Some clients (especially in Japan) have developed specific rules they impose on the language used in patent translations. Some of these rules are useful, some are dubious, and some have hardened into industry 常識 (common practice).
For example, some clients suggest avoiding the words while and when because they can introduce temporal dependencies into the scope of the disclosure. Some translators even avoid using if for 場合 (they always use in the case instead). Function words like if and when can be problematic in specific cases, but it is ridiculous to try to exclude them from patent translations altogether.
On the other hand, an example of a sensible rule that all patent drafters and translators follow is to introduce elements using the indefinite article a and then refer back to them subsequently with the definite article the. For example, “A device comprising a sphere, wherein the sphere rotates counterclockwise.” However, some patent translators take this rule to the extreme and use the indefinite article in front of every new noun that appears in the text. So you end up with ridiculous constructions like “A device comprising a sphere, wherein a surface of the sphere is reflective,” when of course it should read “the surface of the sphere is reflective,” since a sphere has only one surface. Similarly, there is no reason to write “a weight of the sphere is 100 kg” when you mean “the weight of the sphere is 100 kg.”
Ridiculous practices like using “a weight” when you mean “the weight” accomplish nothing but making the patent more confusing, thereby irritating the examiner or litigator who is unlucky enough to read it. Patents are already difficult enough to read and we shouldn’t add to the confusion by blindly following dubious made-up “rules” in our translations.