Skip to main content

Hi all,

 

I visited the club in San Diego last week (not my club) and they had a rather peculiar sign on their TIU and power station:

SIGN

I thought you know... 60Hz is 60Hz... how different can it be... maybe they are just superstitious? I haven't heard of anyone making this claim before, (that DCS performs better when run off a brick).

 

Anyways... I like to be quantitative.... so when I got back to my club I measured one of the Lionel bricks they are talking about versus the output from the Z4000. In both cases the outputs are unloaded. Here are the time domain waveforms:

HD

The Z4000 has serious harmonic distortion compared to the Lionel brick so maybe this is the root cause of their observations. A low pass filter could clean it up, but it's high power and low frequency, so it'll get hot and be big.   It's actually quite plausible that this unwanted broadband content could disrupt enough of the DCS spread code to drop a packet.

I guess maybe there is something to this (good catch San Diego 3-railers!)..... Did anyone else notice these harmonic distortions yet? or see degradation on DCS from using the Z4000 instead of a brick?  I'd be curious what others think....

 

~Adrian

Attachments

Images (2)
  • SIGN
  • HD
Last edited by Adrian!
Original Post

Replies sorted oldest to newest

Curious what the symptoms are that make them come to that conclusion?  No mention of the symptoms.

The Brick is a straight transformer, so you will get a purer sine wave fixed at about 18Volts.

The transformer is a transformer.  It is electronically controlled to vary voltage, but designed to be closer to a pure sine wave then other manufactures chopped wave. So there certainly will be a difference.

They are trading off some voltage droop with the bricks.  The Z-4000 has a higher unloaded voltage available, so that you can get an 18V loaded voltage.

Would like to know the symptoms though?  G

Last edited by GGG

I've looked at the waveform from a few Z-4000s and I have never seen one that bad. What is interesting is that the output of the brick is also pretty terrible, which speaks to the incoming waveform from the utility. If you look carefully at the brick waveform, you can see a tiny discontinuity at about 45 degrees past zero crossing, which is about where the big blip occurs in the z-4000 as well.

According to the patent, the Z-4000 basically applies PWM to the utility supplied waveform to reduce its voltage according to the position of the throttle, and also uses some switched energy storage inductors to smooth the output, much like the series inductor and free-wheeling diode in the secondary side of a SMPS. There is no DC link. I wonder if the serious flat-topping of your utility power is making the Z-4000 unhappy!

There have been a few threads in the past showing waveforms of the Z-4000. Actually, they looked pretty good. Wonder if your Z-4000 has a problem!

The Z-4000 patent is an interesting read. It also gives more or less complete schematics.

At least two things cause harmonic distortion on the power lines.

- Anything with a motor, and the iron is saturating. Basically a non-linear inductance. I have seen this on scopes when working with power tools, even when the power tool was operated from an electronic AC power supply that re-creates a smooth sinewave. The purpose of the power supply was to maintain 120Vac 60 Hz or 220, 230, 240 Vac 50 Hz at the outlet.

- Anything with a bridge rectifier and a capacitor afterwords, which is just about any modern electronic device. Once the capacitor charges up, until the voltage increases to the point the diodes in the bridge can start conduction, very little current flows. Then alot flows. So its like a non-linear resistor. Once the voltage increases to that point it sags from a sine wave. Same thing happens with our trains with modern electronics, even if a pure sine wave was at the outlet and an ideal transformer.

Wild guessing the cost of making or buying a suitable power low pass filter would exceed the cost of just buying bricks.

Never had a Z-4000, does it have a full on position? Some light dimmers have that, when in the position no triacs in the circuit.

PLCProf posted:

I've looked at the waveform from a few Z-4000s and I have never seen one that bad. What is interesting is that the output of the brick is also pretty terrible, which speaks to the incoming waveform from the utility. If you look carefully at the brick waveform, you can see a tiny discontinuity at about 45 degrees past zero crossing, which is about where the big blip occurs in the z-4000 as well.

According to the patent, the Z-4000 basically applies PWM to the utility supplied waveform to reduce its voltage according to the position of the throttle, and also uses some switched energy storage inductors to smooth the output, much like the series inductor and free-wheeling diode in the secondary side of a SMPS. There is no DC link. I wonder if the serious flat-topping of your utility power is making the Z-4000 unhappy!

