Working on a unit can mean different things: keeping it as a documented example, returning it to the way its maker built it, or changing how it behaves. Those aims are not interchangeable, and a choice that suits one unit may be wrong for another.
Start with the evidence#
I consider the unit’s condition, what’s known about its design, and the quality of the available references. Clear photographs and notes help separate original features from later changes. If the evidence is incomplete, that uncertainty is part of the record.
Repair toward the original design#
Where I do repair, I favour like-for-like parts and the maker’s original arrangement. The parts that limit the output and keep it isolated were chosen for a reason, and a substitute that “works” may not protect in the same way. Repairs should be documented, including what was changed and why, so later readers can distinguish original parts from additions.
Modification is a separate question#
Raising the output, removing a limit, or adding a mains supply to a battery unit is not a repair. It turns the unit into a different, untested design, and the person connected to it carries the risk. I treat any such change as a design project in its own right, with its own assessment, and not as a quick improvement. Batteries, Chargers and Mains Power explains why the power source matters so much.
Function is a separate question too#
A unit that switches on and produces a tingle is not necessarily sound. Conversely, keeping a unit in a collection does not require returning it to use. Deciding whether one is fit to connect to a body requires direct assessment by someone with the training and test equipment to measure its output. This article is not a repair or testing guide.
Explain the decision#
For each project, I want to describe the options considered, the evidence behind the choice, and what remains unresolved. That makes the write-up more useful than a simple before-and-after photograph: it shows how the unit was approached and where caution shaped the result.
There is rarely one perfect answer. A transparent, well-documented decision is a good place to start.