....
You could, but in order to meet the requirements of the standards, it is very likely that there is a thick layer of conformal coating on the PCB mount components that would need to be removed first.
Perhaps one hope is that separate boards were used by LPA Excil at the time of design, but from their later work they tend to like to include everting on a single PCB.
I suspected the construction would be like this both for electrical protection / isolation and physical immunity to vibration. My thought was based around separate boards / modules.
The highest output domestically available plug in transformer seems to be 50V DC. Getting 110v isn't in itself a problem, a UK to US power adaptor would do that. However it would still be AC. I need to be able to rectify that.
I do retain the second option of gutting it and replacing the innards with a 12V system.
Simply rectifying 110V AC will give an average rms 110V DC but derived from UK mains with 100Hz ripple and a peak of 155V. Further adding a smoothing capacitor to reduce the ripple to around 10V will result in around 150V DC not 110V. I suspect the specification for the train 110V DC has a tolerance of at least 10% and maybe more but not 35%.
110V AC is considered relatively safe and hence the industrial hand held tool voltage. For DC 50V is considered the maximum safe voltage. Hence why a lot of telecomms uses 48V.
If you could source two small isolated 48V DC supplies wired in series that would be 96V. Most though will have a far higher current output capability than you need and unless sourced as surplus expensive.
Just like the voltage requirement needs to be confirmed to be 110V DC so too does the current draw. I would expect the current draw to be low given that a typical 40W equivalent LED household bulb will around 6W. A test would be twelve 9V alkaline batteries ( e.g. Kodak from Poundland ) in series and a current meter / multimeter in current mode in line to measure. I would be surprised if more than 100mA as that is 10W. Taking one then two batteries out would find if the lamp still worked fine if a little dimmer from 99V then 90V.
..... getting 110v dc will be difficult and expensive.
That depends on your definition of difficult and expensive.
My approach if a little under 110V would suffice is two small transformers each with two 18V secondary outputs. The smallest other than PCB mount will be around 12VA and the regulation of small transformers is generally around 20%. With the secondaries in series the total will be around 80V DC at 50% load. Rectify this with a small bridge rectifier and follow with a 100uF capacitor so the ripple is around 10V / 10%. With a 100mA load the nominal output would be around 100V DC.
With a higher input e.g. one 18V transformer and one 24V transfomer there would be enough overhead to regulate to 110V DC without too much heat dissipation but to me over the top to replicate a rail supply for a lamp.
Add fuses and fuse holders on both input and output. Parts cost I guess around £30 plus a suitable enclosure. If this was hidden just a simple diecast box at perhaps £10, if exposed something suitable and decent looking would cost rather more.