There have been a few threads in the past showing waveforms of the Z-4000. Actually, they looked pretty good. Wonder if your Z-4000 has a problem!

The Z-4000 patent is an interesting read. It also gives more or less complete schematics.

I didn't catch the tiny discontinuity on the brick measurement. I'm going to go back and measure 3 more Z4000s, maybe I just picked up a bad one. I'll electrocute myself trying to make a resistor divider to measure the utility waveform too.

illinoiscentral posted:

At least two things cause harmonic distortion on the power lines.

- Anything with a motor, and the iron is saturating. Basically a non-linear inductance. I have seen this on scopes when working with power tools, even when the power tool was operated from an electronic AC power supply that re-creates a smooth sinewave. The purpose of the power supply was to maintain 120Vac 60 Hz or 220, 230, 240 Vac 50 Hz at the outlet.

- Anything with a bridge rectifier and a capacitor afterwords, which is just about any modern electronic device. Once the capacitor charges up, until the voltage increases to the point the diodes in the bridge can start conduction, very little current flows. Then alot flows. So its like a non-linear resistor. Once the voltage increases to that point it sags from a sine wave. Same thing happens with our trains with modern electronics, even if a pure sine wave was at the outlet and an ideal transformer.

Wild guessing the cost of making or buying a suitable power low pass filter would exceed the cost of just buying bricks.

Never had a Z-4000, does it have a full on position? Some light dimmers have that, when in the position no triacs in the circuit.

An LPF is a tough one. As a circuit designer I went there immediately too.... but after a few seconds of thinking.... It's not just an RC or LC low pass on a breadboard.... the train draw amps of current, so whatever it looks like would be giant 10-100W rated components that cost a fortune. That time-domain measurement was done at the full position actually.

 

 

Here are some Z4000 waveforms I recorded.  For the record, I've never seen the Z4000 have an issue with DCS on a number of layouts, and I've seen no issues here.

Hover over the pictures to see the voltage and load.  Note that the probe setting was wrong when I shot these, so it should have been 10x what is indicated on the status line.  The first three are with a 2 ohm load at various voltages, the last two are no-load shots.

Z4000 'scope Picture 2 Ohm Resistive Load at 4.9VZ4000 'scope Picture 2 Ohm Resistive Load at 12VZ4000 'scope Picture 2 Ohm Resistive Load at 19VZ4000 'scope Picture No Load 11VZ4000 'scope Picture No Load 21V

Attachments

Images (5)
  • Z4000 'scope Picture 2 Ohm Resistive Load at 4.9V
  • Z4000 'scope Picture 2 Ohm Resistive Load at 12V
  • Z4000 'scope Picture 2 Ohm Resistive Load at 19V
  • Z4000 'scope Picture No Load 11V
  • Z4000 'scope Picture No Load 21V
gunrunnerjohn posted:

Here are some Z4000 waveforms I recorded.  For the record, I've never seen the Z4000 have an issue with DCS on a number of layouts, and I've seen no issues here.

Hover over the pictures to see the voltage and load.  Note that the probe setting was wrong when I shot these, so it should have been 10x what is indicated on the status line.  The first three are with a 2 ohm load at various voltages, the last two are no-load shots.

Ok. Maybe the power in southern California is just lousy? Anyways I'll repeat this measurement here this week and see. Thanks for those!

Anecdotally, the "best" DCS signalling performance is with DC voltage on the track (HO, G-gauge).  It stands to reason that since the DCS signal co-exists with the power, the cleaner the power the better the DCS signaling.  Since few guys run DC on O-gauge, I'd think in order of preference (for DCS signalling) you'd want a brick with pure-sine,  a Z-4000 with almost pure or synthesized sine, and in distant last-place a chopped-sine controller.  I suppose if one studied the Z-4000 patent or schematic there might be a scenario where a Z-4000 might create a cleaner AC track voltage than a brick if the raw AC wall voltage is noisy.

Add Reply

Post
The DCS Forum is sponsored by
×
×
×
×
Link copied to your clipboard.
×
×