Video Gallery  


Who's Online  

We have 118 guests and no members online


Tuesday, January 9, 2007

Computer Control - Painful first steps

frontpage hit counter

Under construction

I am a computer guy, it is all I have ever been. DCC & the realism of today's QSI sound systems pulled me back into the hobby I left some 30 years ago. Read any write up on DCC in any magazine or on the web and real quick, you get this picture of the PC and DCC integrated, sort of a symphony of beauty and control. From the beginning, computer control has been central on my mind.

Actually Lenz helped me a great deal with these notions. How, very simple, my first DCC set was a Lenz SI. If you took the time to look it up, you would find that I did not purchase a system with a throttle, but one with a serial port, and software included. I hooked it up, plugged it in, and installed the software on my computer and I was running my trains with my mouse in less than an hour. Oh, was I pumped. I am computer integrated already! The next day I was completely wireless running my Atlas Midland Central layout with my Ipaq, way cool, oh, am I in deep.

Ultimately, I evaluated four different products, KAM, JMRI, Railroad & Company and CTI-Electronics. I settled on CTI as my solution. Today my trains are automated using CTI. It took over 13 months to make it happen. To be clear, it took me over 13 months, from November 2005 until early January, 2007, to make a locomotive behave like a trolley under DCC control.

My criteria
My initial test suite was simple. I wanted to run a locomotive as a trolley. Simply Go back and forth on a single piece of track, waiting different amounts of time on each end of the track and blowing the whistle upon arrival and departure. I thought it was going to be cake, a very simple request. I never dreamed it would be as big a deal, and expensive as it turned out to be.

First stop, KAM
Since the software Lenz sent me was Train Commander by KAM, it was natural and expected that would be my first software experience. Train Commander is a nice throttle and the network features built into KAM make it a VERY attractive product. KAM's website, it turns out is as confusing as its documentation. While this is unacceptable, it is also the norm for all of the four solutions I examined.

KAM's Railroad Commander was the product that was suppose to do what I wanted. Certainly it took some time to get feedback working on my Lenz, navigating the LR101/LB101 is not trivial. Wiring DCC for block detection feedback is more complicated than DC blocks, belying all of the DCC is simple to wire nonsense. But I digress, back to the software issues.

I can tell you that Railroad Commander is a pretty nice automated CTC panel, reasonably straight forward to get working. My only issue was that I could not get the contrast and intensity of indicators on the screen as bright as my old eyes required. KAM does not rely on scripts to work. Their GUI reminds me of a level of complexity found in intermediate Access programming and about as well documented, I might add.
The program was originally developed in German and I had German translation messages pop up. I found errors and confusion in the documentation. I was never able to get any level of automatic operation. After two hard months of working on this software, the bubble was burst, I declared model railroading was not ready for computer automation. Also, as a piece of trivia, the KAM Yahoo support Forum is highly moderated.

Railroad and Company
As I was winding down my KAM effort, RR & Co looked like a good solution. It has the best documentation of the four offerings, excellent, no, the best advertised capability and a really awful software trial arrangement. The 'quick timeout' just turned me off. The price is very high, around $500.

To clarify my comments about the price of RR & Co, directly from their website 3-8-07.
The basic software is $349, add $149 for network and $149 for sound.

Couple the price with the trial approach and I bailed out. If they would have given a nibble approach, get me hooked at an affordable price, I might have been an RR & Co user today. But the risks are just too high, I'm not about to spend $500 for another piece of software that just fails to meet the claims. I have already reserved that money for Bill Gates.

Perhaps the best support and certainly the most loyal fans belong to this solution which is certainly priced well. Decoder Pro is a viable, perhaps best of breed decoder programmer. I have used it successfully and certainly can see where a club with many locomotives and visitor would swear by this product. I find the documentation to be confusing and somewhat ambiguous. JMRI is for those who have a high geek content. This is one solution that can and does run on Linux, if that is what you'd like to do. Their Panel Pro is a CTC solution and I know that they can automate train movement. Their Yahoo support forum is extremely active and I'd say one of the best around.

Recently on the JMRI yahoo forum, two messages stand out in my mind.

