On Mon, 26 Oct 2020 10:45:04 -0700 (PDT), Tim R
wrote:
On Monday, October 26, 2020 at 1:38:41 PM UTC-4, Ralph Mowery wrote:
In article ,
says...
Hey, I like it, but we can complicate it more. I think we need a Raspberry Pi or an Arduino in the circuit controlling it.
Outside of the programming the Arduino is very simple to use for many
things. They can be bought for less than $ 5 from many places in China.
That is often less expensive than many of the relays it would take to do
the job.
I need to learn more about that; this is one of those technical advances that kind of sneaked by me. I was trying to figure out if I could replace my washer timer with something like that, 30 year old Maytag and can't get parts, but then it started working again.
snip
I setup an Arduino (Uno) as the heart of a semi-automated controller
for my BIL's extensive model railway layout.
The test track I have here is just a 3m straight length of (OO) track
on a bit of timber with a small passing place in the middle section,
joined to the main line with a couple of sets of points.
There are diodes in one track leg near the ends to kill the power to
the loco if it was to overrun (but allowing it to reverse out).
There are 4 coded IR sensors on the track, two near the ends, two just
outside the points.
Arduino ramps up the PWM, loco pulls away and accelerates (with
inertia) to full speed till it passes the sensor past the distant
points then decelerates (again with inertia) and then creeps to the
furthest sensor then stops.
The points change (capacitor discharge via electronic bridge drivers)
then the loco pulls away, round the bypass and to the other end.
Rinse / repeat.
The idea was to have some automated bits of track that ran stuff at
the back of his main layout, just to keep something running while he
was working on the layout or running other stuff himself.
The full project was going to be a main line with two sidings at each
end.
The track would be broken up into at least 5 sections, the main run
and the 4 sidings. All 5 could be driven independently (but as one)
and each section 'sensed' for current draw to determine occupancy /
transition.
3 locos (probably railcars / trams) could be placed on the track, two
in any of the 4 'ends' and one on the main line.
On startup, the Arduino would test each section of track for occupancy
(test current pulse), set the points to one of the empty sidings and
drive any loco on the main line into an empty siding.
Now it would randomly select one of the two trains facing the empty
siding at the remote end and set the live line to that siding. It
would then slowly run out onto the main line (past a sensor),
accelerate to full speed (user setable with a pot) to the remote
sensor then slowing though the points and to the end on the empty
siding. We could add a halt / station in the middle if we wanted.
Re evaluate and run again, either randomly or cycling though all three
locos / units.. ;-)
During out initial trials we learned we would probably have to select
a small subset of similar spec models from his extensive collection to
ensure the performance of each loco was similar.
Because he has quite a collection, he really didn't want to go
digital, hence all the intelligence had to be in the track.
We might also need to use the Arduino to drive an independent PWM
controller with feedback, to allow better control.
I have an Arduino Mega running the 3D printer I built a few years
back. ;-)
I have a Raspberry Pi3B running Open Media Vault (NAS) on a 3TB USB
drive and another running TVHeadend with a couple of HD TV tuners.
All good fun. ;-)
Cheers, T i m