Let's make sure you're using the controller properly.  Assuming your train is address 3:

1) Select 3 on the controller, 3 lights up solid
2) Momentarily press the function button, 3 blinks
3) Momentarily press the 8 button, this should toggle the sound (either on or off), no change to the controller

On my DCC controller, I momentarily press the Function button to toggle the active system in and out of function mode.  I'm new so I'm probably using improper terminology here.  The light adjacent to the active system (e.g. 3 for train, 9 for turnouts and crossovers) will blink if in function mode and will remain lit if not in function mode.  I don't ever hold down the Function button while pressing a second button.  I must be out of function mode in order to switch from one system to another.

I think there are a couple of issues here: (1) Your meter is possibly inadequate to accurately measure the voltage frequencies associated with DCC.  I believe the fundamental frequency to be in the range 8 to 10 kHz with harmonics much higher.  I believe your old Lionel train provided power at the line frequency of 60 Hz.  (2) Measuring voltage drop along a track with no load is a waste of time.  There will be none (voltage drop that is).  About as useful as testing a battery with no load.

Assuming your meter is capable of measuring DCC voltage (even if not precisely), check if you are loading down your power supply (controller).  Do this by measuring the output of your supply (at the supply output), first with no load, then with both trains running at top (safe) speed with lights on.  If less than a 10% voltage drop, should be ok.  Now measure the voltage at the location of concern (along the track) with the both trains running at top (safe) speed, lights on, one passing through the location of concern.  If less than 10% drop from unloaded voltage, I would think you would be ok.  The simpler method would be to use empirical data, i.e. just see if the trains run ok.

I doubt there would ever be a significant voltage drop within the track rails, regardless of load.  The drop would occur at the rail-to-rail joints (electrical connections).

Another problematic characteristic of the voltage waveform might be the frequency content.  I think it's pretty high and some meters might not be able to detect the True RMS voltage accurately at the higher frequencies.  My Fluke meter seems to work fine but it's a much better meter than those cheap ones you see at Lowes.  For sure, you must put your meter on AC.

Really weird, was happening almost half the time I attempted to park the trains.  Hasn't happened again for several days.  I think I'll just keep an eye on it for now.  I don't think the room temperature has changed appreciably during this period (I'm assuming an A/D converter zeroing problem inside the controller).

Quentin, I should have mentioned that the Baldwin does have a significantly higher likelihood of derailing over turnouts than the Echo Valley train.  Due to very limited space, my track curves and turnouts are all 18 inch radius.  That could be a factor.  Actually, I'm very impressed with the Echo Valley train set after getting the Baldwin.

It seems to run fine.  Speed scale quite a bit different from the Echo Valley train.  Have to crank the speed control knob up further for the same speed.  This causes me to readjust the control each time I switch back and forth between the two trains.  No big deal.  My only complaint is the short and long horns.  The Echo Valley train produces a loud, deep horn that sounds like a real train.  The Baldwin produces a quiet, high pitch "toot" that doesn't sound like a train at all.  It's so quiet, you can't really hear it unless both trains have stopped.  The other sounds are okay.

It is apparent that the locomotives contain the motors that drive the wheels.  Just looks to me like the electrical power (pulse-width modulated with serial data) is being picked up off the tracks by the tender, processed, converted to appropriate level/polarity DC voltage, and sent to the motor inside the locomotive.  I think the locomotive doesn't have any intelligence, just a reversible, adjustable DC motor and a pair of wires connected to the LED.  I'm just guessing though.

Mine: Baldwin 4-6-0, Southern #1087, locomotive & tender.  Also: Echo Valley Express train set (2-6-0).

Quentin, in the foreground: Baldwin 4-6-0, Southern #1087, locomotive & tender.  In the background: Echo Valley Express train set (2-6-0).

I assume the derailing occurs only upon entering the heel (common end) of the turnout.  I've returned a couple of turnouts for the same thing.  Bought on Amazon so no problem.  Another issue I have run into is the weight of the cars, especially the last car of the train.  I had to rearrange my cars so I have a heavier car at the rear.  That took car of that problem.  I've seen lots of write-ups on this website concerning derailing on the turnouts.  It appears to be a common problem with Bachmann turnouts.  Apparently, the switches do not hold tight up against the rails.

I'm sure this is a stupid question but which is powered directly from the track: the locomotive or the tender?  Bachmann HO trains.  I assumed the tender receives electrical power from the track and drives the  DC, reversible motor in the locomotive using 2 wires and powers the LED headlamp in the locomotive using the other 2 wires.  I think the DCC decoder and speaker must be inside the tender.

While we're on the subject of crossing gate:  I bought one from Bachmann thinking it would be electrical, turn on warning lights, and drive the gates up and down.  Turned out it simply has a pad between the 2 rails connected to levers that move the gates down.  Or at least attempts to.  One goes half way down and both just wave up and down as the train passes.  A real disappointment.  If I'd bought it on Amazon, I would have returned it.  If there is an electrical crossing gate, I would like to know about it.

My uncle, who is no longer with us, had one in Oklahoma City.  A much older house.  The clay soil acts like a big flower pot and causes the basements to flood unless on a hill with some good drainage.  Tulsa is a bit different than western Oklahoma City; Tulsa has "soil."  We don't have anything that could be referred to as soil; just rock-hard clay.

Good info about the possibility of wiring the metal frog on my crossovers.  One of my two trains often momentarily stops when crossing one of my two crossovers.  Always gets going again but it is annoying.  I will take a look for a wire.  If I find one, I might need some help as to where to connect it.

This might be a little better.

Layout via AutoCAD.  Haven't figured out how to expand to full page.