☰

West Exchange Sidings

Problems

This layout was designed with 4 NX panels, one in each corner. This makes operating the layout hard work, either a lot of walking or a lot of shouting. So it has been decided to go to a single panel operated by JMRI, with and optional front panel to operate the yard for shunting. This will involve replacing all the point controls so that they can be operated by DCC using a separate DCC power district.

The documentation for the current wiring is patchy, out of date, and wrong. Even the tag numbering cannot be relied upon. Also some of the D connector to tag strip wiring is different and needs correcting.

Lastly the operation of some of the points using Seep point motors is at best patchy, those near the power supply working better than those at the other end of the layout. This clearly needs to be fixed.

Finally, there has been an issue with bending of D connector pins. This could be due to trying to force 2 male D connectors together, in this case eliminating one set of connectors would help.

Solutions

The 2 short 3ft straight boards at the end of the layout are used for the test track, so need local control as well as remote DCC control. The plan is to use the current control switches on the top of the layout, the single capacitor point CDU's will be replaced by the MERG design dual capacitor design and a MERG DCC steady state decoder module. The MERG CDU's need 24 Volt dc, so need a convertion of ac to dc.
MERG DCC steady state decoder.
Kit 53 firmware in C.

The fiddle yard points, 12 at each end, can be controlled by 6 MERG pulse decoder modules, controlling 4 points each.
MERG DCC pulsed decoder.

The scenic front of the layout, has 7 points and one end and 8 at the other.

1. One possiblity is to use the Cobalt IP point motors. This would make the point movement quieter, but they are fairly expensive at £25 each, 18 for £450.
   There is an issue with the address being forgotten. DCC Concepts reported "We did have a part batch of Cobalt Digital IP point motors that could suffer an address memory loss when a high voltage spike occured on the DCC bus. Most commonly as a result of a short, but in the case of some systems - when they are turned on/off.
   A separate clean DCC power district may help with this.
2. Using the same MERG pulse decoder modules as the fiddle yard, with the current Seep point motors, will be slightly louder but only be £64. When they are back in stock. However, some points are on another baseboard, so using the steady state board and CDU's would be necessary for those points.
3. A third alternative is to use small radio control servos. The small 9g servos are 20 for £24 each. The MERG servo mount is 6 for £4.53 each. The MERG 4 servo control board is £5.53 each. And the MERG DCC steady state decoder module, will drive 8 points for £14.60. For a total of £94 including postage. But we probably also need the servo setting box £16. and the servo turn on current can be excessive and will need some turn on delay for each group of 4 - 6 servos.

Power supply

Inside the box are 2 x 160VA torroidal transformers, giving 4 outputs of 18 Volts at 4.44A. Each of these are connected to multiple pins on a 36 way D connector. Transformer 1, output 1 is connected to pins 1/2 and 20/21, and connected to 5/6 and 13/14 on the next D connector. Output 2 is connected to 6/7 and 25/26, and then to 23/24 and 31/32 on the next. Transformer 2, output 1 is connected to pins 13/14 and 31/32, then to booster U and V. Output 2 is connected to 18/19 and 36/37, then to command station U and V.

There is a Lenz command station and a booster, the booster is probably not needed most of the time. The output 36 way D connector has pins 1/2 command station J. 5/6 AC T1:1. Xpressnet L M A B is connected to D pins 8, 9, 10 and 11. 13/14 AC T1:1. 17/18 booster J 19/20 command station K. 23/24 AC T1:2 Lenz C D E from CS is on 25, 26 and 27. CD out is on 29 and 30. 31/32 AC T1:2. 35/36 booster K.

There are 10 fiddle yard tracks, 10 stationary trains at less than 100mA each, is less than 1 Amp. 6 running trains at less than 500mA is another 3 Amps. So the booster is probably not actually needed for running. What is needed is isolated power districts to allow some operation to continue when there is a derailment. We can have a switch so the fiddle yard DCC can be either from the command station or a booster.

New Wiring

Each baseboard has 2 x 25 way D connectors, connected to a 36 way tag strip. So some connections are on 2 or 3 D pins to carry more current. It looks from the above list, as if the first 4 on each connector may have at least 2 pins.

