Looks like a common problem. I was affected too. Bought mine (also version I)  in 2009 as second hand (still payed 4600 Euro then).  Later had it serviced by Nikon and got a SWM motor replacement installed. Finally a few years ago after seeing very little use and thus resting for some month I detected that it could nearly not be focussed any more. After sending it to NikonGermany again I got the reply that it needed a new inner focusing helicoid but that is not availabe any more. Finally I got it back having lost the little remaining ability to be manually focussed.

The lens was a revolutionary internal supertele-zoom with constant maximum aperture when it came out (and little to none none of the manifold "non-Pro" supertele-Zooms were available then) had great close-up capability and performed well in low to medium ranges but was found out to be subpar at longer distances. I dont need it any more but I share the disappointment that an expensive lens like that cannot be brought into life any more. It is the only lens of many I own that became unusable.

The person I spoke to offered to repair the lens for me.  I will have to pack it up and send it to China.  He said he's fixed a dozen of these and counting.  The second generation has this issue as well.  It seems really common, I saw some on eBay for dirt cheap prices, noting this issue.  I know someone coming over so they will be the mule to bring the lens back in March.  Once I get it back, I will make a post.
I wonder if there are any image quality differences between the first and second versions.  It might be worth getting the VRII with this exact issue and having it fixed.  I examined the lens' Flickr page.  Some copies do show obvious softness at long distances while some appear to be perfect.
 

Market day. January sun.

D850 Angenieux 35-70/2.5

Does anyone know the optical formula for this lens??

I took this photo of Joseph Weizenbaum a while ago, when I met him in Berlin. Who is? https://www.weizenbaum-institut.de/en/institute/our-guiding-principles/about-joseph-weizenbaum/

Tech: FM-2n with 1.4/50mm Ai and Fuji Velvia 100, scanned with the Nikon D850; 2.8/60G; current Nikon Film scanner scew in contraption.

Why do I post it?

Because we face a very strong form of ELIZA-effect (humans like to converse with machines more than with humans live or through machines): https://www.technologyreview.com/2025/04/08/1114369/ai-companions-are-the-final-stage-of-digital-addiction-and-lawmakers-are-taking-aim

I like that photo of the German-Jewish genius, his enigmatic and observant gaze.

I believe the language of the future is being formed; humanity still has to evolve. Sometimes I think we're almost primates.

My suggestion would be to put an extension ring (the longest you have) between the lens and the camera, then see if you can focus at short distances using the zoom ring.

The basis of the test is the lens conjugate equation: 1/f = 1/u + 1/v, where f is focal length, and v is the distance behind the lens where an object at distance u is sharply imaged. If u is really big (infinity, as you might say, blushing slightly) 1/u is really small or zero, and 1/f = 1/u, meaning an object at "infinity" is sharply focused on a sensor the focal length behind the lens. To focus closer - to make u smaller - you need to make v bigger - move the lens away from the sensor. In the large format world you use the bellows to move the lens, but in the 35mm world that is done by moving lens elements around. But there is an alternative way of focussing closer: change f. If f is smaller, 1/f is bigger, then for the same 1/v, 1/u is bigger and u is smaller.

If you can focus using the zoom ring, the lens elements are where they should be and moving as they should.


The reason to use an extension ring is that the smaller u is, the bigger 1/u is, so the bigger the change in f to make 1/f = 1/u + 1/v, so it is easier if you di it at close focus.