Measuring the TMCC/Legacy track signal strength

There are more marks than spaces in a NOP...more ones than zeros.

Stan, that reasoning seems to make sense, I only recently started looking at the measurement panel and noticed that I was getting two different frequencies. 

However, figuring that I triggered it a bunch of times, and I always came up with the same frequency for the same base unit seems to suggest something else, though I'm not sure what.  I would have thought I'd capture the other frequency at some point for any base.

Let me bring back this thread. A few days ago I had lightning hit near my house. I had a few electronic devices that were affected and it seems my Legacy base was one of them. It was plugged in at the time of the strike and was working fine beforehand. Now all my TMCC and Legacy engines have flickering headlights and are responding erratically to commands. I switched out the Legacy system for my old Cab1/Base 1 and that works fine.

I built Dale's test circuit and I measure 3.8 VDC which is twice what Dale says it should be. Any ideas what's going on?

Ken

Which diode type did you use, and what is your meter make/model?

Dale,

I used the same diode you specified, a 1N914/1N4148. Also used a 4.7K resistor. The meter is a Sears Craftsman digital multimeter set on the 20 VDC scale.

Last night I took my Legacy system to a friends layout, hooked it up, and it had the same symptoms. Flickering headlights and erratic operation. I could hold my hand over the engine and the engine would operate normally. His Legacy system worked fine so I know the problem isn't related to my layout or a grounding issue.

Ken

Did you use his wall wart?  If so, we don't need to worry about a break in the grounding from your wall wart to the U-ground wiring of your house.

My Legacy Base is one that I got because it was damaged by nearby lightning.  The output transistor and a capacitor were damaged.

Dale,

 I didn't think about trying it with his wall wart. I used mine on my base to make sure the problem was with my stuff.

I just checked the resistance from the ground pin on the wall wart to the outside of the power plug and got 0.4 ohms. From the ground pin to the inside of the power plug reads 0.4 ohms also. Doesn't seem like these should read the same. Which one of the plug connections carries the house ground to the base? Getting exactly 8 VAC from the wart.

Out of curiosity I checked a Cab1 wall wart. Ground pin to outside of power plug - 0.4 ohms. Ground pin to inside of power plug - 4.3 ohms.

Ken

You may be looking at limitations of your meter.  If the Legacy power supply actually supplies power, clearly the power output isn't shorted.  It sounds like the Legacy power supply is probably OK, but like Dale says, the definitive test would have been to run on his brick.

With the TMCC Base you are reading just the conductor connected to the outer shell in one case, and the conductor PLUS the resistance of the secondary winding of the wall wart in the other case.  The Legacy power supply is probably similar, but since the wall wart is larger, the secondary winding would have a lower resistance.  Seems like there should still be a difference, but a smaller difference. 

Dale, unless you have a pretty good meter, you aren't going to read any difference.  I get .15 ohms difference between the two secondary pins and the primary ground pin.

I have not read EVERY post here, but couldn't a field strength meter with a 50 microamp meter sensitivity and an adjustable series resistor along with a coil tuned to 455KHz become your roaming signal strength meter?  Sounds like a simpler solution, although not thought all the way through.  This could easily be built into a car.

The R2LC has an internal pin that reflects signal strength of the TMCC track signal.  It's been suggested that a car be constructed to broadcast that indication to a remote receiver.

I took my base back to him and tried it with his wall wart. No difference, same flickering headlights and erratic control. Whatever the problem is, its in the base. I guess it goes back to Lionel when they start accepting repairs again.

Ken

I sent my lightning damaged Legacy set back to Lionel shortly before Aug 1. Got it back a couple of weeks later with a note that they replaced a resistor, a zener diode, and the charging circuit (which failed shortly after I bought it). Hooked it back up to the layout and had the same problem as before, intermittent response to commands on my TMCC engines, one Legacy engine that defaulted to conventional, and curiously a Legacy Y3 that operated just fine. Called Lionel and got to talk to the tech that worked on it. He said they always check the repaired base operation on their layout with a Legacy engine and a TMCC1 engine. My base worked OK for them. So they gave me a new RA and a shipping tag, and I sent it back.

A couple of weeks later I got a call from the tech apologizing for not having repaired it the first time. He took a scope and checked voltages and found that instead of the 5v p-p or better he should have read (I'm assuming he meant at the output of the command card) he got .3V. Changed the command card and fixed it. The thing both of didn't understand is why it worked the first time on their layout. Apparently some Lionel engines can cope with a weak signal and work OK as my Legacy Y3 did.

In any case I was impressed that I got a personal phone call and a lengthy apology for having to send it a second time. I consider that excellent customer service.

When I got it back, hooked it up to my layout and everything worked fine. I got a new command card and a new charging card and all it cost was the initial shipping charges to send it to Lionel the first time. Thank you Lionel Service!

Ken

That's not an uncommon failure and not an uncommon miss by the techs. I checked out a Legacy base for a nearby friend and found that the base which had been to Lionel also had a .3 vpp output from the U terminal output. He sent it back with a note as to what was wrong with it. And as you noted, some engines on a smallish layout maybe will respond with a weak signal.

I sent a note to Mike Regan suggesting that they buy one of my Track Signal Meters to routinely check the signal output, but never heard anything from him.  The idea of using trains on a layout to verify the integrity of the Track signal is ridiculous.  Either it has the 5V p-p signal, or it doesn't!!  Once you verify that the full-strength signal is present, you can operate trains to see if the functions are correct.

It isn't uncommon for the track signal to be reduced or distorted, and this crippled signal can sometimes operate some engines, but it is not RIGHT.  If the technician has a 'scope, he should look at the signal as a matter of course.

I found my first TMCC base with a weak signal a couple weeks ago Dale.  Had you not mentioned checking the signal I would likely not have thought of that.  I just use the 'scope to check them, but I did try the little diode trick as well.

I believe that there are a lot of people out there who are battling Command Base weak-signal problems, thinking it is layout or locomotive problems.

I have repaired a few TMCC Bases with that problem, and also my Legacy Base, which I bought from a Forum member with both of us knowing that it had a lightning-induced problem.

What's the most common component failure that causes this issue?

For that problem I typically see damaged output capacitors (there are a series and a shunt capacitor) and output transistor.  I have seen a few units with severely damaged power supplies, but I am not sure if that is also lightning induced.

Sometimes the waveform is distorted, typically clipped on one half due to wrong biasing.

It just occurred to me that I should include a "loaded output" test with my signal strength meter that puts something like a 50 ohm resistor across the Track Signal output to verify that both the no-load and loaded conditions are OK.

I didn't know what to use, so I was using a 1K resistor across the output.  50 ohms sounds low, does it have that much drive?

I just made some measurements:

100 ohms  -2.4dB

68 ohms  -3.6dB

51 ohms  -4.8dB

There is visible waveform flattening on the lower edge of the waveform for both the 68 and 51 ohm loads.

Since the signal did not drop to half yet, the output impedance is lower than 51 ohms, although the distortion is probably a more limiting factor.

Testing with 1K ohms isn't exercising the output very hard.

Sounds like 100 ohms might be a good value to use.  I don't think I'd want to drive it to distortion.  I agree that 1K may not have been optimal, I just wanted some sort of load on it to take a look at it.

Adding shunt capacitance causes a rise in output voltage up to about 3.3nF, with or without a resistive shunt.  On the same DUT above, the increase was 1.15dB max with no shunt R. With 3.3nF and a 100 ohm resistive shunt the rise dropped to about .4 dB rise.

This is a TMCC Base.  The Legacy Base has a more complicated output network that might react differently.

Edited and fixed my numbers...

On the TMCC base, it's an emitter follower with a 100 ohm resistor from the emitter to common. So the output impedance is lower than 100 ohms...the transistor being in parallel with it. And Dale just measured a db drop around which looks like an output impedance of around 36 ohms. A 100 ohm load seems like a good test of the circuit to show that it's all there and healthy. 

One of these days we will see a schematic of the Legacy stuff...I'm guessing it would be the same or similar, but could be an op amp follower I guess.

Interesting about the capacitive loads. Wonder what the average layout looks like to it.

The Legacy output stage is very similar to the TMCC version, but it has some L and C components between the emitter follower and the outside world.  Of course it is surface mount.  When I repaired my Legacy Base, I upgraded some of the components to hopefully improve the robustness.

The signal source is very different, starting with a square wave generated by counting down the crystal clock with two different countdown ratios for the frequency shift.  The sine wave is derived by LC filtering and then fed to the emitter follower output stage.