The NEMA sockets are dated to April 1998.

Polycarbonate refractor bowls offer some degree of light distribution.

Internally, the lanterns are in an altogether cleaner condition. They are equipped with an inline fuse carrier that houses a BS 1362 fuse rated at 3 A, which seems to be a means of protecting the high frequency electronic ballast. With both lanterns, the fuse carriers have been wired out, suggesting that both lanterns have required replacement ballasts in their respective lifetimes.

The internal labels provide some interesting information - when these lanterns were made, they were still known as MI 26s in the UK; the more global 'XGS 103' designation not being used until the early 2000s. The lanterns were made at the Philips factory in Hamilton, Scotland ('HN') in June 1998 (8F).

The lamp in the lower lantern dates from October 2007, having been fitted second-hand in 2017.

The one-piece electronic gear units are located on the inverse of the lamp reflector. Originally, the reflectors were attached to the insides of the canopies but now hang loose by the internal wiring, owing to the plastic supports long having snapped. Owing to the MI 26 having been designed with conventional gear components in mind (that attached to the inside of the canopy too), the electronic gear version is something of an adaptation. The canopy moulding was not updated to feature the screw pitch of the new ballast (not that it would fit particularly well, owing to the position of the NEMA socket) and so, whilst the ballast does fit on the reflector, a corner of its circuit board has to be cut away in the factory during assembly of the lantern, as otherwise, the corner fouls one of the fixing hole screw points for the reflector. As the replacement ballasts do not have the corner cut away, if all are intact, this suggests that the lantern has received a new ballast at some point. Owing to this design flaw, the ballast in the upper lantern is only supported at one end, whereas, on the lower lantern, one of the reflector securing screws has been removed.

The v-shaped clamp seen earlier grips the bracket pipe when these two nuts at the back of the lantern are tightened. A stamped-on date nearby suggests that the canopy was made on the 30th March 1998 (it is the 28th March for the other lantern)...here's where things became extremely spooky - this date twenty years later was when the lanterns were removed from service. The 30th March 1998 is also the date that a Thorn Gamma 6 replaced the GEC Z5698U on the column behind my old house - the installation that formed the genesis of my street lighting interest - everything happens for a reason!

The pictures below feature the column that supported the MI 26 fitted with the part-night cell, after the P852 had been fitted.

Along with the official Phosco label, a larger label is also present - this allows the lantern's exact credentials to be viewed from ground level. The various parts of the code translate as: C = CU Phosco, 13 = Wattage, P5 = Designed to meet a P5 lighting class, and D = dimmed output after 21:30 (75%) and again after midnight (50%). A Royce Thompson Oasis 2000 photocell set at 20 Lux (though, in reality, this seems to be set considerably higher) photocell is employed.

The canopies were given a thorough cleaning on Thursday, 2nd April 2020 (thanks Dad!).

I had intended to paint the canopies, following this cleaning; however, I decided to leave them unpainted, owing to how well the cleaning alone had improved their appearance.

The two lanterns were reassembled, and two Oasis 2000 photocells, both dating from August 1998, were placed in their NEMA sockets, though neither will be used - the NEMA sockets are now wired out.

Other internal modifications, not shown here, include reinstating the in-line fuse carriers, and, in the case of the upper lantern, cutting the corner away in the replacement electronic ballast, and installing a new second screw to secure the ballast to the gear tray.

Free-standing 2 m columns were fabricated by TAS Engineering, of Burton-upon-Trent, to a design that utilised the original outreach brackets, and secured them in the same way that they had been secured to the footpath columns. The two columns and brackets then went to a local refurbishment company for painting in RAL 6006 - the colour that lighting columns in Derbyshire (but excluding Derby itself these days) are painted. Appropriately, the two columns are positioned alongside two of the artworks created by Chris Cyprus for his 'Northern Lights' series, which was inspired by the distinctive colour of SOX lighting contrasting with pale blue evening skies. The lantern replaced by the P852 seen above is first.

The lamp seemed to take some time to warm up; I expect that it dating from 2013 and being well-used played a part in this. Eventually, the full colour was reached, however.

Testing with my energy monitoring device revealed the following results:

Test Voltage (V)	Current being drawn at full power (A)	Measured wattage (W)	Apparent Power (VA)	Frequency (Hz)	Power Factor	True Power (W)	Difference to rated wattage	Percentage Difference
240.2	0.18	39	43	49.9	0.89	38.48	3.48	9.94%

The second lantern was installed a few days after the first was.

The lantern, shortly after being powered up.

This lamp too seemed to take longer to warm up than normal...unless, the lack of any SOX lighting on my road these days has caused me to forget the amount of time that the lamps take to warm up!

Evidence that this was an old lamp was indicated by the electrode area retaining a slight red glow throughout the hours that I ran the lantern.

The test results differed slightly to those of the other lantern:

Testing with my energy monitoring device revealed the following results:

Test Voltage (V)	Current being drawn at full power (A)	Measured wattage (W)	Apparent Power (VA)	Frequency (Hz)	Power Factor	True Power (W)	Difference to rated wattage	Percentage Difference
243.4	0.15	35	37	49.9	0.96	35.05	0.05	0.14%

Lantern operation video: