Lionel Transformer Sine Wave Video

Once again, I praise Mike for doing this Video. I know that I do not have near the knowledge that a lot you all have in the electronics field.  But I think when I maybe have half the time that Mike has in repairing locomotives, I "might" think about criticizing him....but not now. I only wish I had a tenth of the knowledge that he and his time have on repairing these little toys that give all of us joy.

I know that maybe all he said was not "exactly" right in your mind, but I am so thankful that he made this video that helps this ignorant hillbilly. I now know more about this hobby which makes this a more enjoyable hobby for me.

He posted 4 slots up from you.

Emphasizing the point!  But sure would like to see a sillyscope trace of the ZW-L.

Well said. Maybe a little strong on the "criticizing" him...I think more like expanding and discussing his message. I think it's been healthy. And I have learned something from every post...some things I knew, some I didn't, some I still am not sure about, and all of them I thought about which is the really good thing.

Excellent video that illustrates how "tricky" the world of compatibility can be these days!!!  Many thanks to Mike and Lionel for taking the time to produce these educational videos.

David

A few comments about the video:

1.  A 180W brick connected directly to the track will give more power than going through any of the "chopped wave" controllers.  First, there is a voltage drop across the output transistors in the controller's output stage that doesn't exist with a direct connection.  Second, the controllers reserve a portion of the waveform for control of the bell and whistle/horn, and half of this reserve can never reach the track.

The 180W Powerhouse brick has a 2-stage electronic circuit breaker that is much superior to the dumb thermal circuit breaker in the TPCs.  The TPC does not have any active sensing of the output current.  Adding a TPC does not increase the protection.

For strictly Command operation, the TPC is unnecessary and a waste of money.  Peter, don't change a thing!

2.  I believe the argument regarding peak value and charging capacitors does not apply to TMCC devices that run Pullmore  AC motors.  The AC voltage to the motor is controlled by a "chopper" similar to the controller's chopper, without any peak storage devices.

3.  The comments about the PW ZW's output being limited to 3.4A is ridiculous, as mentioned previously by others.  It has 14 amps to share among all outputs as required.  Also, the venerable PW Z transformer has probably the highest voltage output of any model train transformer - 24V open circuit and about 14A continuous output rating.

The video seemed to be pushing new product.  I think the point Mike is making about the PW ZW is that four CW's have a higher output than one ZW.

The basic transformer inside the CW-80 is a mighty fine unit.  I tested one of those at 8 amps for 12 hours with very little heating!

What voltage did it put out at 8 amps?

Dale I believe the capacitor charge cycle was explained as the voltage that goes to the coupler and smoke unit not necessarily to the motors exclusively. It is very hard to make a video like this and make it understandable for the layman and still informative for an expert. I think Mike did a pretty good job for both.

Ron

let me throw this into the mix and see what the brains here think (since it is a weekend and mike is out until tuesday).

he and i had a long discussion about power and the benefits of moving to modern chopped sine wave power from my post war ZW's

i took my 9 power districts (formerly 3 on one ZW, 4 on another ZW and 2 on a third ZW) and put them across 2 180w bricks with 2 TMCC direct lockons.

i had feeds for each power district from my circuit breakers, so i was able to mount them onto two gang bus bars for easy testing and management of connections.  i feed the "hot" wire from the lockon to the bus bar.

on one power district sits an 11 car passenger train with 10 of the cars having lights and which is pulled by a lionel milwaukee bi-polar.  if that district is alone on one brick/direct lockon combo, it is fine.  if i start adding other districts, the lockon breaker trips. if it put it on the other lockon, the same thing happens.  

have i maxed out the capacity of the brick/lockon combo?  that seems unlikely, but i cannot see what else it would be.  what happens if that train starts rolling?  from a strict wattage standpoint, the two bricks don't equal what i had with the ZW's.  do i need to add another brick/lockon or two???

I saw the video and at first was impressed by the presentation. However after some reflection I think it is a bit misleading.

The first test compares a CW80 and a 1033 put through a TPC or whatever. The CW 80 puts out 18vac RMS and peaks at about 25. The 1033 puts out 16VAC and peaks at around 22.4 volts. So on circuits charging capacitors in engine circuits in a command environment the CW80 would have about a 2.5 volt advantage since capacitors charge to peak. However a TW puts out 20 volts at full throttle and in a command environment would peak at 28 volts,better than the CW80. Since running in command and the throttle is at maximum I see no advantage at all in a chopped sine wave since if run through the TPC it is chopped anyway.

In a conventional environment I see the advantage in in theory of the chopped sine wave. If both transformers were set at 6 volts,a capacitor circuit in the engine would charge to about 8.4 volts from the 1033 but 25 volts with a CW80 since it maintains its 25 volt peak. If the capacitors are sufficiently large they would drive the circuit they are powering.   Even a BCR should charge on the first throttle rung or 5.4 volts with a CW80. Nice I guess if you have legacy locomotives. However I have a lot of PS1 units and they do not like the shark fin wave.

Might have been a better design to make the new transformer coming out pure sine wave and chop the circuit with the engine electronics,then the transformer would be universally compatible. I see no advantage at all of the chopped wave in the command environment,if run at full throttle the shark fin wave would be closer to pure sine wave anyway. In fact might as well just hook the bricks straight in,why have a throttle in command?

The PWZW realistically is good for about 200 watts continuous or about 10 amps or maybe 11 amps at 20 volts. If used in command mode all that could come out of one tap. Normally only 2 taps would run 2 loops in conventional,all 4 would not normally be used with 3+ amps each,so this is a bit misleading. The other 2 taps might do a few lights or accessories and not draw a lot of current.

Lastly he says 20 amps can be safely put to the track. Maybe I guess,but 10 amps to a loop would be more on the safer side. Anyway the 20 amps could be accomplished with 2 bricks in parallel, and 4 of them would be cheaper with the same wattage than the new transformer and has good breaker protection in them. One poster says he lessens the arc welder effect by reducing track voltage to around 15 volts. While this reduces potential,it increases amperage through a Triac or whatever driving a motor. At a lower voltage more amperage would be required to do the same amount of work. As along as it falls in the parameters of the chip tolerance it is OK but it is something to think about.

I have a question for Dale M or one of the engineers. If a TW and a CW 80 each powered a load through a wire and were set at 6 volts. Would the CW80 voltage have less resistance going through the wire than than the TW since the voltage still peaks around 25 for the CW while the TW peaks at 9 volts? If I am wrong about any of the above please correct me also.

Dale H

Convert the cars to LED lighting. In command each car will use 1 watt or less. It is a lot easier to do this than push more amperage into the system.

Dale H

Rob, my data for the CW-80 transformer (original version) alone with a line voltage controlled at 117V was:

No Load - 17.9V

2.58A - 17.5V

8.80A - 16.5V

10.4A - 16.3V

These numbers yield a source impedance of about .155 ohms.

Somewhere I have some test data on my two PW ZWs, but I can't seem to locate it.

I should also test 135W and 180W bricks.

The point of the video was that if you are running

conventional a chopped wave works better than

a sine wave transformer. With the chopped sine wave

you can reduce the RMS value so the train will run at

a lower speed but will still have a high peek voltage for the

smoke unit and coupler operation. This will work up to a

point than as the RMS voltage is reduced so will the peek

voltage.

If you are running in command than a sine wave transformer

will work because you have a full 19 volts for the smoke unit

and coupler with speed controled by the command control

system. There is no advantage to the chopped sine wave in

command.

Dale:  I use a modern ZW with four 180 watt bricks that has Lionel's digital readout meters attached.  What do you think the voltage should read at start up?   I ask because I only seem to get a 17 volt reading and if a passenger train and several engines are on that track, it goes down to 16 volts. I have used a digital meter and confirmed the voltage at the track.  Is this normal?

Last year I had a gentleman at York install special LED lighting in a diner passenger car but once it was delivered, the lighting would go up and down in intensity.  When I discussed the problem, I was informed that the circuit was designed for 18 volts and would not operate correctly at anything less and since Lionel's bricks put out 18 volts, there must be a problem with my wiring or layout.  It has been a very frustrating issue as I have over $150 invested in the upgrade.

I would appreciate any thoughts you may have.

