"After looking at the the photos showing the new signals illuminated, the uneven nature of the lighting screams incandescent optics. This leads me to believe that CP hasn't invested thousands of R&D hours in some sort of new technology. After all, if they had, the result sucks. Why spend money for no improvement and why spend money on a technology with a limited lifespan?
This is GRS territory. Why not stick with what they have and what the maintainers already support? If you compare photos of the side of one of the new housings to those of any other GRS SA, you will see an identical triangular ratio between the door hinge pin, the mount hinge pin, and the mech mounting stud. Their wire-coupling assembly on the bottom of the housing, retained by three bolts, also looks like it would be a drop-in replacement for the GRS part. Almost certainly there is an SA mechanism in that box. Therefore the guts aren't news, the housing is the Main Event here.
Well there you have it. There is a good chance that inside these new housings are classic old GRS electro-mechanical searchlights. John said he'd ask around and try to get some confirmation on his theories and when/if I hear more I will update this post. Unfortunately CP is increasingly abandoning the whole searchlight motif on its former D&H and CP main lines, even the Unilens and LED type. Here's to them learning to change their errant ways.
What's so special about the housing? It is NEW. I have a little passel of captive dwarf signals and over time I've taken each of them apart for restoration. There is very little that can go wrong with an H or H-2 housing and the H-5 is the Cadillac of searchlight housings. However, the mid-to-late era SAs have a primary design weakness in the mech rail. The door latch pulls on a loop on the end of the mech rail. Over-tightening, common on heads with failing gaskets, results in the latch pull snapping off the end of the mech rail. Replacing the mech rail forces a confrontation with the bolts that retain the mech rail and the sighting device. I've never successfully removed those bolts without shearing or deforming the stamped bolt end inside the mech rail. Ironically it is easier to remove these bolts from cast housings, as every single aluminum housing has exhibited an aggressive galvanic reaction between those parts. I've scrapped a signal head where the aluminum was so pitted that water was seeping into the housing.
In the course of restoring one of my SAs it was necessary to re-tap the housing and make new bolts on the lathe. That's too much work. If I were running a signal shop and someone brought me a pile of SA heads and asked me to refurbish them to new, I would be very thankful for the near-term job security. I would be busy for a while.
Analysis: CP has shelves of serviceable SA mechanisms sitting around that represent a significant investment. Why not put them to work? The SA housing is flawed, so it is easier to send the critical dimensions to a fabrication house, where the contractor can CAD, stamp, mill, and weld a product far cheaper than an in-house recycling project, and for the equivalent cost of new housings from Alstom. Alstom doesn't want the business anyway; why else would they price the SA into extinction?
What won't break? The door latch. What isn't rotting? The door seal. What isn't leaking? Bolts on the top of the housing, certainly replaced by welds inside and out. What is back in productive use? SA mechs.
Time will tell if this is a successful experiment. The serial number would lead one to believe that they have made at least 623 of them. CP walked away from over 100 searchlights and searchlight mechanisms when they abandoned the Chalk River and portions of the North Bay Subdivisions in 2011. Perhaps that is also a sign that they have more than enough spares stockpiled to last until they are done with searchlights.
As an aside, if someone wanted to design a drop-in SA replacement with no moving parts, they could simply replace the mechanism with a disc carrying a multi-aspect LED diode array. The outer doublet lens would have to be discarded and replaced with plexiglass. If a replacement required focused optics, a solution could be engineered by a team of three in under a week. It would feature three compact LED clusters feeding into a prism. The optics would be mounted in a dummy frame that slid into the mech rail, aligned on the mounting pins, and lever-locked in a secure position. Nothing complicated there, but it is far cheaper to buy a stock colorlight head and bulldoze all the rusty old fossils. The real problem is the neglect suffered by the old equipment, which sometimes exacerbated the design flaws. Time will tell if these new searchlight housings fare any better, but I commend CP for trying. For now it seems to be an improvement on the SA housing. Too bad it is so ugly."