Weekend Update 1/8/17

[Angus Shops]

I ripped out another chunk of the old layout in preparation for the move. I must say, I built some solid bench work all those years ago! About 3/4 of the layout has now been demolished. So far my efforts to salvage the portions of the layout that include the more complex trackwork have been successful, with some minor, repairable, damage. However, I'm doubtful about the Field townsite; this scene is 18' long so it would need to be cut in two through about a dozen tracks. I'm not sure I can do this neatly enough to get it back together properly. Also, this was the first part of the layout constructed 18 years ago, so it includes my earliest hand built turnouts, some of which probably aren't up to current standard. So, even though I'm happy with the arrangement of tracks, I'm thinking I'll scrap it and start again...

So the layout has been reduced to 4 piles: the salvaged sections, garbage bags full of demolished scenery (less than I thought!), lumber cut to firewood length, and scrap wood not suitable for the fireplace.

Geoff

[Dave V]

Another Wild West Models structure begins to take shape in Aspen on the Colorado Midland. I intentionally modeled the paint as if it were peeling.  Note the extra tall first floor of the DPM buildings (common for the era) versus the presumably older wooden structure to the right.

[VonRyan]

Installed a decoder in a Key Imports PRR T1 today.

[cbroughton67]

My other project this weekend was finishing off the last two of a trio of Arnold U25C's. I repainted them into a later (and more accurately-colored) L&N scheme, and added DCC, a working nose headlight, and a working Gyralight over the cab. I used a Digitrax DZ146IN decoder, since it has 4 functions and fits the plug on these locomotives, but I prefer TCS decoders for light functions, because I think they operate LED's much better than Digitrax.

[ Guests cannot view attachments ]

[Mike Madonna]

Installed a decoder in a Key Imports PRR T1 today.

Very interesting... any details??

[peteski]

My other project this weekend was finishing off the last two of a trio of Arnold U25C's. I repainted them into a later (and more accurately-colored) L&N scheme, and added DCC, a working nose headlight, and a working Gyralight over the cab. I used a Digitrax DZ146IN decoder, since it has 4 functions and fits the plug on these locomotives, but I prefer TCS decoders for light functions, because I think they operate LED's much better than Digitrax.

Very true!  I was hoping that finally Digitrax's 6th generation decoder firmware (like the DZ146) would finally have improved LED effects, but they still work very poorly. And according to Digitrax's documentation they did supposedly made the 6th gen. decoders more LED friendly. 

[chuck geiger]

Working on Ellis Union 76 in Wenatchee. This area has been several scenes on the layout. Wash of
Pavement Black and Hippo Gray acrylics. Test fit tanks. A new chain link fence with small Evergreen
rods for posts and bars with Kirkland brand (Costco) green tea bags mesh for fencing.

[cbroughton67]

Very true!  I was hoping that finally Digitrax's 6th generation decoder firmware (like the DZ146) would finally have improved LED effects, but they still work very poorly. And according to Digitrax's documentation they did supposedly made the 6th gen. decoders more LED friendly. 

I can't tell any difference between the Digitrax decoders from 12 or so years ago and newest ones of today with regard to light functions. They were and apparently still are designed to work best with incandescent bulbs, not LEDs. Supposedly, CV62 (keep alive rate) can be set for LEDs, but even after setting the first digit to 0 as recommended for LEDs, the impact is marginal at best. With bulbs, you only have to vary the voltage to change how bright or dim the bulb appears. With LEDs, it's totally different - they have to be pulse-width modulated to vary the brightness. An LED will light up if it smells an electron nearby, so varying the voltage has little affect.

[garethashenden]

I found a kit in a drawer. Half completed, superglued rather than soldered, and painted the wrong shade of brown. All done by me circa 2009 and left forgotten since then. I found it a few weeks ago and decided to try again. Following advice from here the kit spent about 18 hours in a jar of acetone which did an excellent job of removing the superglue and most of the paint. I then cleaned up the brass and set to work putting it back together again. I have now got it just beyond the stage that it had gotten to, but I have done a much better job of it this time.
Southern Railway (UK) "Queen Mary" Bogie Brake Van:


Hopefully I'll be able to finish it in less than eight more years...

