Lionel Legacy Base Causing DCS Signal Problems - Resolved -

@gunrunnerjohn posted:

Well, getting a "complete" list of all the causes is most likely the impossible dream!  By your own admission, there are varied causes, and I'm sure we haven't actually properly identified the root cause of many of the interference issues.  For instance, Joe posted a unique issue that I had never heard of before!  The TMCC being a separate layout and not connected to the tracks with DCS is a new one, I had never heard of that issue before!

Whoa.... I am very sorry here GRJ.

I traced the issue to a ground wire that connected my upper 3 rail loop to the rest of my two rail. The TMCC signal was connected thru an internal connection on something, that I now forget. I believe it was the Z4000? The 3 rail had a separate TIU so that wasn't it.

It drove me nuts and I only shared the video to show it happens. Nothing else was dragging the signal down. NO TMCC engines on the rails.

Later on when they became available, I had two more PH180 bricks (on order for a year?) so I could separate the Z4000 from my outer O scale 2 rail loop.

Since I got my 2 rail MTH Bigboy, I connected the Z4000 again for higher voltage and power. So I have to remember not to plug in my TMCC base!

Also, my experience is that signal degradation is worse with PS2 engines than PS3 engines.  What is the reason for that?  Thanks for all the input.  I am getting a real education here.

Michael

@Michael Cimba 040217 posted:

In my situation I am using the WIU and the DCS app on my phone so even if signals between the remote and tiu could be an issue it wouldn't be in my case.

OK, well that rules out corruption of the RF signal in this case since if it were, it would likely degrade the WiFi signal to a point where nothing would work.

Just throwing this out as a question.  Is there anything a particular type of track system could be doing to amplify the legacy signal?  I am using gargraves phantom track and ross switches.

As I understand it, in the TMCC/Legacy world the track acts as an antenna.  How much track do you have?  If it was a length of 1000 feet (1/2 wavelength) or multiple thereof, the signal can reflect and "resonate".  Based on your earlier post it sounds like it's doubtful you have that much track.

That said, RF is a tricky business I wouldn't rule out a particular combination of length, materials, and shape to making a more effective antenna out of the track.  However, characterizing that benefit/degradation would require someone with a lot of RF knowledge and likely some RF modeling tools.

@Michael Cimba 040217 posted:

Also, my experience is that signal degradation is worse with PS2 engines than PS3 engines.  What is the reason for that?  Thanks for all the input.  I am getting a real education here.

My guess is that the processor doing the signal extraction (Adrian mentioned it was some sort of convolution, which to me says it's a DSP of some sort) is not as powerful in the PS2 engines as in the PS3 engines - which makes sense.  Typically in that world, if you increase the size of the Fourier transform used you can pull more signal out of the noise at the expense of integration time, i.e. the result might take longer to get.  However if the DSP is faster, you might get it for free in PS3 chipsets.

@SteveH posted:

Please forgive my unfamiliarity with the DCS system.  John, between which devices is the DCS signal strength measured and then displayed on the remote?

The "signal strength" measurement is really a misnomer.  What is actually being checked is sending packets to the locomotive from the TIU and then counting the returned responses.  If you are reading 10's, that means all packets are being responded to and received by the TIU back from the locomotive.  However, that doesn't actually measure the signal strength on the track, it is really just checking if there's enough signal for packets to go back and forth.

@rplst8 posted:

As I understand it, in the TMCC/Legacy world the track acts as an antenna.  How much track do you have?  If it was a length of 1000 feet (1/2 wavelength) or multiple thereof, the signal can reflect and "resonate".  Based on your earlier post it sounds like it's doubtful you have that much track.

That said, RF is a tricky business I wouldn't rule out a particular combination of length, materials, and shape to making a more effective antenna out of the track.  However, characterizing that benefit/degradation would require someone with a lot of RF knowledge and likely some RF modeling tools.

Actually, the track is NOT an antenna in the world of TMCC.  The "antenna" inside the TMCC locomotive is picking up the radiated signal from the earth ground (3rd prong of house wiring), the track side of the signal is connected through the wheels to the electronics directly.  In this regard, the RF propagation model is kind of upside down, but the effect is still the same.

