Hello all DCS experts,

The National Capital Trackers will be experimenting with various modules and wiring configurations in an attempt to get a reliable DCS signal around the layout. We are soliciting any and all suggestions on what has proven to work for modular layouts, and what experiments we should take.

The modules are currently wired according to the Tinplate Tracker standards using 4-pin Jones (Cinch) plugs on a 4-wire bus loop.

Our initial experiments will be:
- test alternatives to connecting the TIU to the track - via lighted track lockons or via a track wiring harness.
- use aux power for signal injection or run power through the TIU.
- break the layout into blocks and run a star harness for power feed to the blocks.
- location of balancing lights.

Thanks in advance for any constructive input this forum can offer,

John

Hi John,

I would suggest you start with power going through the TIU to your existing buss. I assume you want more than 10 amps capacity, so try using two or more TIU channels and splitting the bus into two or more sections. Leave out buss jumpers and use insulating track pins on "bridge" sections so the channels are isolated. In fact, your best bet is to install a center rail insulating pin on one end of each "bridge" section on the entire layout. Doing this means that each module is a separate "block" with only one pair of wires to each isolated section of track.

For light bulbs, start with a #1445 or #1447 bulb on each TIU output.

Early reports on DCS seemed to indicate that modular layouts worked well unmodified. This might have been because of the strong and consistent wiring throughout the layout coupled with the possible effects of the connectors. Hard to say.

I'd recommend you set up a "base test" layout configuration first as you need a consistent layout size to make the comparisons. You'll also need to set the variable channels to fixed to handle the two mainlines. As a former Tinplate Tracker, I'm going for the minimum amount of feed sources.

Since Tracker modules feed from the buss, I'd use insulated pins between modules to cut feedback.

I'd start by separating the layout into two equal-sized blocks. From there, the first test would be feeding from the end of the block using one channel per mainline -- a Fixed1 and Variable1 for block A's mainlines and Fixed2 and Variable2 for block B's mainlines. Check the signals at that point to see what you get. Depending on the length of the blocks, you may or may not get signal degradatation or feedback.

Next, I'd add the light bulbs on the TIU outputs and try again.

The next step I'd try is using a pair of terminal blocks -- one for each mainline and feed each from a single fixed channel. Put a light bulb on each terminal block and check your signal strength again.

After getting good baselines, I'd try another layout configuration. You may end up having to split larger layouts into more block sections.

Good luck. Let us know the results.

John

I sent Steve an E-mail. I live near by and have attended meetings on occation.


The Trackers layout was at York and I believe was still running DCS. Carl installed it very early one and actualy had the TIU positive leg in series so they could run 30 amps through an output. It was the early over lay set up we used to talk about but I forgot the name.


From what I have hear most of the modular layouts have worked well on good orginal grid wiring. I think having the center pins pulled really helps in this case.

Jamie

quote:

Originally posted by CSX FAN:
John

I sent Steve an E-mail. I live near by and have attended meetings on occation.


The Trackers layout was at York and I believe was still running DCS. Carl installed it very early one and actualy had the TIU positive leg in series so they could run 30 amps through an output. It was the early over lay set up we used to talk about but I forgot the name.

From what I have hear most of the modular layouts have worked well on good orginal grid wiring. I think having the center pins pulled really helps in this case.


Jamie

Jamie:

What you describe sounds like "passive mode" where the track power DOES NOT go through the TIU, but rather the TIU is paralleled onto the power leads and an aux power input is used to power the TIU itself. That allows full (30-amp) current to go to the layout without frying the TIU while allowing the TIU digital signal to superimpose onto the track voltage. I've tried it that way and it does work with varying results (depends on the integrity of the layout wiring, of course.) The only thing you lose is the E-stop (aka Panic Button) option with the DCS remote.

I've tried "passive mode" I think it works providing you follow Barry's wiring suggestions. Paired wiring, terminal blocks, light bulbs etc.

If you expect 2 wires from the output of a tiu channel connected to the track to super impose the Dcs signal on the track something similar to a Tmcc command base it's not going to work very well. Just my 2 cents.

Tracker John, I think the first step is for use to all use the same brand of track. Not the giant mismatch of all the manufacturers. The second step is to make sure all track and track pins are tight and clean. We need to break the layout into insulated blocks (center rail only). I will be getting together with the other trackers on Saturday. Hope to see you there!

...keep the rails polished...

John,
Here is what the Lehigh Valley HiRailers use.


And in operation...


We have run successfully with up to a 24x52' layout using this configuration and standard 6 pin molex connectors. Nothing special, no breaks in the track. 8-10s all the way around, except near one turnout. (I suspect a bad solder joint.)

