I just got an hour to get started on that last little piece. The plywood has been cut out right from the start. (Lying on the floor).
And now I’ve also cut six small pieces of wood so that I can screw the purchased steel angles under the plywood. (Located on top of the plywood down on the floor). It is strange. But I never have exactly the screw needed – in this case 3.5 x 17 mm countersunk spun screw. So a trip to Bauhaus was necessary.
The plywood for the parade-strecke is as good as fitted: It is adjusted, the angles are screwed onto the chassis and there green parquet flooring is glued in all the places where the plate should be mounted. Now, only Stopp felt and rails, track switch servo and wiring are missing. Sensors and decoders are fitted.
But that’s the smallest part of the work. The larger part is the definition of SVJ as well as parade-strecke in JMRI – including virtual signals and the whole. That took a few hours.
I put the rails on the plywood to line up where to drill and saw. It’s a great feeling to see the layout completed.
Parade-strecke is having all kinds of faults: There is no sound damping between the rails and the plywood. Rails as well as plywood are loose and can really make up a sound base. But still, there is hardly any noticeable difference between rails with green parquet flooring and parade-strecke. But SVJ with Stopp felt is much better damped. Now we’ll see if Stopp works just as well on the parade-strecke when I get it applied.
I also just mounted a servo and a lantern in the new turnout. It’s impressive as soon as I forget. So for next time (and especially for my own sake): First insert the large piece of plexiglass. Then the servo (including the steel wire) and finally the small piece of plexiglass.
Parade-strecke and thereby the entire layout is done:Stopp felt is damping the sound better than the green stuff for parquet flooring. But there is not much difference. So I will not bother to change anything.
It is a strange feeling to have “finished” building the layout. I don’t quite know what I’m doing next. There are enough candidates:
– Warrants and scripts for all trains.
– The latest changes to SCWarrants have never been pushed to Github and hence to the new releases of JMRI. I still have to update to the latest JMRI version and then push those changes. This is what I did on May 29, 2017, allocating and deallocating the entire route of an SCWarrant as a unified operation.
– All the eventually 10 turnouts lanterns need power in order to include that little decorative element.
– The cameras must be mounted in the shadow yards.
– I have another showcase to be re-built and hung on the wall.
– Tunnel portals so that the transition between the visible and the hidden parts of the layout become more well defined.
– Hiding the front shadow yard, i.e. to make a faceplate so that the track is hidden.
– Modules or rather frames for the modules which are to form the landscape and the road for my Faller car.
– And then hopefully there will be a power supply and some plugs etc., so that I can make the final rebuild of the servo decoders. I must have changed a bit on the parade-strecke turnout settings, but it can’t be done until this power supply thing has been done.
– The semaphore signal for Spøttrup station must be built and equipped with a servo.
Video surveillance at shadow yards
I have started playing with WEB cameras. I have received a total of three web cameras from China. One that cannot be adjusted in focus and two that can.
The cameras will be used to show on the PC screen what is going on in the shadow yards.
I had hoped that I could connect all three (and preferably one more) either to the PC via a USB hub or to a Raspberry Pi. But none of these options work.
If I connect the cameras to the PC via a USB hub, there are two problems: The cameras will only show half resolution and only one camera at a time can be used.
Also, if I connect the cameras to a Raspberry Pi (model 2B) I can only use one camera at a time. And here I can’t get live pictures, but only a snapshot approx. once per second and delayed for several seconds before it appears on the PC screen.
I have ended up with a compromise where I connect two cameras directly to the PC, using a USB connector on each side of the PC, which I think means two separate USB controllers. The third camera must then just generate snapshots with a one-second delay. For that camera is connected to a Raspberry Pi.
If I need more cameras, it may require a desktop PC with even more USB controllers. But I have ordered two things that might save my old laptop from this hurdle: An expresscard with 2 USB 3.0 ports for 5 Euro. I hope it contains another USB controller. And then my PC has a firewire connector. I hope that means another controller, and have ordered a firewire / USB cable for 1 Euro.
The changes I made to JMRI on May 29 and June 11, 2017, should be included into JMRI.
That is why I have updated to the latest version of JMRI and committed my changes.
It wasn’t easy, because I couldn’t compile the new version. I got an “out of heap memory” error, which usually means some error where the compiler is running in an infinite loop. But after a few hours of trying different things, I found out that I needed to update to the latest version of the compiler.
But now I am in doubt if I should also include a change I did in the ECOS as to whether I should keep a change regarding turnouts not being set if JMRI thinks / knows that they already in the desired state.
It may be better (if JMRI does not have the correct status) to not include this change.
The JMRI change is committed. It is a real pleasure to work with Bob Jacobsen and the rest of this community, who in this case, as well as many other cases, embraced the improvements and helped it through tests, etc.
In the meantime, I have completed the back of showcase number two, as well as cut the balsa triangles for the shelves, which I will glue later tonight.
Unfortunately, I got three shades of LED strips. It is not just in the picture that there are two white, two green and two violet shelves. Furthermore, the glue is bad on the green ones. They fall down. So I will probably have to apply a different kind of glue. Or I try my luck and buy another roll of LEDs and hope that is better.
That’s how it is when buying cheap in Jem and Fix.
The problem is that there are also two different kinds in the lower showcase. And one of them is very similar to the green one, but is from another roll. The green roll was brand new when I cut for the two shelves.
There is white light on all shelves. Jem & Fix swapped with a smile – because I could show a voucher. Remember to save it!
And then I have thrown myself into something that gives me almost physical discomfort and at least shaking hands: My good Märklin MY, that crashed to the floor and bent its buffer, needs repair.
And then I had to break the buffer completely. Then I peeled off the buffer itself (which has an edge broken and which is available as a spare part which I must order) and then I put the part of the “locomotive hull” that I have broken off on a toothpick so I can work better with it and so that I can observe when it is straight. A little filing (the material is very soft, so it is pressed “slightly”) and the two parts fit together and are ready to be glued back together:
The two surfaces can be glued together with super glue. At first, however, I tried with a small hammer as well as with the corner of the table, if I could get the corner of the surface on the hull, where the body of the locomotive seems to be bent 0.4 mm in, back into place. But the hammer just destroys everything (I managed to damage the decorative hook on the tow bar as well as the matte black paint a couple of places on the bottom edge and I almost damaged the headlights), and I could not squeeze the hull back with hand and table corner:
In the meantime, I have ordered the buffer itself, which the locomotive must do without until I get it.
Now my trains are again secured against crashing to the floor.