Dale Manquen wrote a significant piece on TMCC Signal Basics, and he clearly explained it much better than I will.  Sadly, his website is no longer up, but I captured that topic here.

Dale's TMCC Signal Basics.pdf

Dale's description is spot on.

@rplst8 posted:

My guess is that the processor doing the signal extraction (Adrian mentioned it was some sort of convolution, which to me says it's a DSP of some sort) is not as powerful in the PS2 engines as in the PS3 engines - which makes sense.  Typically in that world, if you increase the size of the Fourier transform used you can pull more signal out of the noise at the expense of integration time, i.e. the result might take longer to get.  However if the DSP is faster, you might get it for free in PS3 chipsets.

I believe when Barry was around he stated that exact same thing.  I know both the Rev. L and PS/3 chipsets got a much more powerful DSP signal processor for better performance.

@gunrunnerjohn posted:

The "signal strength" measurement is really a misnomer.  What is actually being checked is sending packets to the locomotive from the TIU and then counting the returned responses.  If you are reading 10's, that means all packets are being responded to and received by the TIU back from the locomotive.  However, that doesn't actually measure the signal strength on the track, it is really just checking if there's enough signal for packets to go back and forth.

@rplst8 posted:

Actually, the track is NOT an antenna in the world of TMCC.  The "antenna" inside the TMCC locomotive is picking up the radiated signal from the earth ground (3rd prong of house wiring), the track side of the signal is connected through the wheels to the electronics directly.  In this regard, the RF propagation model is kind of upside down, but the effect is still the same.

Dale Manquen wrote a significant piece on TMCC Signal Basics, and he clearly explained it much better than I will.  Sadly, his website is no longer up, but I captured that topic here.

Dale's TMCC Signal Basics.pdf

Dale's description is spot on.

@rplst8 posted:

I believe when Barry was around he stated that exact same thing.  I know both the Rev. L and PS/3 chipsets got a much more powerful DSP signal processor for better performance.

John, Thank you for taking the time to explain what exactly the DCS "signal strength" means.

On behalf of all of us newer forum members, an even bigger thank you for saving and re-posting Dale Manquen's TMCC Signal Basics article.

I highly recommend that anyone with a TMCC/Legacy system read it.  For the first time, I think I now understand the two different parts of the Legacy 455kHz transmitting antenna:

1st - Connected to the Legacy base's U post is the Outer track rail.  This serves as the antenna base (but as you point out, also a direct conductor to the Loco)

2nd -  the "earth ground" connected from the TMCC/Legacy base back through its wall wart to the House wiring, is essentially the "whip".

From the receiving Loco's perspective, these two electrically opposite Antenna "Poles" (sort-of) both need to be somewhat balanced.  When the signal strength of one Pole overwhelms the other, poor reception by the Loco results.

Sorry for the TMCC tangent here, but this is really good information, and I'm a bit excited to finally understand it, I think.

Now back to DCS...

@SteveH posted:

On behalf of all of us newer forum members, an even bigger thank you for saving and re-posting Dale Manquen's TMCC Signal Basics article.

I highly recommend that anyone with a TMCC/Legacy system read it.  For the first time, I think I now understand the two different parts of the Legacy 455kHz transmitting antenna...

Yes me too.

As to the OPs issue - I still wonder what exactly is at play that could cause the issues he's seen.  Even in the thread that GRJ linked from Adrian, while I think it's understood why TMCC can interfere with DCS in some situations - characterizing exactly what the situation is where one person's dual TMCC/DCS layout won't work - but many many others do, still seems elusive. 

Could the length of the house wiring play a factor?  Also I wonder - in situations where this interference occurs is there a way to determine if the TMCC signal in the air (or the track) is exceptionally strong compared to a "normal" environment.

Yes, thank you GRJ and others for all the invaluable info on how Legacy/TMCC and DCS works.

Would having the Legacy Base plugged into an outlet with a GFCI on the circuit do anything to the signal?