[peteski]

I can't tell any difference between the Digitrax decoders from 12 or so years ago and newest ones of today with regard to light functions. They were and apparently still are designed to work best with incandescent bulbs, not LEDs. Supposedly, CV62 (keep alive rate) can be set for LEDs, but even after setting the first digit to 0 as recommended for LEDs, the impact is marginal at best. With bulbs, you only have to vary the voltage to change how bright or dim the bulb appears. With LEDs, it's totally different - they have to be pulse-width modulated to vary the brightness. An LED will light up if it smells an electron nearby, so varying the voltage has little affect.

Digitrax's Mobile Decoder Manual 2.0 (from http://www.digitrax.com/news/2014/02/02/digitrax-decoder-manual-20/ ) specifically states:
Series 6 decoders have optimized LED and lamp lighting algorithms, have improved scaleable speed stabilization (Back EMF), work with Digitrax Power Xtenders to improve performance in the presence of power interruptions and have configurable FX3 pulse function on all function outputs.

But we both know that they still haven't perfected those LED lighting algorithms. 

As far as the functions (and also motor control) outputs go (this is for all brands of DCC decoders), they all use PWM (not analog) outputs.  Decoders do not have true analog variable-voltage outputs.  But incandescent bulbs have different voltage/current characteristics than LEDs and they have very slow response compared to LEDs.  LEDs are also more efficient than incandescent bulbs so the PWM pulses need to have very low duty cycle to make the LED appear to glow very dimly.  For some unexplained reason Digitrax has not figured that out yet. The problem is in the Digitrax firmware code because the hardware they use (chips) is very similar to other manufacturers (who figured out how to properly dim LEDs).

[nkalanaga]

Jason:  Very good find.  I wonder if they were really "standard", nationwide, or if this was a VGN standard, and other roads varied slightly?

[up1950s]

Late Christmas gift to myself 

HOn30 14ton Sterns-Heisler.

Dainty lil guy .

I would like its big brother in N . Why the easiest logging loco hasn't be done  suxs .

[Chris333]

Dainty lil guy .

If I counted correctly there were only six 14 ton Heislers ever made and only one was narrow (36") gauge. I like earlier stuff for my narrow gauge. I was looking for a HOn30 18 ton Climax and found this instead. It is smaller so it fits my needs perfectly. I like em' dinky.

[tehachapifan]

Digitrax's Mobile Decoder Manual 2.0 (from http://www.digitrax.com/news/2014/02/02/digitrax-decoder-manual-20/ ) specifically states:
Series 6 decoders have optimized LED and lamp lighting algorithms, have improved scaleable speed stabilization (Back EMF), work with Digitrax Power Xtenders to improve performance in the presence of power interruptions and have configurable FX3 pulse function on all function outputs.

But we both know that they still haven't perfected those LED lighting algorithms. 

As far as the functions (and also motor control) outputs go (this is for all brands of DCC decoders), they all use PWM (not analog) outputs.  Decoders do not have true analog variable-voltage outputs.  But incandescent bulbs have different voltage/current characteristics than LEDs and they have very slow response compared to LEDs.  LEDs are also more efficient than incandescent bulbs so the PWM pulses need to have very low duty cycle to make the LED appear to glow very dimly.  For some unexplained reason Digitrax has not figured that out yet. The problem is in the Digitrax firmware code because the hardware they use (chips) is very similar to other manufacturers (who figured out how to properly dim LEDs).

Well, the gyralight effect I have programmed to an output on the drop-in Digitrax decoder for the Atlas MP15DC works quite well. Plenty of difference between the bright and dim phases of the effect much like my Zimo's. There even seems to be a slight reflector flash effect upon every other bright phase. Could the resistors being used have an affect on the quality of the effect?

[daniel_leavitt2000]

Dainty lil guy .

I would like its big brother in N . Why the easiest logging loco hasn't be done  suxs .

I always liked Heisler's symmetric and balanced design. I wonder if they were more reliable than Climax or Shays.

I would think that this engine would be a lot easier to make in N than the Atlas Shay. Even the running gear can be exposed and look prototypical. Looking at photos, there sure were major variations. I think the closed cab, boxed frame version would work best for a model: the frame can hide some of the mechanism and provide weight down low. Motor can sit in the coal bunker and cab. Sprocket gear from the motor to the drive shaft. In fact, this design was used for one of the very first steam engines ever in N: the Arnold 0 series BR89 0-6-0. The truss frame and sub frame versions would be too top heavy I think.