News:

Please read the Forum Code of Conduct   >>Click Here <<

Main Menu

Radio Control for HO

Started by Atlantic Central, October 22, 2007, 11:50:23 AM

Previous topic - Next topic

Atlantic Central

#15
Regarding the "cost" of DCC, I spoke with a fellow modeler who uses Digitrax. His layout, which I helped design, has the following stats:

fills a 30' x 60 room
400' double track main line
aprox. 1200' of track total
stagging yard for 25 trains
engine terminal with turntable/roundhouse
freight yard for 6-8 thru trains + 10 track storage
branch line
more than 100 turnouts
more than 60 loco sets (an F7-ABA being one loco set, etc, so it is easily over 100 powered units)

His DCC specs, he uses Digitrax

8 power districts w/power supplies and circuit protection
1 auto reverse unit
stationary decoders for all 100+ turnouts
5 radio throttles (friends bring theirs as well)

He does not have any signaling or computer interface for dispatching, yet!

He indicated that he has spent well into the figure I listed above ($7900.00)!

The only people that don't want you to think it costs this much are the ones who have a vested interest in getting you to use it.

Sheldon

SteveJ

This is a very enjoyable string for a new railroad modeller like myself.  This hobby is served well by Sheldon with his experience and approach to wireless control for his railroad.  This type of thought and determination is very enriching to me.

In my short time 'thinking', 'constructing' and 'thinking' again I realize that simplification is a good mantra when making things work and the pocket book.  I hope you let us follow your development in this subject as it develops. 

Like Mike I anticipate...

Atlantic Central

SteveJ,

Thank you, I will keep posting as the details are worked out. It appears that the computerized block control may be possible with the TE and/or low cost solid state switching may be available to keep the cost of the manual system really low.

Research and testing continues as does construction of layout modules.

Sheldon

Jim Banner

Sheldon, nobody is disputing your friend's estimate of what he has spent on Digitrax.  But I suspect he spend quite a bit of that back when prices were considerably higher than today.  Why do I say that?  Firstly, a layout of that size is rarely built overnight.  Secondly, 8 power districts each with a booster and power supply is old technology from the days before electronic circuit breakers.  At least, I suspect your friend does not run 80 to 100 locomotives at a time and does not need all that power.

I can remember a time when installing Digitrax was expensive.  The first one I installed, the decoders were over $100 each and the command station, with one throttle, left just enough out of a $1000 dollar bill to buy a bottle of coke (we didn't have cans in Saskatchewan in those days.)  At those prices, I wouldn't twitch if you told me your friend spent over $20,000 on Digitrax.

But I came up with a much lower price based on today's prices at Tony's and based on a single 8 amp booster driving up to 8 double headed trains through a multi-channel electronic circuit breaker.  I believe you were talking about up to six trains/twelve locomotives which certainly does not require 40 amps to run.

I noticed that your friend included signals in his price.  I did not.  Did you include signals in your costs of computer block control?  I suspect they may be cheaper with computer control, at least, the software to control them is cheaper than using the hardware approach.  I have no idea what it would cost today to do a layout with  computer block control using commercial equipment.  I suspect it is much more attractive than when it was either industrial input/output cards or build it yourself.

One thing I am wondering about with using TE radio control.  The outputs of the receivers are basically high level outputs, 0 to 12 volts backed up by enough power to run your trains.  This suggests to me high level switching between CABs which is still often done with relays.  On my own layout, 32 blocks times 8 relays per block would have been 256 relays.  Reliable relays at the time I was doing computer block control ran about $10 each.  For that reason, I went to low level switching and used CMOS gates at pennies a piece.  I was wondering how the system you are using connects CABs to blocks.  Or does it use some different scheme?

It is not just new railroaders who are enjoying this thread.  Some of us who have been in model railroading since steam was king enjoy it too. 
Growing older is mandatory but growing up is optional.

Atlantic Central

#19
Jim,

My friend does NOT have signals, he may or maynot do that latter. And the layout has been constructed within the last 18 months. Many of us from the group helped build all the benchwork in only about 6 weeks last year, he did all the track and wiring himself, now a select crew has begun the scenery! He buys everything from Tony's

I do understand the difference in power distribution and did not discuss why he chose so many power districts as apposed to the circuit breakers, but, most of his locos have sound and are on the layout all the time. He and others have had problems with the high current draw of these sound locos. Some are on dead tracks, many are not. Personally, I would have all those things on "dead" tracks until I was ready to use them. I will agree that it does not have to be done the way he did it.

One last thing about DCC, or I should say Digitrax, I personally REALLY dislike their throttles. While I have "learned" to use them, I still dislike them. Not user friendly in my opinion.

I also understand that using the TE with the computerized block control will require power switching rather than the low level switching available with the the throttle cards from Oak Tree Systems, but they also have a 3 amp solid state relay card that is way less expensive than relays, 8 relays per card $35.

My original computerized block control estimates partly included signals because with Oak Tree Systems the detection is built in his throttle card and signal logic is built in the software. Extra output cards and the signals would be the only expense. Even without cbc, I will most likely go to him for signaling do to his intigrated low cost approach. He is doing lots of signal systems for DCC layouts these days!

At this point, since the new layout configuration does not have the large amounts of hidden track as the one orginally planned with the computerized block control, I'm not sure it is needed. The walk around block selection seems more than adequate. especially with larger blocks and multiple locations for asignment. I like the idea of push buttons and pilot lights at the beginning and end of eack block as well as possibly repeated on a dispacters panel.

When you are walking with the train you are throwing the turnouts anyway, asigning the power in this ways seems really easy, dare I say almost intuitive. And, as I said before, I think there is a tendancy to make blocks too small. I think that in most cases you can take what most would consider three blocks and make that a zone and do cab assignments at that level. Maybe still break it up for signaling, but assign the cabs at the zone level.

Next challenge - get the price of that circuity as low as possible.

I don't want to sound too cheap here, I spend thousands evey year on trains, but its just a hobby, and less money spent on one aspect of it means more for some other aspect. And we all know it has never been an inexpensive hobby.

As always your thoughts and experiance are greatly appreciated.

Sheldon

Atlantic Central

#20
Update:

Continued testing has revealed the following:

As I am sure any of you with TE experiance already know, transmitters on the same freq. lock each out out while transmitting, but ones on different freq. do not. This will make it more practical to use a different freq for each cab. As currently designed the layout will require 8-10 cabs, this works out great!

This will also reserve the channels for use with accessory receivers if I work anything out with them.

The Bachmann locos with the inexpensive imbedded DCC decoders ran very nicely once the decoders where bypassed and the capacitors where removed from the motors. Headlights will have to be rewired. Rewiring two locos (or maybe a few more if I buy some Sharks) still beats installing decoders in 150.

The pushbutton cab assignment system is looking more and more like the best answer. Wiring will be moderately complex but equipment cost should be minimal with good shopping.

Sheldon

jsmvmd

Dear Sheldon,

This is an intriguing post.  I had looked at TE years ago as a possible OS for my Aristo Roger's set, and had rejected it since I did not need a wireless throttle system.

Now, as you know, I will be getting a block wired HO layout 11 ft X 22 ft, and I have to choose a control system.  Until you posted here, I was set on DCC.  Now I am not too sure.

I will contact you after I have studied this and have learned to speak more on the subject of control systems.  As a neophyte, you can understand I cannot ask even the most rudimentary question. 

My son is at the local vo-tech studying electronics.  His instructor is an EE who builds all sorts of neat stuff, has his house wired for every kind of control system you could imagine (sort of like "Back to the Future" and "Flubber"), is cobbling up a go-kart from ATV and other components, etc.  A very neat guy who all the boys look up to.  Thus, I will talk to him about this and enjoin him as a local source.

Once I get a few things sorted out I will contact you.

Thank you and Best Wishes,

Jack

Atlantic Central

Update:

I have built a test version of my proposed push button cab selector using relays. Mr Hines (who taught me most of my electrical engineering) would be proud. It works exactly as planned.

I have also fitted the TE's with cooling fans, fused them down to 3 amps on both the input and output, and mounted them to their power supplies as a unit. Makes a nice package and will be easy to mount under the layout.

I am now looking into if the cab selector circuit can/should be solid state or just stay with relays. Quality relays of the type needed are now readily available on the surplus market at very low prices. Price alone may dictate relays since some have been found as low as $1.00 ea, previously a $10 item. That would make a $1200 investment only $120!!!!!!! But even the $1200 is only a fraction of the DCC decoder cost.

I continue testing different locos at different speeds and conditions. No new problems, with the previous imbeded decoder problem being the only thing so far.

There are a number of "warnings" out there about Pulse Widith Modulation overheating motors, but no problems yet. One competitor, who models in N scale, and sells a pulse width throttle, claims his is safer because it has rounded pulses, then he goes on to talk about running a switcher at switching speeds for 5 or 6 hours? I don't think I am up for any 5 hour operating sessions with the same loco. And my experiance is that HO is not N scale.

N scale is a world of its own with those small motors and tight fits. I point this out just to warn those in N scale that my tests do not include any N scale equipment and there may be different risks with N scale. I know at least one N scale modular group did use the TE a few years back and did develop a very sensitive overload circuit for their application. They did not publish any info on motor overheating but did say they used the Pulse Width setting most of the time.

Next I will be testing a few detection systems with the TE. The real test will be to see if a bais voltage for standing detection interferes with the TE, and/or which detection schemes work the best.

Sheldon

Atlantic Central

#23
Update:

I have found a few more locomotive anomalies, none of which is a big deal but I am reporting them just the same.

On pluse with control the Life Like Proto 2000 0-8-0 runs great but the directional headlights do not work. Both front and rear lights come on all the time regardless of direction. They work as designed on the linear setting. This is really not an issue since as a switcher having both lights on is actually more prototypical. Other Proto steam tested did not have this problem.

On linear control, the constant lighting circuit in the Athearn Genesis F units does not light until way after the loco is moving, but it works very well on pluse width control. so this too is a non issue.

As I continue these tests, I become even more comfortable with the bushbutton "fast-slow" method of operation. This was actually the one thing I was most skeptical about but now am most happy about. The smoothness of even fast stops and start is amazing. And, there is a simple but realistic feel to appling power in this way.

Sheldon


jsmvmd

Dear Sheldon,

I can remember some anomalies similar to yours when I  tried to run the first Bachmann X-mas Annie on my Aristo 1.8 amp power pack with PWM.  It ran OK, but there might have been a few "glitches."  I also seem to remember that it was said (perhaps by TOC) not to run Bachmann LS on PWM.  Does that seem right?  Dave, please correct me if I mis-stated.

Thank you.

Best Wishes,

Jack

Atlantic Central

#25
Jack,

I must admit I can't offer many thoughts on large scale. I know some have said some large scale brands/items do not like PWC, but I do not own any large scale and have no personal experiance with it.

The HO problems I have found are all with "circuit boards", not with the motors themselves, so they could all be corrected easily if need be.

And I am very happy with the TE performance and am moving ahead with using it for my layout.

Sheldon

jsmvmd

Dear Sheldon,

I am quite intrigued with your comments, and am very interested with the TE for my application. Will be contacting you soon with details. All my stuff is Athern or Atlas or Proto, with a Bachmann switcher that was run with conventional power and block control. So it would seem this might be up my alley.

Best Wishes,

Jack

Atlantic Central

Jack,

You know how to reach me, and it will be a pleasure to assist you any way I can.

Sheldon

Atlantic Central

#28
Update,

As the testing continues the 3 AMP fusing has protected the TE receivers from several "errors". This is good to know. As suspected, common rail or common bus wiring should be avoided. While common rail does work with the TE if the whole train is within blocks controlled by one receiver, the passing of cars with metal trucks from a block controlled by one receiver to a block controlled by another causes the same speed surges as trying to pass a loco from one receiver to another. And, with common bus wiring and reversed polarity in such cases causes a short.

Completely isolating each block eliminates all these problems. With compete isolation metal wheels/trucks have no shorting effect because only one rail becomes "connected". this has the side benifit of eliminating any chance of gaps bridged by standing cars causing problems, a common problem in the old days of common rail DC.

Another side benifit is the ablity create automatic "dead" zones between blocks by staggering gaps. The staggered area will only be powered when both blocks are on the same cab.

The block selection circuit is fully refined to the fewest number of relays and the application of digital and/or solid state circuits is being explored. Availablity/cost of suitable relays may well be the controling factor. The block selection circuit will provide great flexibilty in operation since it will allow all three types of operation "walk around", "dispatcher" or "display" running for shows.

Having all three types of operation intergrated into the control system has always been a major goal. It seems the combination of the TE with the pushbutton block control will provide that at the lowest cost and easiest, most intuitive operation.

Sheldon




Atlantic Central

#29
Update,

Suitable componants for the relay based version of the block selector have been found at good prices, making any solid state approach the same or more costly and in some ways more complex.

The relay circuit has been refined to the simplest but most flexible circuit and will provide a number of possible options to the installer. These include dispatcher and local control with an unlimited number of local positions being possible, default startup settings if desired, emergancy stop, block over run protection, and panel priority and lockout controls.

The basic circuit is for a 4 cab system but can be stacked or daisy chained in several ways for more cabs.

Developement of a circuit board for the relays will begin soon and more products have been ordered for testing the TE with various detection circuits for signaling/train indication.

Serveral other modelers have expressed various levels of interest in the system and its application to various types of existing layouts is being explored.

Rubber ducky antennas have arrived from Aristo to replace the metal ones. Indoor range is more than adiquite with the rubber ducky. Some large scale modelers find them to limiting outdoors and Aristo is now supplying the telescoping metal ones with the TE, but can provide the rubber ducky for a small charge as a service part. As a future test, I may experiment with an even shorter rubber ducky.

Sheldon