This installation of TCS T1 and DP2X-UK Decoder is for HO Scale Mehano Thalys T675 and was performed by Phil Dennison, Aliso Viejo, California, USA.
I believe that Mehano sold three versions of the Thalys model. A basic version with no interiors in the cab or carriages and plastic pantographs, a DCC ready version with interiors and optionally powered metal pantographs, and a DCC sound fitted version. Two packs of three intermediate carriages were available to match, with interiors, which allowed construction of the complete prototype train.
The first of these was sold both as a four vehicle pack and in a train set with track and transformer. I think some of these included two powered locos.
The version I have is the DCC ready four vehicle train pack (catalogue number T675), with one powered and one dummy loco. The powered loco needs DCC control of motor and front/rear lights, the dummy just needs lighting control.
This photo shows a power car before work has begun. The powered and dummy cars are almost identical from the outside. Under the set number 4321, in very small writing, the powered car is numbered 43210 and the dummy 43219.
Body removal is very easy, unscrew the screw shown by the green arrow. The chassis moves slightly forwards as it comes out of the body, there is a small lug on the rear.
The blanking plug has been replaced by a DP2X-UK decoder. There is plenty of space, so a DP2X would work just as well. The green arrow marks the lug on the rear of the chassis which must be located in the body for re-assembly. The wires connecting to the roof are for the option of power via the pantograph. The switch for this is inside the body, and so unlikely to be moved by mistake, so I decided to leave the wiring in place.
Each truck/bogey had a capacitor connected across the pickups - the orange disks marked by green arrows. I cut these out in preference to unsoldering, as the soldered joints are a little awkward to access. NOTE: These capacitors MUST be removed for proper isolation. Leaving them intact will corrupt the DCC signal and cause regular failures.
That completes the 'DCC ready' part of the process. The dummy loco is not fitted with a decoder socket, and so must be hardwired. This is complicated by the fact that most of the LEDs are surface mounted on a small printed circuit board.
Here the PCB has been pulled out of its location, it was clipped under the transparent part at the top of the photo. For this view I have unclipped a section of the underneath of the body. This is unnecessary unless the PCB is to be modified. On the board there are four LEDs, two yellow and two red. A third headlight is mounted in the roof, but not visible in the photo as I have moved it to gain access to the PCB.
The conventional approach to directional lighting is to separate the circuits for the forward and reverse lights. A function only decoder can then be used to supply power and two different connections for forward and reverse lights.
On this PCB, the tracks can be seen which link the LEDs. These would need to be cut to power the lights separately. The PCB in the nose of the power car is different, the lights are separated. This does allow normal DCC control, so the ideal option would be to obtain one of those boards. Unfortunately I haven't found a source for such parts.
So the problem is how to reproduce DC style directional lighting via DCC. Most decoders do not have bi-directional function outputs. But the motor output is bi-directional. So I decided to use a T1 motor decoder instead of a function only decoder, which can then be suitably programmed. There is plenty of space inside the model, so the larger size is not a problem.
The T1 decoder is attached to the weight in the centre of the chassis. I cut the wires which connected the chassis to the body, and connected the red and black decoder wires to the truck/bogey pickups. The orange and grey wires are then connected to the two wires leading to the lights. One of these is at the rear due to the option for pantograph power.
The decoder should be given the same address as the one used in the powered loco. As fitted so far, the lights would behave like those on a DC loco, only on when in motion. Instead the T1 decoder cane be programmed so that the motor outputs behave more like a function decoder, i.e. at a fixed voltage, but retaining directionality.
Setting CV61=68 enables button control of the motor circuit, and a power level for the lights should be set in CV133, I used 100 (the range is 1-255).