Any troubles we have had were due to a semi disorganized multi-operator environment, not the equipment. What is important is to get handhelds, TIU addresses and engine addresses organized BEFORE a show! Things go a whole LOT smoother! (If this sounds like the voice of experience...) I whole heartedly recommend a totally separate "add track" for anyone bringing an engine that is not already "correctly" added in a manner to account for multiple operators. I also recommend NOT using the main layout for adding engines during a show. All kinds of whacky stuff can and does occur.

Chris
LVHR

Chris
LVHR

The Standard Gauge Module Association would like to add DCS to our next display at the November 1st and 2nd Train Fair in Syracuse, New York. Our existing wiring scheme has a track hot, an accessory hot and a common ground for each of the three tracks. The resulting nine wire bus is connected to adjacent modules using nine pin Molex connectors. Currently we connect each track hot to one throttle of a MTH Z4000, its accessory hot to the second throttle and the ground to the ground post. (Each track and its associated accessory hot has its own Z4000 transformer, providing ten amps to each throttle.) This allows us to use sections of insulated outside rails grounded by passing wheel-sets to activate relays for signals and other accessories. The 12 gauge bus wires are connected to the rails at two points on each module. We measured the voltage drop at our last event, and found it to be 0.1 volt for each 100 feet of track length. Each track loop will be about 250 feet long at Train Fair, although this will probably be increased at future events.

DCS seems like the perfect way to allow the operation of multiple trains on each loop. Unfortunately, Standard gauge trains use a lot of power, and 10 amps would typically be enough for only two trains. Our idea would be to divide the track loops into power blocks, providing ten amps of power to each from a TUI. Any pointers that could be provided by this group would be greatly appreciated.

Some of my observations of DCS operations at NCT shows:

1. Sometimes it works great.

2. Sometimes it is extremely erratic with signal levels ranging from 10 to check track. Sometimes the problem areas are mobile and shift to another layout segment during efforts to achieve a better signal.

3. We have observed this many times:

quote:

using the main layout for adding engines during a show. All kinds of whacky stuff can and does occur.

4. Yards and turntables typically have degraded signals (but this is where we want the most reliable control of engines).

5. Passive mode is desirable for high-amp multi-train operation but signal levels are typically better when we run power through the TIU.

6. TIU channels may output different signal quality.

7. Using 100% tubular track does not assure good signals. Some shows with good signal levels are a mix of track types.

So, we want to build a "cookbook" approach to adding DCS to a modular layout and have confidence that, after spending a hectic 2 hours schlepping the gear and setting up the layout, we can power up and operate DCS engines without fussing for another two or more hours.

After considering Frank Maguire's discussion of signal transmission issues, we have challenges.

John

John, thanks for the link. Frank Maquire's description of the dcs operation is very helpful in understanding what is going on. Every post I read on this subject builds my understanding, so thanks to all. I still find it interesting how complex these trains are becoming. I hope this does not lead people to stay away from command control reading these posts. Joe P.

quote:

4. Yards and turntables typically have degraded signals (but this is where we want the most reliable control of engines).

Add tiu, Trying to run everything from one tiu is going to be difficult at best.
In my opinion "Super mode" is the answer, You can have up to 5 Tiu, that's 20 channels, Each channel has it's own signal generator.

Try using a tiu channel just for the turntable stalls.

John,

I am currently working towards a "cookbook" solution for us as well.

Two tries so far have yielded less that satisfactory results. But I plan to keep trying.

Here is a thread about changing the wiring, which of course led to a lot of DCS signal discussion.

http://ogaugerr.infopop.cc/eve/forums/a/tpc/f/57660482/m/3601053174/p/1

The above referenced signal transmission thread was also very useful.

Chris,

Is that a surge arrestor with the green lights?

Bill,
Yes. It's a 4 channal Frank Qualls special. Check the HiRailers Unlimited website. I have each track and the acc voltage protected.

The TIU is wired for pass through on the Red side only. The Black side (common)is tied together with a heavy brass bar on the output side, so there is only one wire feeding the common side of the layout. This way, you don't blow the fuses inside the TUI!

Wire is all 14ga stranded.

Chris
LVHR

The National Capital Trackers team of Steve and Matthew K. (L and N), Frank H., Watson M., Bill R. (Farmer Bill), Bill C., George T. (GET Tracking), and Regis H. spent a hot Saturday in a non-air conditioned hall experimenting with DCS using a simple rectangular modular layout configuration.

The good news is the team developed what promises to be our cookbook approach for achieving high DCS signal levels on a modular layout. Their report follows.

