Thanks to this forum I have found out about the need to reform capacitors.
For a lightweight variable AC supply is there any problem using a common dimmer switch through a voltage doubler?.
I am not sure of range etc or any unforeseen problems.
I intend to use the circuit below.

You only reform Electrolytic caps and that is with DC. There is trepidation before I even see a circuit.

Photo uploaded.

Anything running directly from the mains without a transformer, is to be considered intrinsically unsafe. Dimmers do not put out a good waveform. Even the reformers in my 1938 Valve & Circuit Tester use a transformer.

The unforseen problem is likely who organises your Funeral?

"Silicon Chip" Has featured two reformers. The one I use, built onto the Bench PSU, uses the "Rodney Champness one". It actually requires a transformer (Which my PSU has already) and uses an LR8 regulator that will handle 450V DC It utilises a switch to step the voltages via a divider chain. Secondly you can discharge the cap, via a 10K resistor, which is a lot less stress on the cap & switch.

It is not difficult to build.

These are a vastly safer option.

NB. "Variacs" aka "Slide regulators" are "Auto transformers" These are also unsuitable and just as dangerous as there is no isolation from the mains. At around 130VAC OP; a 30mA RCD is unlikely to trip (tried it with a tripping device): Be warned.

There is a point where one does not bother reforming decades old electrolytic caps. I mainly use it for new caps that have been stored a while, or to ping the duds. Some new are duds: Not common.... but!

With regard to reforming, I am one who just avoids it. All the bother with setting up to do the job and it ends up costing more than simply replacing worn condensers with new ones - to me, this, along with worrying about whether the reforming worked or not, just doesn't warrant the time spent on reforming.

Recapping a radio will cost what... $20-$30? It's playlunch money and if the goods are good, they won't require replacing again by the current generation of collectors and restorers.

Point taken with the safety aspect. I did read that dimmers will damage transformers hence the direct AC.
There are conflicting opinions regarding reforming - to reform, stuff the cans or fit modern caps under the chassis.
I guess, unless the set is museum quality, that disconnecting the cans and fitting caps underneath would be the easiest option.
In this case I will do that.

My "reforming" is pretty basic - turn on the radio for about a minute, turn it off for 10minutes, then turn it back on again.

The last few radios were done with a light globe in the mains lead, because the globe will change its output as the capacitor reforms.

But in the end if the cap fails, just replace it. No big deal.

I think that circuit is ingenious, and a good way to explore the characteristics of vintage electros that you do come across e.g. ones that have not seen service over a number of decades.
Safety is paramount, I am thinking the source of DC must be isolated from mains, could come from e.g. plug pack, start on say <50v...
Insulated terminals, nice to add a light bulb(s) to circuit.
I would be interested in what you find....

As noted I use mine more for testing than reforming. I do not bother attempting to reform the old red Ducons. The other thing I use it for is powering a "B" rail for odd ball faults, or if I think a filter cap has gone funny* (beware of voltage dividers).

One can (valves removed no other power) run the "B" rail up to the rated or surge volts in steps & measure the current, or disconnect one end of the filter cap & check it individually. I did that recently where there was a transformer melt down. I disconnected the "B" from the Metal Rectifier & powered the rail. After which the rectifier became 1N4007's.

NB The LR8 in that reformer being abused, will lock up on overload.

* New caps can fail. I have had a couple of instances where the 450V RB's I will not use near anything that behaves like a 5Y3 /#80 have not lasted and I have had some 500V types start to conduct at a hyperbolic rate, before getting to there rating. This is where the reformer is a useful tool and why I always monitor the "B" rail on startup.

I do think what I have suggested is not only a safer option but their voltage control on the DC side more practical.

You can if you have one, consider the "Variac" followed by a Valve radio transformer to give the isolation: Salvage?

A good test/diagnostic for HV electros is a megger, switched to either 250 or 500v depending on the ratings of the cap being tested.
It is essentially a pretty high resistance from a HV source with a voltmeter across the test cap. so you can watch a electro slowly charging up.
Any cap with serious leakage current would show up as not charging up towards the source volts, stuck low.
This would tell you a lot about the cap(s) being tested.

Reforming may take a lot longer, in a slow controlled way.

Many moons ago I made up a HV variable DC supply consisting of a conventional HV DC supply with an 809 power triode in series on the plus rail i.e. cathode follower. Its control grid was connected to the slider of a linear pot across the supply. With a voltmeter on the front panel. Useful for experiments, testing...but I digress...

My method is to use a variable voltage DC source. Connect the capacitor negative to negative and fit a 10k resistor between the positive of the supply and the cap. Connect a DC voltmeter across the capacitor and raise the power supply voltage to around 100V. Watch the voltage across the cap. It will rise slowly, sometimes very slowly. When the voltage across the capacitor is within 10% of the supply voltage, raise the supply voltage again by another 100V. If a cap is beyond reforming, the voltage across it will not rise. As well, when the supply is disconnected, the voltage of a good capacitor will drop slowly whereas a poor one will drop much more quickly.
It must be said that the variable DC supply has an isolating transformer within it.
I can post a circuit of the power supply if required.

N believe the formula for leakage is 0.1CV. draw un in uA. the old Lafayette RC bridge here says the limit is 10mA: No way would one with a leakage like that get a berth in a radio.

The stepper is a great idea even for HV caps. Especially if you have ones that one has a better than even idea, will be a dud. You can start at 25V & let it sit there for a while, if it fails to draw excessive current.

Ok, my immediate problem is the lack of a variable DC supply. I will rectify that first.

I've had good results reforming large, expensive multi-section electros in 50's and 60's TVs (provided they are showing no signs of physical leaks and they were not made by UCC) by using the chassis's rectifier, all other valves removed, and a dim bulb.