e-light

Thank you for the warm welcome in the forum
The 20 seconds are my experiences for images that are not tracked!
With “our neighbors” it even worked ad hoc from the car roof with 15 seconds and a 10 second advance trigger! It was 80 KM west of Nantes France early late summer morning on vacation without a tripod or Vixen GP mount!.
I use the camera to slow down, to balance out work; that's why I make slides like I did 40 years ; they're called JPgs ooc today! It shouldn't be work. . . . I use Olympus Master 2.3 (out of the first digital and af Cam system) for crop reduction for the web and here and there B/W conversion. The Sony (a7 + Canon FD Lenses) data will not be changed; Fuji allows itself to be occupied by Olympus; but the camera is called correct. If you mainly use TTartisan manual lenses, of course no lens data is given. I use the X-A1/2/3 because of its Bayer sensor and the compactness with TTartisan lenses.  I must say I am a 23mm / 35mm  looking guy how sametimes use the 35mm/50mm as a "Normal-Optic" ; -)
 

Welcome to the forum.
There are plenty of star trail exposure time calculators available for the various guidelines (“rules”), such as
 https://www.photopills.com/calculators/spotstars
Put the body on a tracking mount and periodic error is the more important obstacle to keeping stars round. On a tracking mount, 3 to 5 minute exposures are typical for each frame for cameras that have the ability to store longer exposures without over saturating the pixels.
For some camera bodies, a 20 second limit is a good idea, but that is more due to the full well depth size for that camera body rather than to prevent or allow trailing. 20 seconds for some bodies is the limit where brighter objects begin losing color by over saturating, turning the pixels white, especially when shooting in jpg instead of raw.