I just checked Triad Magnetics to see if they could make a custom high-leakage-autotransformer (HX ballast), and that wasn't one of there custom options on the website. However, they do list ferroresonant transformers as part of their manufacturing options, which essentially requires the same materials and setup. I am curious:

Has anyone had a ballast custom-made? If so how?

Only home made 50W MV series choke, using materials from cheepeese PL11W lanterns (230V world)...

I just checked Triad Magnetics to see if they could make a custom high-leakage-autotransformer (HX ballast), and that wasn't one of there custom options on the website. However, they do list ferroresonant transformers as part of their manufacturing options, which essentially requires the same materials and setup. I am curious:

Has anyone had a ballast custom-made? If so how?

I have ordered some custom transformers, which is generally the same game.

BUT beware you end up with significantly higher prices for custom made stuff. SIGNIFICANTLY more than you may have been accustomized playing with NOS and second hand magnetics.

You specify either electrical parameters required and supplier does the calculation for you, or you specify exact manufacturing parameters - iron, gaps, wire gauge, #turns, taps. Or sometimes you just note a custom version number if someone ordered such a magnetic before.

Lamp ballasts are complicated in the way that you can run into a few manufacturing iterations until you will really get the lamp power prescribed. Or you can order multiple taps to fine tune, but the price again goes up. Or can have core gaps adjustable.

In practice this is not much needed until you go into a real serial manufacturing. You can get all the multitude of impedances necessary by combining existing chokes in parallel/series combinations and most of the time even HX transformers can be run in parallel. Or specifically for America you can play with CWA capacitor a bit, of course.

I’ve seen this mentioned before on LG, quite a few times and heard of people wanting to make their own ballast’s!
But this type of electrical engineering here seems to be steeped in secrecy and mystery! 🤔, people in Britain that can do it tend to not tell you much about it, i.e, the grade of wire, the wind numbers, the math that goes into it!

Imagine if you had the formula to hand to be able to make your own 200 watt SLI/H autoleak transformer, (something that is so rare in the UK, people have to use alternative gear)!
Maybe Roman (RRK) can shed some more light on this as he did above?😎

I think for anything but series chokes or maximum HX you need at least a field simulation, unless you want to go the route of really many iterations.
With series choke and maybe HX you get away with just adjusting the gap on the final product.

For number of turns you get away with basic "transformer" calculations (N = V * sqrt(2) / (2*Pi*Freq*MinimumCoreCrossSectionArea*Bmax)), this uses to be the starting point. But be aware to include all voltage combinations (open load, warmup, normal burn,...).
The wire cross section is "what fits" (assume the wire occupies an area equal to a square of the same size as the total wire diameter, include the insulation layer), using some margin factor, similar as with a regular transformer.
Then calculate the total wire length, its resistance (don't forget the temperature coefficient and the elevated temperature effect) and then the losses in it (take into account the real currents flowing through each winding; losses distribution across windings should follow their volume/copper mass, otherwise you get too high overall losses for the given overall size).
When losses too high or too low (aka core too big), you need to rearrange the core and start over, it is quite an iterative process.

So it’s not so much shrouded in secrecy, just incredibly complicated 😁

For using standard "custom transformer" services: Unless explicitely advertizing otherwise, these use to have "cookbooks" for, so are able to design from electrical specs standard mains power transformers only. Specialty things like ballasts or even e.g. audio output transformers are usually "above their paygrade", so you must specify exactly your own design.
They may serve you with the offering of some standard sizes for some parts (core foils, bobbins, wire sizes) and some coefficients regarding manufacture (e.g calculation whether the proposed number of turns of the specified wires fit onto the bobbin,...), so things they normally do for the mains transformers as well, but that is it. The rest is on you, the customer expertise, to design the thing (a choke,...) and provide exact recipe for the winding and core arrangements.
This also include the design details defining the required dielectric strength (and to deal with the fact the wire insulation alone won't be able to handle it by far), when the winding is to be used as the ignitor step up pulse transformer as common with the semiparallel ignitors...

