pure sinewave, the RMS value is .707 times the peak value of the sinewave.

A CW-80 has a transformer inside it that produces an output of about 18 volts RMS.  This is the maximum "power" that is available, and feeding it through any form of control circuit that "chops" it will only reduce the amount of power available.

For example, the attached photo compares the voltage waveforms at the input (bottom) and output (top) of a Powermaster.  Note that there is actually less coming out than going in because of those notches that are used for horn/whistle and bell activation.  Those little nips only amount to a few percent of the power available at the input, but the output is and always will be less than the input.

Chopping DOES NOT improve the waveform's power!!  Adding a TPC to a brick doesn't not improve the brick's power! 

The chopped output will not work on many of the older mth proto 1 locos so is worthless to those people.  If  bridge rectifier is connected to the sine wave the pulsating dc would double the sine wave forms on top of the + line and should give more power than the one pulse mike shows.  Maybe he is assuming that a negative chop will occur right after the positive one, but I fail to see how I/2 of the sine wave with a pulse of 30 hz (1/2 of the 60 hz with both negative and positive pulse) being termed more than the full 60 cycles.  Help me if I am wrong on this.  The only thing I know is the cw80 does smoke way more than the 1033.  This explanation only makes it more confusing that mike makes and I agree with Dale. 

And the final product the new steam smoke units I have do not smoke as well like the older legacy ones do and I am disappointed with Lionel.  Maybe they need to sell more transformers so this approach is being used. 

I was taught that the average power from ac is the sine of 45 degrees with is .707 or the conservative average power is .636 of the ac waveform.  1.414 is the sum of .707+.707, the peak of the neg pulse and the positive but mike is only show the peak of the chopped positive pulse.  Where is the negative pulse?

I agree with what Dale Manquen said.

Yeh, 1.414 is impossible as it is peak to peak and that could only happen if both the positive and neg occur at the same time and they are 180 degrees out of phase.

I don't know whether Mike's explanation is technology sound, but the concept is sound.  Peak current/voltage/wattage could definitely affect how hot a wire or component will become more than average current/voltage/wattage.  The analogy I would use is that the damage from the heat of water on your skin is definitely affected by peak temperature, moreso than average temperature.  You could have similar average temperatures and suffer a third degree burn if the peak temperature in one scenario is much higher than the peak in another, even for a relatively short period of time.

I saw the video. The bottom line for the smoke unit (a resistor) is the total power produces heat and thus smoke. it does not matter how you get the power.

  I do notice more smoke with newer locos but part of that is the design of the smoke unit.  

   TMCC does help the smoke production because of full track power but many of the TMCC locos also have different smoke electronics. All IMO but I did work in electronics 40 years before retiring; Don

Remember there are threshold effects.  No smoke or minimal smoke until a certain temperature is reached, however briefly.  This is the principle behind pulse control of speed in DC power packs.  Frequent short pulses of high voltage yield movement whereas average steady voltage does not.

I'm afraid that pulse control of motors is quite different than simply heating a smoke resistor.  The motor analogy is flawed in describing power dissipation of a resistive load.  The pulses will indeed overcome the initial friction of the motor and start movement, that's true.  However, you can demonstrate the same effect with a pure DC supply to a can motor, it'll continue to run at a lower power setting that it'll start, simply because you've overcome the initial starting friction. 

I have to go with the opinion that the RMS value of the power, whether a pure sine wave or any other waveform, is going to be the determining value for how hot the resistor will get.  A plain resistive load doesn't have inertia or friction effects.

For a smoke resistor, the thermal time constant - the time required for the wire and core to change temperature - is much longer than the 1/120th of a second between pulses.

The only valid point I can see is when you are running at half throttle and the "shark fin" is just the last half of the hump of an 18V sinewave.  The RMS (heating) value of this would be higher than a full sinewave of half the height.  But who runs at half throttle in Command - and expects good smoke?

Dale, in conventional mode, I would think that running at half power is pretty common.  So isn't it possible that in conventional mode, a modified wave transformer would give better heating of the element than a pure sine wave transformer?  Perhaps that is the practical experience that Mike Reagan is referring to?

Exactly right!!!

So,  why do the newest smoke units in legacy steam smoke the same amt on low, medium or high and my older legacy fef with smoke low on low med on med and high on high.  All this technical talk does not explain the situation of the new smoke units and why did lionel make them smoke worse?

Any periodic wave shape has a peak, RMS, average value.

A voltmeter or any equipment will respond differently to

each value of the applied voltage. A sine wave transformer,

like the old ZW, will control train speed by changing the RMS 

value of the applied voltage.Changing the ZW RMS voltage

also changes the peak and average voltage.

A CW 80 controller changes the train speed by changing

RMS output but does not change the peak output except

at low voltages. The Mike Reagon video demonstrates this

very well with the coupler operation. With both the sine wave

transformer and the shark fin CW 80 set for the same RMS voltage,

the coupler operated with the CW 80 but not the sine transformer

because the coupler required the higher peak voltage of the CW 80.

The smoke unit is probably the same.

Where can one find this video? Figures I finally step up to a ZW and now I gotta get the new one!  HA!