I want to congratulate the NCT team for a lot of hard work and persistence. Well done, guys!

I also want to thank this forum for an excellent technical discussion of DCS signals and suggestions on what to try in our experiments.

I hope that other modular clubs can try the NCT approach and confirm that it works for other layout configurations. This experiment did not include T-modules, yards, and turntables. However, NCT shows in the coming months will provide us an opportunity to determine if our approach succeeds for varied layout configurations.



National Capital Trackers
DCS Project
28 June 2008

The problems we have encountered with DCS include:
1. The TIU doesn’t see any engines, or the engines do not see the DCS signal, causing Check Track or Engine not on Track.
2. Losing control of an engine as it navigates the layout
3. Can’t add or delete engines
4. Can’t make any modifications to engines once control is established. Smoke, sound, etc.
5. Erratic Track Signal, 0-1-2 or nothing

We decided the only way to address the problem was to build a test layout for the sole purpose of trouble shooting the problem with a goal of at least improving the Track Signal strength. Our modular club adheres to the Tin Plate Tracker standards for module construction. Two main loops of track for each module connected via steel track pins and a wiring harness using 4 pin Cinch plugs for connectors.

The layout we constructed was 18 feet x 32 feet in a rectangular configuration. This approximates a typical NCT layout. The track on the modules was also typical ranging from Lionel O gauge to Atlas, K-Line, and GarGraves. Several switches were also evident, the old Lionel variety.

The main troubleshooting tool will be the DCS remote using the Track Signal option to test signal strength around the layout.

A) The first step was to set up the layout just as we do at any typical show this meant connecting the Track Power directly to the Buss wiring harness via a pigtail connector (utilizing male and female Cinch plugs.) A single MTH Z4000 transformer was used connecting both main outputs. One for the outer loop and one for the inner loop. Track power was not routed through the TIU. The power for the TIU is provided by an auxiliary power supply with the two fixed outputs of the TIU connected via another pigtail harness in a similar manner. The Buss was continuous with lights placed across the posts of both fixed TIU outputs.


Results: Both loops were checked using the Track Signal option. The same engine, running at approximately 3-4 mph, was used on both the inner and outer loops. We were not disappointed. The strength of the Track Signal varied from nothing to 1-2. We duplicated the problem, now to fix it!


While the layout was being assembled we looked for any bent track or loose fitting track pins and repaired them. Once the first pass of Track Signal checks was completed we cleaned both loops of track using several Trackman 2000 track-cleaning cars, plus wiping the track with an isopropyl alcohol cleaning fluid. The track was dirty but retesting showed no visible improvement.

B) The second step was based on input from the OgaugeRR Forum. We kept the power and TIU configuration the same, but removed the module-to-module center track pins and replaced them with fiber pins essentially eliminating the mechanical track-to-track path for power and the DCS signal. The only electrical path being the wiring buss. We started by breaking the buss and adding modules one at a time testing the signal strength as we proceeded.

Results: Initially the signal looked good, 8 to 10’s, but when more than 6 modules were added the signal deteriorated to 1-3’s. We attempted to “tune” the layout with bulbs connected to the buss using specially built test Cinch Male and Female Connectors with a bulb across the white or outside power wire. This was tedious and time consuming but achieved acceptable signal strength. Moving the bulb from one end of a module to the other caused significant signal strength degradation. Although satisfactory we ruled this out, as each set up would require tuning and the time required would be prohibitive. It is also noted during the testing process we improved the signal strength on the GarGraves modules by burnishing the center rail. Readings improved from 1-4.

C) We entered the third step confident that we could break the layout into four sections and provided separate feeds in a star pattern for the DCS and power connections. The Z4000 and TIU connections remained the same, but with four parallel feeds. We terminated the buss again with light bulbs across the end of the respective four busses.

Results: Track Signal testing commenced, but with no improvement. The best Track Signal we managed was 1-3 or nothing. We all were a bit frustrated at this time. Our 5 + hour effort had nothing to show for it. A lunch break was in order!

D) During our break we discussed our T/S plan and the apparent lack of any significant progress. We expected at least some improvement. So we came up with a few modifications to our plan. We did away with the star pattern cabling, but left the fiber pins intact between the modules. The wiring Buss was reconnected normally, but at the halfway point around the layout the Buss was disconnected. The big change this time was disconnecting the Z4000 feed and routing the two main Z4000 outputs through the two fixed ports on the TIU. We also left the auxiliary power supply in place.