Michael

Well today I finally had some time to do a good amount of testing on the layout.  Tonight I discovered something very interesting.  Up to this point I had connected my test track only to Fixed Output 2 of the TIU.  I started moving the track connection to the different terminal blocks.  First I tried Fixed Output 1.  I got a perfect 10 DCS signal with the Legacy Base plugged in to power and the "one wire" connected to the Fixed Output 1 terminal block common.  I left the "one wire" connected to the Fixed Output 1 terminal block and I moved the track wire to Variable Output 1 and then Variable Output 2.  In both of these cases I got a perfect 10 DCS signal with 1 PS3 engine and 1 PS2 engine powered up on my test track.  So then I setup a second test track.  I moved the 1st track connections back to Fixed Output 1 and I connected the 2nd test track to Variable output 1, leaving the "one wire" connected to the common of the Fixed Output 1 terminal block.  Again I got a perfect 10 DCS signal with 1 PS2 and 1 PS3 engine on test track 1 and 1 PS2 engine on test track 2.

So, at least it appears that the issue is only on Fixed Output 2 of the TIU.  As soon as I unplug the power to the Legacy base the signal on Fixed Output 2 goes to a perfect 10.  I still can't explain what's causing it, but it seems it is only affecting 1 TIU output instead of all 4.

I'll be curious to see your input on this and if this info helps anybody else dealing with this problem.  Thanks.

Michael

Michael,

It could be that you have a weak DCS signal on Fixed Out 2.  If you search the forum, there are reports of this happening with some TIUs that are early Rev. L.  Look for posts from Adrien! (yes with the ! mark).  Gunrunnerjohn also knows a good bit about it and may be the guy to talk too if you need it repaired.

Given that it's only one channel, it's very likely that it's the DCS signal generation and not anything to do directly with Legacy.

@gunrunnerjohn posted:

Given that it's only one channel, it's very likely that it's the DCS signal generation and not anything to do directly with Legacy.

GRJ, Is this a possible issue with the TIU that needs repaired like rplst8 stated, or is it just a matter of trying to improve the signal on that channel with the 18 volt bulb etc.  What was the issue in the TIU that caused signal problems on Fixed Output 2 only and what is the fix?  Is the fact that I get good signal on that channel when the Legacy Base is not plugged in still indicative of a low DCS signal generation.  I will try to test that channel a little more today and see if the signal remains strong without the Legacy Base plugged in or if it fluctuates.  Anything else I should be looking out for?  Thanks.

Michael

A "good signal" is relative.  If you're counting on the track signal test that MTH does reading it on the remote, that's not really a true signal strength test.  All that does is tell you how many DCS packets out of ten can make the round trip to the engine and back.  However, you can have significantly degraded actual signal amplitude on the channel and still get a "perfect" test as far as the MTH measurement is concerned.

The only true test of actual signal strength is to monitor it with a 'scope and see what actual levels exist.

With that being said, if you have read some of the discussions about DCS signal strength and the factors that influence it, including connecting TMCC/Legacy, you'll see that it's far from a simple subject.

The Rev. L TIU DCS output signal drivers are much more susceptible to damage by track transients for some reason.  You'll see several threads on replacing the 74ACT244 driver chips to fix signal output issues.

Thank you GRJ!  I think I can borrow a scope to measure the signals.  Can you point me to the proper procedure for setting up/hooking up the scope to measure the dcs signals and what I should be looking for in the output? Thanks.

In order to measure the signals, you'll need a pure sine wave transformer as you have to add a filter to remove the 60hz power to see only the DCS signal.  Then you need to whack together a little filter to knock out that 60hz signal, and the result will be just the DCS signal.  I set the DCS remote operating mode to SPEED and make sure there's an engine loaded in the remote.  You can then press the STARTUP button whenever you want to see the DCS signal, the SPEED setting prevents the remote from stalling when it doesn't hear from the engine.

Once you get everything hooked up, you should see something like this.  The amplitude of the signal should be at least 10V P-P.  Ignore the overshoot spikes and look at the actual signal, the denser part of the waveforms.

I didn’t read all the posts so maybe it’s been mentioned. I read a long time ago not to run the legacy wire through the TIU. Hook it directly to the track.