Thanks,

Don

Dale, the shark fin waveform only reaches the peak value for output voltages equal to or greater than half the maximum voltage.  Each half cycle of a 60Hz sinewave is 8.3milliseconds, and any shark fin less than 4.15ms won't start until after the peak voltage has passed.  At the 6V of your example, the shark fin's turn-on point would be well past peak.

The current waveform is somewhat different for a capacitor charging from a sinewave vs. a shark fin.  For the sinewave, the charging current starts when the rising voltage sinewave reaches the value to which the capacitor has discharged during the zero crossing portion of the sinewave.  For typical circuits with adequate energy-storage capacitors, the droop will be about 20%.  That means that the charging current starts at about the 80% point on the rising waveform, and probably continues until about maybe 90% on the falling waveform after peak.

Converting these percentages to angle (out of 180 degrees) and time, the turn-on point is at 53 degrees or 2.4ms.  The turn-off point is 116 degrees or 5.3ms.  The current charging the capacitor is therefore a large pulse 2.9ms long. 

Since the short peak current must be quite high compared to the average or RMS current, the resistance drop through any wiring will be higher than for a for a smooth sinewave current of similar average or RMS value.

The current waveforms will be identical for a sinewave and a shark fin as long as the shark fin's voltage turn-on rise is before the 53 degree point.  This would be true for output voltages of about 75%-80% of full throttle and above, or around 14V for an 18V source.

I hope this was indeed an answer to your questions.

If the loco has an AC motor, wave forms with other than a sine wave will result in less efficient operation and heating of the motor?  

Don, the output current for a modern ZW with meter passes through 3 devices on the way to the track - a diode, a FET (Field Effect Transistor) and a current-sensing .01 ohm resistor.  As you pull more current, the drops across these devices increases.  At 10 amps output the drop can easily be 2V.  (There is also some resistance/impedance in the brick's coils and circuit breaker.)

LED lighting with a regulated source can give you nearly constant lighting over a fairly broad range of track voltages.  For the nominal 12V 3-LED strips that many of us use, we run the LEDs at about 10V to get a reasonable brightness.  With a fullwave rectifier and an energy storage capacitor, this requires a sinewave voltage with a peak value of about 13 volts or about 9V RMS.  Most of the PW transformers start at 6-8V to have enough voltage to cycle the E-unit.

(One advantage of LED lighting is that the color doesn't change when the light output is reduced.  Since the eye responds to light intensity logarithmically, we don't note much of a change for small voltage fluctuations.)

Does the CW 80 transformer have a circuit breaker ?

How can you get 10Amps at 16Volts, which is 160, watts when the CW 80

is only an 80 watt transformer.

Yes...It blinks it's green "on" light when there is a short and recovers when the short is removed. I think it's an overcurrent electronic switch rather than a circuit breaker.

I really appreciated the discussion of the waveforms and how the CW-80 works.  I knew CW-80s produced somewhat chopped power, but not just exactly how.  Mike made it clear how and why modern power supplies (I will be technically correct and not just call them transformers) provde better loco respondse and smoke.

I use three CW-80s (one each for my shorter loops - which run smaller trains with less power demand - and one for the BEEP underneath the lake surface on my water skier/boat-on-lake) and a Z4000 on my two bigger loops where I often have lighted passenger cars and multi-loco trains with power hungry postwar motors.  I assume MTH does something similar since I get similar performance between the two types except in cases where the additional power of the Z4000 is needed.  

I have wondered about this.  A pure sine wave is . . . well, a pure, single frequency.  The chopped waveforms produced by these power supplies are not - they have harmonics - higher frequencies - as some components injected into the track.  I've wondered if this causes DCC or any system that signals the loco control digitally through any problems - the harmonics interfering with the digital control signals.  I assume Lionel engineered legacy and TMCC so that its CW-80s, etc., don't interfer, etc., but i wonder if they engineered them to avoid causing PS1 and 2 problems (actually, the Macheavillian side of me has wondered if they engineering them to cause problems, but forget that for now . . . ) I wonder if CW-80s and PS1 and 2 are ascompatible, and vice versa.  I run conventional so it would not be a big problem for me, regardless . . .  

