Adventures in TMCC & Legacy... Trying to master the 455Khz track signal. I've built the test car, and it is finally working..

Hi guys, great conversation going here, keep it up. I'd like to see this topic end up being THE PLACE to come do discuss everything related to the 455Khz signal. That's why I'll be transitioning the title to better reflect what the conversation is all about.

My layout is an interesting testing ground with plenty of quirks and mysteries to solve. There are lots of theories out there on how to make our trains run better, and with any luck we can prove or disprove many of them in this thread.

I am not a "scientist". I do not have the electronic background or education that some of you have. What I do have is a general understanding of many of the concepts, which I have picked up from many of you over the last few years. I also have a willingness to learn and try new ideas based on your suggestions.

I had a wonderful phone conversation with Dale last night. We talked about more tests that should be done as well as my grand layout control scheme and how it was all going to tie together with C/MRI, JMRI and TMCC/Legacy to create a fully automated CTC mainline. The theory for all of it is quite sound, as most of those components have been around for a long time, and are well tested. The secret to reaching the end goal is perfecting the track signal for smooth, RELIABLE operation.

Sorry I got a little behind in replying to your responses this morning, there were so many, I got a little overwhelmed.

Dale, the radio should be here next week so I can do the sweep (and check some locos). I have to reconnect the bases and track power, in order to do the chicken wire disconnect test.

Chuck, maybe we should try to rename the term "ground plane" to something like "earth ground carrier". It would be more accurate and descriptive.

BobbyD, Nick is right, another Legacy base would not have been the answer for this problem. All that would have said was one base worked and the other didn't. This was all about probing the bad base to figure out what made it bad. The fact that it was intermittently bad, made it confusing on one level, but also made it possible to ultimately pin down the problem.

Interesting. so without  purchasing an oscilloscope, is there a simple way to check the strength of TMCC / Legacy signal on your layout. What areas on the layout would you test?

J Daddy posted:

Interesting. so without  purchasing an oscilloscope, is there a simple way to check the strength of TMCC / Legacy signal on your layout. What areas on the layout would you test?

Yeah, John, so far there hasn't really been much talk of a scope. Dale's meter seems like it will do all the tasks necessary for now, including probing suspect areas of the track for comparative values. The cheap AM radio is to sniff around for 455Khz noise emitters. That level of sophistication I can deal with.

Big_Boy_4005 posted:

Chuck, maybe we should try to rename the term "ground plane" to something like "earth ground carrier". It would be more accurate and descriptive.

I kind of like "ground wire antenna"...

But it'll never happen. Ground plane is so catchy .

First post repair test: PASS

I reconnected both bases and plugged in the power supplies. All engines stayed at rest. No more "Runaway Train" or "Unstoppable" (both bad movies ). Cab-1's work as expected.

cjack posted:

Big_Boy_4005 posted:

Chuck, maybe we should try to rename the term "ground plane" to something like "earth ground carrier". It would be more accurate and descriptive.

I kind of like "ground wire antenna"...

But it'll never happen. Ground plane is so catchy .

It is rather confusing, isn't it.  With the TMCC signal the true 'ground plane' as the term is used in other radio applications, is the track with it's U post half of the signal.  Then we are using 'Earth ground' to broadcast the other half of the signal.  Very backwards from how folks brains want to think of things. 

JGL

Dale suggested renaming it a while back. I think it was suggested to be name ground signal radiator or something like that.

Here is that thread.  As you can read I was initially opposed to changing the term due to the widely recognized but incorrect term "ground plane" was instilled into the TMCC culture but today I try to use the term "ground signal radiator" and explain why.  I think Dale's initial thought was probably something we all should have started using a while ago.  Dale has done more for TMCC / Legacy signal understanding than anyone I know.  We could call it the "Dale Effect"  LOL!

Dale, the radio should be here next week so I can do the sweep (and check some locos). I have to reconnect the bases and track power, in order to do the chicken wire disconnect test.

Big Boy 4005

Can you please explain what the use of the radio is for, I guess I missed something. Thank you

Ray

Ray, the 455Khz track signal frequency is just below the regular AM dial. Dale feels that it is close enough when tuned to the low end ~530Khz, that it can detect potential interference emitters around the train room. As we were discussing this in our phone conversation, Dale had an Ah Ha moment. It relates to what Mike Reagan said in this topic 3 1/2 years ago regarding the unshielded coil on the Railsounds board. What Dale realized was, the radio could probably detect that "noise" too. This way you shouldn't have to open up the engine to find out if it's a problem.