Alright, so with the information I am gathering here, if I don’t know how to do it myself, I am out of luck with trying to get someone else to do it. I definitely knew that the price would be very high, I wasn’t planning on getting anything any time soon, but just wondering how someone would go about this process and if anyone did before.

Oh well.

I’ve seen this mentioned before on LG, quite a few times and heard of people wanting to make their own ballast’s!
But this type of electrical engineering here seems to be steeped in secrecy and mystery! 🤔, people in Britain that can do it tend to not tell you much about it, i.e, the grade of wire, the wind numbers, the math that goes into it!

Imagine if you had the formula to hand to be able to make your own 200 watt SLI/H autoleak transformer, (something that is so rare in the UK, people have to use alternative gear)!
Maybe Roman (RRK) can shed some more light on this as he did above?😎

There is no secret or mystery in calculation of a transformer, including a leak transformer. All the math is already 150 years old A tricky part may be there is no direct recipe to make a transformer or an inductor, you need get a balance of cost, weight, size  and efficiency. All these parameters are mutually dependent!  Add also a necessary 'fun' measure to a hobby project. Designing a balanced ballast may be an ok job professional, but too mundane to complete as a hobby.

First thing you have to respect in designing a magnetic part is saturation induction of a core, called Bmax. As Bmax defines a maximum of counter-EMF in a winding that works against an applied voltage, generally it will dictate a number of turns per volt for a given core cross section at a given line frequency. Okay. Next you have a turns ratio between primary and secondary approximately as a ratio of line voltage and OCV required. Note that in practical leak  transformers secondary winding is usually fed from a tap of primary, so the ballast is in fact somewhat in between of a choke fed by an autotransformer and true leak transformer. Next you add a magnetic shunt. Calculating magnetic resistance is complex, one may even tune magnetic shunt gap experimentally.

And so on.

As for (re)making a SLI/H ballast. As usual, before you start to make anything, the first question to ask is 'why'. Depending on the answer you can go different ways. Let's  say the answer is 'I just want to light up a rare lamp correctly'. Well then you have to remember that any leak transformer is electrically equivalent to a combo of an (auto)transformer and a choke. Making this is significantly easier than making a custom leak transformer. You can reuse some scrap transformer for an OCV required and then add a group of chokes calculated to proper impedance for prescribed lamp current.

Or the answer may be 'I want to have a historically correct replica of the ballast!" Okay then, the best you can do is to find a surviving original and take all the measurements. Turns can be measured non-destructively. Modern magnetic materials are likely slightly better than historical ones, so you may measure the core dimensions and re-create it with modern laminations.

Or you can answer 'I just want some fun)' Well that's legitimate for sure. Making a transformer/choke is relatively trivial. I did some at 50Hz and HF too. A winding machine is not complex and not very expensive, generally it is a bobbin holder with a counter and a hand or motor crank. Enamelled wire is easy to buy, even some modern varieties available, even off Aliexpress. You can cannibalize some scrap transformers for magnetic materials and coil formers.

Non destructive turns counting? Tell me more!

Usually there is some space to add some test turns of a wire over a transformer or choke bobbin. Let's say one or ten or whatever. Now you apply a known small AC test voltage on a winding you like to count and measure what you get at a test winding with a good voltmeter. Voltage rate will give you turns rate. If you expect a significant current in a measured winding, you can account for it, too by measuring the current and resistance of the winding that you are measuring.

For laminated iron cored transformers test winding can be placed somewhere nearby the main bobbin on a core kern too, you likely will still get tight enough magnetic coupling between the windings. Some care should be taken if there are air gaps in the core.

Thse same experiment goes for a winding with some taps if you want to know turns ratios, of course.
Sure with hard epoxy potted magnetics you are usually out of luck, until you are motivated enough to attempt to decap.

Well darn, on both of my ballasts, there isn't even a little crack to pass through some magnet wire. Unfortunate.