Lee, see my post above (yesterday) about CW-80s and PS1 engines, which I think means that some early PS1 engines were not compatible with any chopped waveform transformer.  Looks like MTH made later model engines to run with any late model transformer. I haven't noticed any problems running PS2 engines with a Lionel ZW and power bricks but then I use purely command operation except for testing purposes.

"Convert the cars to LED lighting."

highly impractical, given the number of passenger cars i have.  

also, hard to believe the old ZW's could handle this with no problem and the newer gear cannot.

I tend to disagree with the side coming down on not using the TPC in command mode.  This is all I use since I stopped trying to run command loco's with a ZW.  I NEVER ran/run at 18 volts.  Even with the ZW, I'd run at 14-15 volts.  TPC's are dialed down to a similar level. While a TMCC loco has parts engineered to run at 18 volts everything else didn't.  I'm with Dale H in terms of knocking down the power requirements rather than trying to to throw more and hope for the best.

Things I've noticed.  

The only TMCC equipped loco I own with open frame motor runs much better in conventional mode using the TPC (It also run much better with the Legacy controller but that's another story), especially at lower speeds.  Dale M's explanation of the on board triacs in the motor board explains that.  In conventional, they aren't doing anything.  None TMCC open frame motors run much better in conventional using the TPC.  I haven't seen any improvement/difference in can motored trains in conventional other than the TPC offering finer steps/increments (72 vrs 32).

At 15 volts I get enough power through the TPC to fire the couplers and get decent smoke output.  This wasn't always the case with the ZW.  

As far as power protection goes I have the Lionel command lock on (the little shed) that I use for command mode operation.  Device works as advertised, auto reset and all.  It's too bad it can't work in conventional mode (it shuts itself down once the track voltage drops below 10 volts).  

Regarding noise/harmonics.  I suspect that the noise generated by dirty track/pick up rollers, dirty commutators, worn brushes, damaged or missing bearings, no lubrication, etc contribute more to signal degradation than the modern modified power supply outputs.  The issue with early QSI/PS-1 boards were design issues (assumptions that the trains would be powered off of PW power supplies) and were addressed fairly quickly.  

A series-wound AC/DC motor does not have a synchronous rotating magnetic field like an induction motor or polyphase motor; hence, the harmonics don't adversely affect the motor very much.  There might be a slight increase in eddy current loss in the core, but I doubt that this would be significant.

Regarding the harmonics and the Command signal, TMCC/Legacy's signal at 455 KHz would be the 7,583rd harmonic of 60 Hz.  The command signal is FM (Frequency Modulation) which is fairly immune to fixed frequencies due to line harmonics.

How I love the "good old days".   Two wires to the lockon, turn the black handle on the 1033 transformer and their went my 2023 loco.   Didn't care want wave it was.

I was thinking of this when I read your post.  I have an old MTH PS1 Northern that just will not run on any CW-80. It tends to stall frequently as if it had been told to stop, etc.   Frankly it does not run that well with my Z4000 either, only a tad better, but I have an old postwar transformer that I think is pure sine wave output and it likes this a lot more (to be honest it does not run too well at all anymore: old and tired).

All my PS2, and my one PS3 engine, and all my Legacy locos run as expected (except for mechanical problems) with either Lionel or MTH power - but then I run conventionally which is not a stressful test of their ability to distinguish between harmonics and actual sigital control signals.  Still  . . . I took about a recent Legacy loco (U30C) and was impressed by the metal grounding shields and such placed at points around the eelectroncs boards, to the extent it seems clearly somebody has really thought about shielding from interference - so I assumed it had been an issue. 

Thanks your you comments and postings on this.  Helpful.

Quote from Chuck:"Regarding noise/harmonics.  I suspect that the noise generated by dirty track/pick up rollers, dirty commutators, worn brushes, damaged or missing bearings, no lubrication, etc contribute more to signal degradation than the modern modified power supply outputs.  The issue with early QSI/PS-1 boards were design issues (assumptions that the trains would be powered off of PW power supplies) and were addressed fairly quickly."

Yes, it is a harsh noise environment to design for, actually - not military harsh but not as simply as for some sumsumer products, that is for sure.  Both Lionel and MTh seem to have done a good job, because, like I said, except for very early locos, all of mine seem to tolerate just about anything.

