Monday, April 20, 2009

Hiatus

I need to take a break from the hobby. I need time to think. Something happened and changed my life and I lost enthusiasm for the hobby, and now I have to concentrate all my efforts on my family. Not health reasons, so no need to worry.

I know few of you follow my blog, all I can say is there will be no updates for a while, don't know how long this while will be. Thanks for understanding.

Friday, February 13, 2009

Winter Layout Party, week 7

Final day! February 13th is Trainboard Winter Layout Party deadline. I haven't done much during this last week as I feel the scene is pretty complete. It's also been a crazy week at work.
I can say mission accomplished! The scene started as bare benchword and ended up 90% scenicked.

Here is a before and after comparison:

Sunday, February 08, 2009

Winter Layout Party, week 6

One weeks to go! Final rush for the Winter Layout Party at Trainboard.com! Deadline is February, 13th 2009. This week I added vegetation to the horseshoe curve hills, I extended the static grass fields area towards the Callender spurs and added little details like Atlas telephone poles as described in an earlier post.

I started by adding some trees and bushes here and there. I made my trees out of Zeeschuim (aka Teloxis Aristata), a product very popular among modelers in Europe which is very close to the Super Trees line offered by Scenic Express in North America.
I explained how I made these trees in my previous layout's blog. The only differences this time around are:

  • I painted the bare tree gray with a spray can (this is a very important step. Zeeschuim tends to have a light tan color, not actually realistic, see the photos on my previous layout link above);
  • after gluing the foliage, I sprayed the tree a black color from underneath to simulate shade;
  • then I added a bit of yellow spray paint from above to simulate sunlight.
This subtle steps result in much more realistic trees. Here a short video:


SP Coast Line in N Scale - More Scenery from Denny on Vimeo.


Too much talking, onto the photos! In case you are wondering yes, most of the following are Helicon Focus manipulated images, I am addicted already :)

Wednesday, February 04, 2009

Fine tuning DCC decoders for smooth running

I found out (at my expenses) that the lack of B-EMF on cheap DCC N scale decoders results in poor slow speed running. This is especially true on my Intermountain Railways SD40T-2 and SD45T-2, aka Tunnel Motors.

From that moment on, I got myself B-EMF equipped decoders exclusively. However, factory settings does not seem to satisfy my hunger for smooth, creeping slow speeds. Here are some CV setting I figured out by trial and error. Setting provided for TCS and Digitrax decoder.

TCS Decoders
- First off, I make sure B-EMF is active
CV61 = 1

- Next, I add acceleration and deceleration rates respectively:
CV3 = 4
CV4 = 4

- Then, since my command station supports it, I use the 128 steps loadable speed table. Then I disable analog mode operation, I don't plan to use my engines on DC, so I don't need it; some says analog mode sometimes could lead to weird behavior in DCC, for example when track is a little dirt.
CV29 = 18 (128 speed steps + analog mode off)
CV29 = 22 (128 speed steps + analog mode on)

- I programmed the speed table and played around with values of CV from 67 to 94. I usually start with CV67 = 2 and gradually end up with a CV94 = 140. This will reduce top speed and give smoother performances at low end.
Here is my typical loadable speed table:
CV   Desc     Value
67    Step 1    2
68    Step 2    4
69    Step 3    6
70    Step 4    8
71    Step 5    10
72    Step 6    13
73    Step 7    16
74    Step 8    19
75    Step 9    23
76    Step 10    27
77    Step 11    32
78    Step 12    36
79    Step 13    40
80    Step 14    45
81    Step 15    50
82    Step 16    55
83    Step 17    60
84    Step 18    65
85    Step 19    70
86    Step 20    75
87    Step 21    80
88    Step 22    85
89    Step 23    90
90    Step 24    100
91    Step 25    110
92    Step 26    120
93    Step 27    130
94    Step 28    140

For some weird and unknown reason, starting with CV67 = 1 and increasing the following CVs values by 1, end up in having the engine running at speed of light when throttle is set at speed step 1, especially when gradually slowing down from a higher speed step on the throttle. I have to further investigate.

Please note that Dither (CV56 and CV57) is disabled when B-EMF is on! TCS does not provide fine tuning setting for B-EMF, they state it is self-regulating. Read more on TCS B-EMF (PDF document).


Digitrax Decoders
First check if the decoder has B-EMF. For example, the cheap DZ123 doesn't have it and I would rather not use such a decoder for my N scale engines - been there, done that. However Digitrax came up with DZ125 which has B-EMF and it's even a little smaller than DZ123!

- I begin by programming start voltage, acceleration rate and deceleration rate. I program 0 in CV2, because fine tuning B-EMF settings leads to better results, more on this later.
CV2 = 0
CV3 = 2
CV4 = 2

- Then, I fine tune B-EMF settings, playing with CV55, 56 and 57. The Digitrax manual is not very clear when it comes down to these settings, however what I figured out is:

  • CV55 controls how much compensation the decoder needs to apply when sending the next speed command to the motor, for example under load while climbing a grade. Low values give less intense reaction, and high values give more intense reaction.
  • CV56 controls how fast B-EMF should change from one speed step to the other. Low values here permit a smoother transition between speed steps, even when the command is issued by the throttle and not only by the decoder itself through B-EMF.
  • CV57 controls B-EMF intensity, Again the manual isn't clear, and since Digitrax tends to stick to HEX values, it wasn't that easy to find out that the best value for this CV is decimal 085 (or HEX x55). This value means that B-EMF is on when the loco is run as a single or within an advanced consist. I tend to have B-EMF also when I run engines in consist of 2 or 3, since I never experienced the most common problem of B-EMF consists where engines work one against the other.
Here are my values - these settings take care of starting speed as well:
CV55 = 128
CV56 = 20
CV57 = 85

I find these few Digitrax motor settings enough for my taste, so I won't use a loadable speed table and disable analog mode operations, thus programming CV29 as follows:
CV29 = 2


The above settings are not supposed to work as you expect on every engine, you are encouraged to experiment and hopefully to share your results!

Thanks to Knut Tysnes for pointing out FX3 confusion.