To those who have been following the DLP thread(s):
I have spent the last few weeks experimenting with producing double sided pcbs with my E260 laser printer. What I found is that the hack I have documented is not adequate for DS boards. The repeatability is almost perfect in the X axis, but the Y axis (defined as the axis through the printer) varies a bit, sometimes quite a bit, from run to run.
The problem is that the paper path is not quite straight, which results in some friction and hesitation as the carrier moves, especially as it traverses the space between the input rollers and the transfer roller. Once the carrier enters the printing area it seems to move at a constant rate.
The biggest culprit is the Manual Feed Paper Guide which forces the carrier to actually bend a bit. In fact, if you tape the rear edge of the pcb to the carrier, the top surface of the pcb will touch the 4mm shaft, causing (in the worst case) the timing to be so bad that either the carrier stops moving or the printer errors out.
I have redesigned and redocumented the hack and will be uploading the latest version in the next few days. This only works with the MCU version of the Modification! The changes are:
1 - The input rollers now use a 5 mm shaft and 13 mm rollers instead of the 4 and 10 currently used. I upped the shaft diameter because the 4 mm shaft fits too loosely in the notches. You can still use the 4mm shaft if you already bought it by wrapping 5 or so turns of electrical tape around the ends to bulk them up to about 5mm. The added 3 mm to the roller diameter allows pcbs to easily fit under the shaft as the carrier advances. Both the shaft collars and aluminum spacers (rollers) are now 5mm ID. That said, if you already have the 10mm rollers they will still work, although thick pcbs might be a bit dicey.
2 - The Manual Feed paper Guide is not used and can be discarded.
3 - I have also deleted the Center Bracket. It is not necessary since the corrected paper path now feeds the carrier in a very straight line and does not touch the Center Bracket.
4 - The reflective sensor is not used as the NPIS. Instead we are using the U shaped transmissive sensor that was originally called the Narrow Media Sensor. It is mounted on a small piece of aluminum angle that also acts as a centering guide. This new bracket is mounted on the swing-down Front Cover.
There are a few other changes that make the conversion a bit easier. More details on both the improved hack and double sided boards to come.
Mark