Quote from Dale M: "Regarding the harmonics and the Command signal, TMCC/Legacy's signal at 455 KHz would be the 7,583rd harmonic of 60 Hz.  The command signal is FM (Frequency Modulation) which is fairly immune to fixed frequencies due to line harmonics."

And that would make it very likely that there is an issue. 

This is a very interesting thread!!!!

Just for completeness here, I think we should include a quick description of the Z4000's technology.

I don't own a Z4000, but I believe the voltage control is achieved by chopping the incoming voltage at a high frequency, and then modulating the width of each of those chopped packets.  The packets are then "glued" back together and the high frequencies from the chopping are filtered out to yield a smooth sinewave.  This type of circuit works at a much higher frequency than a simple shark-fin control, but since the controls are all ON/OFF switching, there is only a small amount of loss as heat during the transitions.

I believe this fairly common technique, which is not too unlike the switching power supplies in all of our computers, is unique to the Z4000 in our train world. 

I once saw a Z4000 hooked up to a scope and the output was not a perfect sine wave, but much closer to that than to the "shark fin" form. That strikes me as consistent with Dale's description in the above post.

Hardest part of the job is taking the cars apart. Eventually you will be doing it anyway one by one to change light bulbs. While in there might was well do the conversion After that there should be no need to take it apart again,the LEDs should last a long time.. The 11 passenger cars you mention could draw 50 to 100 watts or more just by themselves with original light bulbs. K-Line with stream lighting even more than that.  With LEDs they would use 10 watts or less total draw. This power saved could be used to run trains regardless of transformer used.

Dale H

I posted a comment about this in another thread, but here I'll pose it as more of a question to those with some electrical/electronics knowledge.  When running trains with a CW80 or Z1000, I've noticed increased motor heating and noise when compared to a pure-sine-wave transformer like the MRC Pure Power, Lionel postwar ZW, etc.  It's present in both can motors and universal/Pullmor motors, but it's more pronounced on the universal motors and actually seems to make them run rougher.  So, my questions are- is this harmful in any way to the motors over time, and what exactly causes the noise and increased heat generation?

"Hardest part of the job is taking the cars apart."

i have 40+ passenger cars.  this is a non-starter.  i have two k-line cars that i cannot get apart.

my point is the old ZW's handled this with no sweat.  with TWO passenger trains of this size on the layout at the same time.

now here is an interesting tidbit.

if i have only two districts on this circuit (the one with the passenger train and another with two legacy steamers heading freight trains), the breaker trips.  if i remove EITHER ONE of the steamers, the breaker does not trip.  explain that?  legacy detection process?

Total the wattage of the trains involved.  You are probably running at the edge of what the ZW's can push put out and the thermal breakers are probably hanging on by a fingernail.  Modern breakers generally trip as soon as you hit the current limit and they don't tolerate momentary peaks the way a PW ZW could.

i routed the power from the ZW's through fast acting circuit breakers on each power district.

Scratch that.  I have to go look at the wiring diagram.

here's another tidbit.  if i put some tape down on center rail and roll one of the steamers onto it and power up. all is good.  if i roll that engine off the tape, no problem.  it is something about the draw at powerup that is pushing me over the threshold.

more detail on the power configuration.  i had (have) 9 power districts separated by insulated rail joiners on the center rail.  each district powered from one channel on a PW ZW.  3 on one, 4 on another and 2 on the third.  the other channels are powering lights, switches, uncoupling tracks, etc.

the power from each channel is routed through a fast acting circuit breaker and then to the track for the appropriate power district.

The 180W Powerhouse brick has a 2-stage electronic circuit breaker that is much superior to the dumb thermal circuit breaker in the TPCs.  The TPC does not have any active sensing of the output current.  Adding a TPC does not increase the protection.

For years, I have run 2-180w Powerhouse Bricks (360w) with a TPC 400 powering each "circle" of track (4), i.e. 8-180W Bricks and 4 TPC 400's.  When a circuit breaker trips, it is always a breaker in the bricks that pop.  I have never had a circuit breaker in the TPC pop, which supports Dale's assertion.   

Well, I have three "circles."  According to your formula, I need 4 more bricks.