Results: Testing began and to our surprise we were seeing 6-8’s around the entire outside loop. The next step was testing the inside loop and again 6-8 on every module. Could this be our breakthrough? At the Buss break we put light bulbs across each feed; we did this on both ends of the Buss cable. To our amazement we immediately saw steady 10’s on each module, both inner and outer loops. Just to see if the bulb placement was critical we reconnected the Buss and made the break at 25% point from the TIU connection with no change. Still steady 10’s. The process was repeated in the other direction with the same results. It looks like bulb placement is not significant, as long as it’s at the wiring Buss break. We even added additional bulbs across both fixed TIU outputs with no apparent effect. These bulbs were removed. Armed with our success we added several additional engines and saw no degradation of the signal or operational problems. We added five18 inch passenger cars all with working light bulbs, again no degradation. We even added a car with a blinking EOT light without incident.

E) The next step was to check each of our TIU’s in this working configuration. We had four TIU’s for testing .Two were Rev I3/4.0, one was Rev H1/4.0 and the last one was Rev G/3.1.

Results: Surprisingly the software level made a definite difference. Each of the 4.0 boxes all returned 10’s while the 3.1 level ranged from 2, 4,6,7,9 in the exact same hookup. The Virginia TIU was at 4.0 and the Maryland TIU was at 3.1.

F) We then reduced the size of the layout by 4 modules
resulting in an 18x24 configuration. Retesting did not show any changes.

Results: Still steady 10’s on both loops.

The NCT DCS Wiring Diagram is posted in Photo Albums.

Conclusions: Isolating each module proved effective. The fiber pins between modules, center rails only, seemed to eliminate any degradation caused by track type or track pin connection. The fiber pins and breaking the wiring Buss simplified the wiring path for the DCS data signal by eliminating multiple paths. Adding the light bulbs at the both ends of the Buss showed a marked improvement in signal strength. The light bulbs used were #53 a 14.4v and 12v.

The biggest improvement came from routing the track power through the TIU. With our modular layout the auxiliary power supply alone was not as effective. By keeping the auxiliary supply connected, even when routing of track power through the TIU, we could be assured the TIU would stay powered up even if track power were turned off.

The biggest surprise was the difference in signal strength relative to the different software level of the TIU. The latest level of 4.0 makes a big difference. Along with this observation the level of the DCS remote is also important. The older remote software level works fine but with a noticeable delay in engine response. Level 4.0 in the remote and TIU had an immediate response.

Respectfully Submitted by the NCT DCS Project Team


**************

Thanks again to the NCT team and the forum for working to develop a DCS cookbook solution,

John

Congratulations to John and all the NCT team!

It's nice to see you're taking such a methodical approach and documenting your results. Your improvement in signal strength with V4.0 vs. V3.1 is similar to what I saw on one large layout I used during beta testing. Areas that were flaky with 7,8,9 turned into 10's.

If you're interested in trying another configuration of "passive mode," I'd suggest you get some heavy amperage inductors and insert them into each wire of the pigtail used between the buss and transformer(s). That will isolate the transformer windings from the DCS signal. I've tested this on two layouts now and the results have been positive. However, neither of these layouts was over 300 sq. feet.

Hi Dave,

I'd like to try that so we can run high amps for multiple trains. A couple of questions:

Why is an inductor needed to isolate transformer windings (an inductor) from the DCS signal?

Do you have a part number for the inductors or should I thread the 12g power feeds through a ferrite donut?

Thanks,
John

Hi John, Good stuff, I'm curious how many tiu are you using? 1 for six modules?

The var channels can also be set to fixed mode. I'm sure you already know.

The club uses a single TIU for show layouts driving two main lines. Typically both of the fixed channels are used. The total number of modules in the loop is rarely less than 20. The DCS team conducted their testing on a 16x32 layout which is 4 corners plus 16 straights.

We see variation in DCS signal strength across TIUs and channels.

FWIW, my personal TIU has the strongest DCS signal out of Variable 1 in passive mode.

Tracker John:

Glad things worked out. Your wiring diagram tends to validate a theory I've had about how DCS works. Your diagram is virtually identical to the old "Thin Net" Ethernet computer networking configuration -- terminators on each end with the "computer" taps in between. Now I'm curious to see how far a buss can be run without severe signal degredation. Congratulations on your success and I admire the approach you guys used in the testing.

Hi Matt,

The Thin Ethernet analogy has me running screaming with loathing. Those were NOT the good old days ,,, workstations would randomly loose connectivity depending on which group of nodes were posered up. We always carried 50 ohm resististors with us and had to install Ts to add resistance along the line. (Sounds a bit like DCS.) Aaaargh!

John