@romiller49 posted:

I didn’t read all the posts so maybe it’s been mentioned. I read a long time ago not to run the legacy wire through the TIU. Hook it directly to the track.

Yes, that's correct.  You want to always connect the legacy "one wire" to the output side of the TIU or points beyond (terminal blocks etc), never through the TIU.

@Michael Cimba
You asked:  "Would having the Legacy Base plugged into an outlet with a GFCI on the circuit do anything to the signal?"

No, because only the neutral wire and the phase wire(s) pass thru the detector coil.

@All:  A related safety hint (or 3):

The most sensitive of these trips at 5 milliamps (.005 A).  It is commonly said that if this amount passes thru one's body, it will knock one off a ladder.  More pertinent for the model railroader would be to say that if one is under the layout soldering wires overhead, one will drop the hot iron on one's face.  Remember that lying on a concrete basement floor  is an effective ground for any type of electric tool with exposed metal parts, if its internal insulation fails.  I once had such a drill.  Go with a battery drill.  There is a reason for the popularity of the 20v battery.      ---Frank M.

I did some more testing this evening and it led to more very interesting discoveries and positive solutions I think.  First I just want to thank the LORD for giving me the wisdom to figure this very daunting problem out on my layout and I also want to thank all of you on this forum for all of the expert advice and information that led me to a probable solution.  I hope my experiences will help someone else.

I will preface this by saying I believe the problem was caused by my wiring as GunrunnerJohn had speculated.  There were 2 different aspects of the wiring that came into play here I believe.

From my last post you will remember that I had narrowed the problem down to the Fixed input channel 2 on the TIU.

First, I had my Legacy Base mounted on my control table between my 2 Z4000 transformers.  The power wire for the legacy base as well as the power wire in the split PD9/Serial Cable/Power for the SER 2 ran very close to and in some instances in the same bundle of wires as the cables going from the transformer outputs to the TIU inputs.  The wires coming out of the transformer going to Fixed input 2 were closest to these Legacy power cables.

Second, I had all of my transformer common's hooked together (2 Z4000's and 4 Z1000's) as stated by Barry in his DCS book.  I ran a wire across my control table top from one Z4000 common to the other Z4000 common and this wire passed underneath and very close to all the wires coming out of the Legacy base.  Additionally I had a common wire connected to 1 of the Z4000's from all of the common's on my Z1000's tied together.  This wire passed very close almost touching the wires from the transformer to Fixed Input 2 of the TIU.

Third, I had my WIU on the control table in close proximity to the Legacy Base and the transformers and all the associated wiring as well.  I moved the WIU to under the layout next to the TIU.   

When I removed the Legacy cables from close proximity to the cabling going from the transformer to the TIU and when I removed the wires connecting the transformer common's together the DCS signal improved to consistently getting between 7 to 10 with a PS2 engine where as before it was anywhere from 1 to 6.  Additionally, I then physically moved the Legacy Base to the floor a couple feet away from the control table and made sure all of the Legacy wires weren't close to the transformer to TIU wires.  This gave me a consistent 9 to 10 DCS signal with a PS2 engine on the track connected to Fixed Input 2 and a consistent 10 DCS signal with a PS3 engine.

So, it appears that this combination of wiring routes and component placement was causing my issue.  I can't explain electrically what actually was happening (there are probably many on this forum that can) but I know the changes I made solved my DCS signal issues. 

Now, I will have to figure out the best way to rerun my wires that will yield the same results as what I am getting with everything "loose". 

Thanks again for all the help in solving this issue.  I hope this information will help someone else from having similar problems!

Michael

Fields and waves.  I suspect the legacy rf signal was inducing noise on the closely run wires. You may have heard at some time that it is a good idea to cross wires at 90 degrees, that is why. Sounds like you got the wires separated now.

I really admire how you stuck to this issue and seem to have learned a lot more about debugging your layout.

I recall a situation years ago at a plant where I worked. During some major expansion the electrical contractors ran low voltage control wires in the same tray as the high voltage power lines, everything running parallel. The controls wires were to variable speed pumps and there was no control. Even pros make this mistake.