There is a segger in the middle of the tube. From the side of the motor there sits additional washer on it. I will definitely doublecheck, but I think that this side of the assembly is pressed tight from the end cap. The end cap is locked with the rim of the tube being bent over.
I received the pump with the rotor completely stuck. During the loosening of the nut it got a bit loose and rotating it back and forth helped it spin again - but very, very noisy. After the bearings disassembly, it was full of rust and maybe another mess too. It seems that the bearings weren't actually that bad, maybe the preload function failure and aggressive and wet gases just made it bad. Unfortunately the cages were too fragile and were broken during the disassembly. Eight balls each. My theory is that the pump was working worse and worse and maybe got stuck not during the operation, but during the storage - because the turbine is in perfect order. However, plenty of rust everywhere in the bottom part.
Thank you for the link to particular bearing. My greatest concerns are about the oil vapor pressure. The bearings live in the output vacuum part, but it still requires to have some oil that doesn't get vaporized at operation temperature and is still able to act as a lubricant. For my foreline pump (rotary vane) I use double vacuum destilated oil, but I have no clue about the required viscosity for the TMP bearings. Most likely it's not the right one as it seems to be too dense.
I already asked Alpine to find the bearing for me and they pointed me to
https://alpinebearing.com/ball-bearings/yca1650c-ghy72/. That's very close to those $200 for a pair. I also checked their lubricant Asonic GHY72, but even the smallest package is crazy expensive. Way more than $200!
At this point I'll try to check multiple bearing options. Let's see what I will find. Anyway all the alternatives are a bit risky - almost anything will work in the beginning, until it get stuck during the operation. Hopefully the Turbotronik will detect high consumption. And hopefully I will be able to detect topo high temperature of the TMP on time.