The MERG CDU's take a short pulse of about 500mA to recharge, the pulse decoders only take about 30mA to recharge and 15mA normally. The CobaltIP's are 5mA and 40mA when moving. LED lights may be 10mA each. So after converting all the points to DCC operation, we can convert the power back to dc. This should then make the signal operation easier by having a fixed 0 Volt line. It should be possible to eliminate one of the 25 way D plug and sockets on some boards.

32/0.2 wire will have a resistance of 0.57 ohms for 100 feet, so fine for the DCC to the rails. 16/0.2 will be 1.14 ohms for 100 feet, so fine for 24 Volt dc which will only carry 1 Amp. 7/0.2 will be 2.6 ohms for 100 feet, so fine for the Xpressnet A and B, or the track section data. The wires to the 20 D connectors and D connectors will increase the resistance a little, 20 x 0.03 = 0.6 ohms, and the DCC pins have 2 or 3 pins in parallel.

The control DCC should come directly from the command station, and must be protected from any rail connection. This is done by MERG DCO's(District Cutouts). Peco insulated frogs can cause shorts as the wheels of some locos pass over them, 54ms cut out delay is faster than the Lenz command station.
Kit 57, MERG DCO.

Tags 14 to 18 should be free, but they may not be. So first we need to free up those tags, then we can rewire while the layout remains operational. However, the D connector wires are not as the above picture on the end 2 boards. While re-wiring the curve boards the booster DCC should be moved to tag 3.

Tag	Desc.	Tag	Desc.	D pins	D pin Colours	Alternate D pin Colours
1	DCC common, Black 32/0.2	19	DCC common, Black 32/0.2	3	Black+Red, Black+Bue, Black	Black+Red, Black+Bue, Black
2	Main DCC, Red 32/0.2	20	Main DCC, Red 32/0.2	3 or 2	Red, Red+Brown, Blue	Red, Red+Brown
3	Fiddle yard DCC, Orange 32/0.2	21	Fiddle yard DCC, Orange 32/0.2	2	Green, Yellow	Blue, Green
4	Signal outer CW	22	FY track 1	2 or 3	Yellow+Red, Yellow+Blue	Yellow, Yellow+Red, Yellow+Blue
5	Signal outer CCW	23	FY track 2	2 or 1	White, Brown	White
6	Signal inner CW	24	FY track 3	1	Purple	Brown
6	Signal inner CCW	24	FY track 4	1	Purple	Brown
7	Block 1	25	FY track 5	1	Orange	Purple
8	Block 2	26	FY track 6	1	Pink	Orange
9		27	FY track 7	1	Turquoise	Pink
10		28	FY track 8	1	Grey	Turquoise
11		29	FY track 9	1	Red+Blue	Grey
12	CBUS High	30	FY track 10	1	Green+Red	Red+Blue
13	CBUS Low	31		1	White+Red	Green+Red
14	Xpressnet A, White TP 7/0.2	32		1	White+Blue	White+Red
15	Xpressnet B, Purple TP 7/0.2	33		1	Orange+Blue	White+Blue
16	DCC control, Green 16/0.2	34	DCC control, Green 16/0.2	1	White+Green	Orange+Blue
17	24 Volts DC, Pink 16/0.2	35	24 Volts DC, Pink 16/0.2	1	Yellow+Green	White+Green
18	0 Volts DC M, Brown 16/0.2	36	0 Volts DC M, Brown 16/0.2	1 or 2	Orange+Green	Yellow+Green, Orange+Green

Where TP is a Twisted pair to reduce interference from the DCC.
The test track oval has the first D pin connections, the fiddle yard and main boards have the alternate connections. These are not compatible so the end fiddle yard and main boards have one of each to convert from one to the other.

Local Connections.
12 Volts for lights, Xpressnet L from 24 volt supply by regulator on each baseboard.
Track sections, may be over a baseboard joint, and need a data bus to JMRI.
Signal control connections, can use DCC control from JMRI.

TOP