I believe a man respected  as Mike is should have his name properly spelled. 

MIKE REAGAN

In the video, Mike Reagan set the output of both the CW 80

and the 1033 transformer to about 15 volts RMS which is a peak of 21v

from the 1033 but 25 volts peak from the CW 80. The coupler

requires 25 volt peak to operate after going thru TMCC so the

CW 80 will operate the coupler but not the 1033.

That makes sense if you can explain how the cut wave has a peak higher than the sine wave peak.

Well, if you're using a standard PW transformer with a pure waveform, you'd reduce the amplitude of the whole waveform, so at 18 volts RMS, you'd have a peak waveform of 25.2 volts, or a P-P waveform of 50.2 volts.  However, if you reduce the power to half voltage, or 9 volts RMS, you'd have a peak waveform of 12.6 volts or a P-P waveform of 25.2 volts.

In the case of the sawtooth output from something like a CW-80 or other electronically controlled transformer, at half power you still have a peak waveform of 25.2 volts, assuming the internal transformer core is an 18 volt unit.  You'd have the same resistive heating value, but it does have different characteristics than a sine wave.

I guess I need to study up on how a cut wave transformer works and how the electronics can hold back the 1st 1/2 of the sine wave and then shoot it up to 25 volts.

Does lionel still use 1/2 wave rectification as they did in the early digital sound systems.  I remember working on a Hudson 783 and was very unimpressed to see just one diode on the ac input and have outside rail as common ground.  I was all into qsi at the time and they were all about floating ground and bridge rectifiers.

It is easy to hold off the series pass devices for the first part of the waveform by simply not triggering them.  With a Triac, once you trigger it, you must wait for the current to drop to zero to "untrigger" it.  With a FET you can turn it off by shutting off the gate, but turning off a device that is carrying a large current can be hard on the device.  It is easier to trigger it ON from a non-conducting state.

For either type device there is an initial current surge because of the capacitors inside the locomotives.

Most of the TMCC circuits use half-wave rectification so that the ground can be tied to the frame and outer rails to pick up the track-borne half of the TMCC signal.  A floating full-wave scheme would not have this advantage.  (In the Base, the ground side of the half-wave rectification is tied to the AC safety/U-ground pin .)

Besides what Dale said, on those early units the chips where older technology and did not need very specific steady voltage requirements.  That diode did a half rectification to power the 5V DC regulator that feed the microchips.  Many of the chips were older thru hole stuff.  The motors and everything else ran on AC.

For QSI and mainly MTH everything ran on DC.  Except the couplers for QSI and PS-1.  MTH PS-2 is all DC.  Letting a light wire or coupler wire touch AC chassis will result in AC on the DC circuit and destroy both boards unless your extremely lucky.  G

In qsi the couplers were also dc from a capacitor discharge to the coil of the coupler.  I installed enough coil coupler kits to a after market system to know, just a small bridge and a 2200 mfd cap. 

The levels are not the same on the current Legacy product as you indicate, but if you feel the FEF-3 had wider range I respect your statement.  The latest smoke systems respond to "labor" and to the "Train Brake" too.  To fully see the dynamic range of the smoke production you need to exercise the full stimulus that affects the smoke production. 

As an FYI: Most of the behavioral changes from early Legacy to current delivered product are driven by Neil's input and the design of Vision product features.  The goal is realistic smoke effects, not overpowering smoke production.

An where did that DC Return too?  Chassis of the loco or the AC outer rail.  Same for MTH PS-1.  That was my point.  G

Jon, so do the smoke levels on the cab 2 set the smoke unit the same and the brake and chuffing levels regulate the smoke unit output.  For instance If I am simulating hard pulling I put on some brake to hold the engine back but have the throttle wide open, or I set the speed settings to slow speed with a wide open throttle and then press a higher speed setting and the engine will really chuff loud and puff smoke like crazy.  Now that is cool and did not try it until you put this post on.  I just thought my cab 2 was goofed up on smoke settings based on the controlled steady output of my fef and gs4, both which smoke very little on the low setting and allot on the high setting.  Thanks again for the explanation and maybe more about the new units should be put in the manuel.

ironlake2, correct.  The L, M, H just set the nominal level.  to see the laboring effect on smoke: if you have a Cab-2 - press the "speed" button and you will note there are 2 soft keys that are marked "efx".  The up arrow "efx" button will increase labor, and smoke will increase in kind.  On "H" and high "labor" you will maximize smoke production.

So great to get this type of inside and insightful information on the Forum.

Thank you Jon.

And thank you Rich & OGR for the Forum.

Gerry

Thanks jon. now I understand how my Milwaukee s3, c&o T1 and santa fe 3751.  yes I really get smoke on high setting and high efx.  great feature once you understand it and I really didn't get it until now. 

I wonder how many other owners of legacy really understand the smoke system now.

There are tons of information in this thread.  Too bad we do not have a forum spot to put it for future use.

Maybe our webmaster will read this and consider this idea.

That ground on the board is AC outer rail.  You can trace it.  Also is some applications, like aux tenders and AA units, the coupler return lead was just tied to the chassis via a screw.  G