Lionel added shields to updated boards, but the boards are $40. For about 10¢ in materials, you can make the shield yourself (labor extra of course).

Elliot, you can also find old dead RS power supply boards with the shields and retrofit them.  They fit on the old boards with the exposed coil.

John, I'm not even sure where to look for old dead boards, I live in a 3 rail wasteland. I think it's easier to make them now that I have the pattern. Dumpster diving behind the hobby store isn't really my idea of a good time.

Where's your sense of adventure?

gunrunnerjohn posted:

Elliot, you can also find old dead RS power supply boards with the shields and retrofit them.  They fit on the old boards with the exposed coil.

My thought...I know where there is one

Update:

I've done a couple of things on the layout since we last talked. I installed a ground plane wire on top of that superstructure that I built. Unfortunately, the results were not what I had hoped or expected. There was no real change in engine performance, though a couple odd things happened during the test. Random engines on different areas of the layout woke up with their sound on. They didn't move, but they were revving at high RPM's. When I went address them with the Cab-1, and shut them up, they would not obey. On one occasion, I tipped the engine to lift the rollers off the center rail, and let it back down. That cleared the sound before I had to resort to ear plugs or aspirin.

I also did Dale's chicken wire disconnect test.  I unplugged the layout lighting which tied the chicken wire to earth ground. Then using Dale's meter, I measured the signal level going to the layout. Remember that the Legacy base measures about 2000 when the layout is NOT connected. With the layout connected, and the lighting unplugged, the meter read 1000. I did a quick continuity test between the chicken wire and a different piece of conduit which was still connected to earth ground. Results negative, no connection. The chicken wire was isolated.

So the layout itself is dropping the signal by half. Then I plugged the lighting back in, and the signal dropped to 950. After that, I turned the lights on and the signal dropped to 920.

I guess the next step is to start using the meter to take readings around the layout, and measure the signal level at the bad spots. How low will it go?

You might be interested in some test results from checking the output of the Legacy base with various levels of capacitive loading.  It may explain some of your results.

 Check the PDF file attached to this post.  When reading, note the voltage scale changes as the capacitance gets larger, so you need to factory that into your reading.  When we get to .1uf, the signal is only about 300mv, probably way too little to function properly.

I see that at .1μ, the wave is very noticeably crushed. I'm assuming that if I had a scope, and knew how to read it, I would see this effect in the bad spots.

Hard to say, but I suspect you'd see a weak signal.  All you need is a 'scope.

gunrunnerjohn posted:

Correct Dale, my point was simply that it was a DC measurement when you look at it.  Their sample application does have the resistor and cap.

Earlier you said:

I recall them saying when the signal was in the 40's, everything was great.  When the signal dropped down past about 30 and lower, the wheels started to fall off. 

What are the units?

The 455 kHz IC receiver meter output is a current proportional to signal strength with a specified nominal uA per dB slope.  The 51k resistor converts the units to Volts per dB.  The 0.1uF capacitor smooths the voltage.  For average signal strength you want to set the meter to DC Volts.   It's puzzling that the photo shows the meter set to AC Volts?

gunrunnerjohn posted:

You might be interested in some test results from checking the output of the Legacy base with various levels of capacitive loading.  It may explain some of your results.

 Check the PDF file attached to this post.  When reading, note the voltage scale changes as the capacitance gets larger, so you need to factory that into your reading.  When we get to .1uf, the signal is only about 300mv, probably way too little to function properly.

So, John,

He has one outside rail for his train detection circuit with isolated blocks. Each isolated block outside rail block break is joined with a .1uf capacitor according to Mike Reagan's video for creating a capacitive coupling to allow the signal to travel that rail.

According to the charts, it appears that value kills the signal.

It looks like it should be a .01uf

Yes, No ?

Edit: photo from earlier reply

Looks like 1042 caps or 1uf

Where are we now? does TMCC  still work?  Has  Carl's post above found   part of the problem?  re the   . 1uf or .01 uf?  .