  • Where is a tutorial that goes step by step and does not confuse me like the documentation?
  • Does someone have code that simply automates a trolley?

The answer to the second question was a link to several pages of code. That pretty much shut down any further interest I had in that solution. I may be back to it, but for now, I am not ready to be 'that' deep into a product.

I stumbled into CTI while looking for C/MRI (Bruce Chubb) information. Confusing web page, I understood their DC automation, signaling and feedback information right off. But understanding the tie to DCC was a real challenge. Not well explained except they say it works together. (I'll explain exactly what CTI can and cannot do with DCC in another post. ) By now, I am from Missouri and everything has to be well understood! The CTI Yahoo group is pretty good, heavily moderated, but very good. Took me a day each time to get an answer and frankly the users from Australia were able to get me good answers to my questions. Their Starters kit is 99 bucks including their software and hardware, it is enough to automate a trolley. (It took less than 10 lines of code.)

Actually, I was able, in less than 8 hours to get a trolley operational running my DCC system from TrainBrain (their software) without using their hardware! Finally, after four products I hit pay dirt. At the moment, I have three trains automated with only 8 sensors allowing them to operate smoothly, blow whistles, bells, stop at stations, run the turnouts and I am a happy camper! I have less than 24 hours of time working on it to get it to the point where I am. As a note and to be very fair, CTI gets the benefit of all of the expertise and learning that I expended with the other products along the way. Looking back though, CTI was just much easier to understand and ultimately get results with. A fine place to start from my perspective.

Click to join cti_users

If you want to spend an interesting but meaningless couple of hours look into the legal stuff between KAM and JMRI. One could make the argument that a direct focus on the product details and documentation would yield better results than the distractions of legal discussions. I'm a free market guy and the best concepts and products win the day.

After almost 18 months using CTI's Train Brain software I'm satisfied with the progress that has been made. I routinely operate my layout running 5 trains simultaneously. But all progress comes at a price. When Lenz offered their new 3.6 upgrade, I sent my system in for upgrade. As I have a spare system because I like to have extra parts on had for occasional troubleshooting, when the first system came back, I sent the second one off. It never occurred to me that I should test my layout with 3.6 as I was consumed with building the new valence and facia.

Testing with 3.6 proved to be troublesome, CTI looses sync with the Lenz system and my locomotives occassionally stop responding to the computer. So I'm temporarily impacted until CTI gets their software back in-sync with Lenz. A check of the JMRI site suggests that JMRI users are also experiencing 'issues' with 3.6 as well. JMRI and CTI are now in a corner, as Lenz has not yet released the internal 'specs' for 3.6.

I decided to step around this set of problems and make the leap to NCE's new Radio system. It was all the inertia I needed to make the purchase. Well, nothing is that simple, the NCE USB is simply not ready for prime time, it works 'funky' with JMRI and CTI has not yet added NCE USB to its code set. So I jumped over to NCE serial port which comes built into the command station only to find more serious issues there.

So here I sit, waiting for the Lenz situation to be resolved, or hope that the NCE issues can be put to bed. In the meantime, I'm running with Lenz and tolerate the highly intermittent issues. Take a look at my entry on my new NCE Radio system to read about my pros and cons of Lenz and NCE.

UPDATE on Lenz  - 10/07/2014


The problems I was having with Lenz 3.6 were NOT caused by Lenz.  Programming TCL was my first experience with real time programming and the issues were caused by poor code design on my part.  
The current problem with Lenz is with the 23151 in particular which for the moment is not recognized by TBrain.  Our patience with CTI will be rewarded soon.  The LI-USB and the Serial port adapter L-101Fwork and have worked all along with TBrain.  There are two versions of the LI-USB and getting the correct drivers from Lenz has always been an adventure.
The 23151 problem is particularly acute as the older devices are no longer available from Lenz and finding one used turned out to be folly for me.
Hope this helps 

More as I learn.


All rights reserved, Copyright 2014, Joe-Daddy.Com



Copyright © 2019 Joe Daddy's Weblog. All Rights Reserved.
Joomla! is Free Software released under the GNU General Public License.