The ability to bring to front or send to back is essential whatever CAD programme you use. You should not fill unconnected areas together in any case (I learned this many years ago - by experience!) and this is exacerbated by the use of .eps, possibly why PPD pointed this out to you. You shouldn't be "filling" tags anyway, all you need is the fill - simply do one tag and copy it wherever you want one. The tags can overlap on to the parts, this is where bring to front and send to back become important. Your final etch drawing should be just the fills, delete or turn off all the lines used in constructing the drawing.

This is a small area of one of my test etches, all these are fills, grey for etch both sides, blue from back, red from front. I use three yellowish colours for layers which cancel out  the others - yellow cancels the front etch so leaves a raised rivet or number for example, khaki cancels the back layer so leaves a connecting tag. There is another layer (not seen here) which cancels both, this is mostly used for corrections applied after 1st build.

All these layers are changed to black and white for printing of course, the colours are just for my convenience to see them on the screen. The two parts look like this when sent to print, front film first.

I also draw all parts separately with their own surround, this makes it easy to move them around the etch sheet or copy to others later. I do a demo explaining this at York and Warley exhibitions and would be happy to help if you catch up with me there, we (Judith Edge kits) also have the computer with us at all the shows we do.

The best thing i ever bought was a prototyping breadboard, that I use to prove my circuits. Using smds on it means attaching wires to them first, although it is often simpler to build a typical loco light board and plug that in. Normal leds simply plug in.

My ‘universal’ circuit for diesel loco lights is 3 x white and 2 x red at each end. All the anodes (white and red) are commoned at each end with a single 1K ohm resistor feeding back to the decoder socket blue pin. All the white cathodes are commoned at each end with a single 1K ohm resistor feeding back to the decoder socket white or yellow pin (see later). All the red cathodes are commoned at each end with a single 1K resistor feeding back to the decoder socket yellow or white pin (see later). That makes 6 x resistors on a small board (see later).

Obviously to get the whites at one to light with the reds at the other end requires the front whites and the back reds to connect to the socket white pin and the back whites and front reds to connect to the socket yellow pin.

The mounting methodology is to put the resistors on a small veroboard with sets of red, white and blue wires running to the loco front end, back end and decoder socket. At each end the leds are mounted to a bit of veroboard using link wires to connect the leds as above, then the resistor board wires are connected to these boards. Finally after proving the installation with a battery, the socket wires are connected.

This is basic directional lighting with none of the extra switching required for day/night or tails off running that many folk may consider essential.

It works for me.

Rob

I've just remembered that I made a drawing for the hopper parts that I made (two of each shape needed). It might be of use but I don't guarantee its accuracy compared to the prototype.

Hopper.pdf

I have some photos of the stay-alive capacitor install I have done on my Dapol Terriers (2mm Scale Association etched replacement chassis). I too followed Nigel's blog, and purchased the components from the german website he mentioned. I used the ceramic capacitors, but after reports that their capacitance gets lower with voltage, and that tantalum capacitors are better, I tried making a similar setup with them instead. Unfortunately, they ended up just slightly larger overall, and I just could not clip the body firmly to the footplate, so the ceramics are still in use.

Starting with the components:

• reel of Kapton tape (very thin, transparent, heat resistant and insulating sticky tape)
• reel of fine enameled copper wire (as used for winding transformers or motors)
• capacitors (look like liquorice allsorts)
• diodes (one schottky, one zener, look like tin cans)
• resistor (labelled 1000 for 1kohm)

The capacitors were soldered together in pairs (end to end) then combined to make the block of 8. A small piece of Kapton tape was put on one end to insulate the central part from a length of thin copper wire (enamel removed by scraping) that was soldered along one side of the block, run across the tape and then soldered up the other side. The free end will eventually connect to the negative pad on the decoder.

The block was than wrapped in Kapton tape, except along the top. The Zener diode was soldered across one capacitor, and a short length of the wire attaches the centre of the block to the Schottky diode, with the resistor in parallel with it. A wire run from the other end of the diode/resistor combination will run to the positive pad on the decoder.

The diode/resistor combination was then moved to one side of the block, and more Kapton tape wrapped everything up tidy.

The two free leads from the block were then carefully soldered to the pads on the decoder. The decoder was already protected by Kapton tape, so two small 'windows' were cut into it with a new scalpel blade first. Polarity was checked and double-checked and triple checked before applying power.

The enameled copper wire leads could then be coiled up tight so the block could stand right behind the decoder in the cab.

A tight fit, but it works.