If you're trying to coupler the TMCC signal, the .1 would be correct IMO.  Capacitive reactance of a .1uf cap at 455khz is around 3.5 ohms, and of a .01uf cap at 455khz it's around 35 ohms.  Since at 60hz it's over 26kohms, I like the .1 to couple the signal.

stan2004 posted:

gunrunnerjohn posted:

Correct Dale, my point was simply that it was a DC measurement when you look at it.  Their sample application does have the resistor and cap.

Earlier you said:

I recall them saying when the signal was in the 40's, everything was great.  When the signal dropped down past about 30 and lower, the wheels started to fall off. 

What are the units?

The 455 kHz IC receiver meter output is a current proportional to signal strength with a specified nominal uA per dB slope.  The 51k resistor converts the units to Volts per dB.  The 0.1uF capacitor smooths the voltage.  For average signal strength you want to set the meter to DC Volts.   It's puzzling that the photo shows the meter set to AC Volts?

I know Stan, clearly the meter wasn't set right, that's all I can figure.  The NJ-HR guys are the ones to ask.

gunrunnerjohn posted:

If you're trying to coupler the TMCC signal, the .1 would be correct IMO.  Capacitive reactance of a .1uf cap at 455khz is around 3.5 ohms, and of a .01uf cap at 455khz it's around 35 ohms.  Since at 60hz it's over 26kohms, I like the .1 to couple the signal.

Well, it looks like he has 1uf, no decimal points, unless I found a different 1042. What does that do to the signal?

oops! 104 Z   .1uf  sorry, didn't zoon enough     Thank you , John.

But, that rail needs a cap from the legacy base wire in series to the outside isolated rail or it gets no signal.

gunrunnerjohn posted:

If you're trying to coupler the TMCC signal, the .1 would be correct IMO.  Capacitive reactance of a .1uf cap at 455khz is around 3.5 ohms, and of a .01uf cap at 455khz it's around 35 ohms.  Since at 60hz it's over 26kohms, I like the .1 to couple the signal.

This subject came up in my phone conversation with Dale. The upshot was that the .1μ caps were almost useless, since the train actually grounds both rails as it passes. The train sort of brings the signal with it as it hits the insulated rail.

We were also both skeptical track "halo effect" where the train moves through a "tunnel" of signal. I would love it if we could prove or disprove these TMCC legends here.

Gregg the bases are hooked up, I can run trains. Not as well as I would like, but that's what we are working on here. This will probably go on for quite a while, unless we stumble onto a miracle. If and when we do, you'll know.

Carl, it's a lot easier to read the label from the bag they came in, than it is to read the component in a photo. Trust me, they're .1μ.

 Thanks for the update...  great thread.

gunrunnerjohn posted:

Hard to say, but I suspect you'd see a weak signal.  All you need is a 'scope.

If I buy one, I'll have YOU teach me to use it. Actually, you'll have to teach me what to buy first.

Gregg posted:

 Thanks for the update...  great thread.

You're welcome. We'll try to keep this fun, interesting and informative in the process.

Elliot, you never know when a miracle is just around the corner, keep the faith.

gunrunnerjohn posted:

Elliot, you never know when a miracle is just around the corner, keep the faith.

Gee John, maybe buying a scope isn't so far fetched after all. It's just a tool, and dealing with all this signal stuff and the C/MRI it could be very useful. The last time I used one was high school physics. It might even help me understand electronics better, and that certainly couldn't hurt.

As for keeping the faith, I am. Even when others around here have "lost it".

I'll go back to my seat now.

Elliot, the digital 'scopes are very cheap nowadays, I remember the days when I would have given my left nut to be able to afford one, now they're less than a Legacy diesel.

Yeah John, I just looked on Wikipedia and saw some pictures of some newer models. I had no idea what was out there. My mind is stuck in a time warp, I thought they were still big clunky boxes. Less than a Legacy diesel, huh. OK, you've heard my two major projects, I'm sure there will be other smaller ones along the way. What do you recommend?

Sorry Carl, I wasn't trying to be mean. I do make mistakes like everyone else. It's good to have as many eyes on things as possible to make sure I'm not missing anything. This is very technical stuff, and John, Dale and some of the others keep bringing up new concepts and theories to consider. I don't get all of it, but I'm trying.

gunrunnerjohn posted:

Elliot, the digital 'scopes are very cheap nowadays, I remember the days when I would have given my left nut to be able to afford one, now they're less than a Legacy diesel.

One of my favorites...Rigol DS1102E

http://www.tequipment.net/Rigo....html?b=y&v=7906

gunrunnerjohn posted:

 

I know Stan, clearly the meter wasn't set right, that's all I can figure.

So has anyone actually mapped/plotted the signal-strength around their layout using meter output signal from R2LC pin-13?   I'm not talking about measuring the 455 kHz strength on the track but the 455 kHz as received via the antenna.  I'm thinking of something like the cellphone coverage maps from AT&T, Verizon, etc.. 

I'd think having a real-time indicator of received signal strength with meaningful resolution (2.1 vs. 2.2 vs. 2.3 etc.) would be invaluable.  You could make adjustments or changes to grounding, capacitance, whatever, and watch the received signal strength change in a way that you'd know if you're getting warmer or colder so to speak.  The pin-13 dynamic range is enormous (60 dB or 1000-to-1) with resolution to see changes of a few percent one way or the other - as opposed to go/no-go observations of whether TMCC commands are making it through.

The pin-13 method has been brought up for years in different threads.  Has it been dis-credited or shown ineffective?

cjack posted:

gunrunnerjohn posted:

Elliot, the digital 'scopes are very cheap nowadays, I remember the days when I would have given my left nut to be able to afford one, now they're less than a Legacy diesel.

One of my favorites...Rigol DS1102E

http://www.tequipment.net/Rigo....html?b=y&v=7906

I really like the sound of that unit. The price is very doable too. Now all I need is to learn how to use it. Unfortunately, it's slightly more complicated than a hammer.

stan2004 posted:

gunrunnerjohn posted:

 

I know Stan, clearly the meter wasn't set right, that's all I can figure.

So has anyone actually mapped/plotted the signal-strength around their layout using meter output signal from R2LC pin-13?   I'm not talking about measuring the 455 kHz strength on the track but the 455 kHz as received via the antenna.  I'm thinking of something like the cellphone coverage maps from AT&T, Verizon, etc.. 

I'd think having a real-time indicator of received signal strength with meaningful resolution (2.1 vs. 2.2 vs. 2.3 etc.) would be invaluable.  You could make adjustments or changes to grounding, capacitance, whatever, and watch the received signal strength change in a way that you'd know if you're getting warmer or colder so to speak.  The pin-13 dynamic range is enormous (60 dB or 1000-to-1) with resolution to see changes of a few percent one way or the other - as opposed to go/no-go observations of whether TMCC commands are making it through.

The pin-13 method has been brought up for years in different threads.  Has it been dis-credited or shown ineffective?

I can just hear the little engineers asking, "Can you hear me now?" Seriously, a layout signal level map would be very cool.

Interesting to read but this is like learning Japanese for me I.e. It's way over my learning curve.

To bad there wasn't a "booster" you could hook up to amplify the signal. Again how do the NJ High-Railers do the same on a monster layout without signal loss?

I picked up a really cheap scope a few years ago. I had not used one since college. I bought it to look at the sine wave from my pure sine inverter. In other words to check 60Hz was being produced correctly. 

It's pretty basic, cost under £100 when I picked it up.I have not tried looking at the Legacy/TMCC signal with it though. 

Needs GRJ, Chuck or one of the other knowledgable chaps to check the specification to see if it will do the job.

It's very portable which might be handy.

http://www.maplin.co.uk/p/vell...t-oscilloscope-n55jn

Nick

Nick, while that 'scope would probably suffice, I'm a fan of buying a tool that will not limit you in the future.  Here's the Rigol DS1102E on eBay: 252273188974 for $350 shipped.  It's a new unit in the US that has been opened but not unpacked.

Stan, I like the idea of using the R2LC output and actually trying to map what you have.  What would be really slick is to have the value read be broadcast to a remote receiver so you could take readings where you can't see the unit.  Failing that, perhaps a "smart" unit that would give you varying tones based on signal strength so you could get readings where you can't really see a meter.

I recently bought one of the 100 MHz digital scopes for around $280.  I love it.  It is much smaller and lighter than my other 'scopes, and much 'smarter'.  One of the other Forum members recommended it to me.

A number of different vendors sell what is essentially the same 'scope.  Mine is the Hantek DSO 5102P.

I wish I could hook up a 12V gelcell and make it portable.