That’s it. Walter’s got the bed supports on Thingiverse too. Ballscrews were about $60 shipped from Golmart on Aliexpress.
I’d suggest trying isolators before shelling out more money. I don’t really understand the aversion to isolators. The cheaper ball screws you’re looking at aren’t likely to be straighter than the lead screws you’re already using anyway.
Running super cheap ballscrews here, got them with the ends machined, I made my machine almost exclusively from Aluminium machined parts so its super-rigid and constrained and I do see very slight banding. Ballscrews are probably better but they’re not a definite fix…
Someone have a link to a “z isolator”? I’d like to investigate these!
@Ben Delarre. : will you have time to talk online this coming Monday. I noticed that you are running a smoothie on your V2. Let me know if Monday is good and about what time to communicate.
A z screw isolator basically disconnects the z nut from the thing being raised so it can only apply force in z. A flat or crowned surface touches the z axis parts, and something like an arm or fingers to keep the z screw from turning. I don’t have one for this community’s kind of machine (not built yet) but I can link to one I made for a different machine: http://www.thingiverse.com/thing:750610
@Jeff_DeMaagd my main concern with z isolators is that the eustathios homes at the top and any z isolator solution I can think of always has some backlash in it, which will be especially evident on the first layer after reversing direction from the z homing. How do you avoid backlash issues when using isolators?
Just make sure the platform can move up and down freely. Either I don’t get backlash or it’s repeatable enough that I can’t tell. The weight of the platform should be more than enough to hold it down.
@Ben_Delarre Just a thought since everything else seems to be mentioned above. Are you using helical lead screw couplers with motors mounted at the bottom? If so, is the lead screw bottomed out to the top of the motor shaft? Helical couplers will act like a spring if there is a gap between lead screw and motor shaft, allowing vertical movement on the Z axis.
@Jason_Perkes thanks for the suggestion, but the eustathios is driven with a belt from the motors to two pulleys one for each z leadscrew.
Same principle or question applies whether they be belt or motor driven. If there are no helical couplers then question is invalid. Possibly might trigger a thought though. Would be interested to know the fix in the end 
Yeah no helical couplers. Its belt from motor to z axis leadscrew with pulleys directly mounted onto the leadscrew. Just switched filaments to some petg and the banding is much reduced, still there but tiny amounts. Now I just have bobbing on retracts and layer changes… If it’s not one thing it’s something else!
Blobbing on retracts, reduce all retract related speeds and accels to real slow then dial back up slowly. E axis acceleration or retract acceleration is one most overlooked. Reduce to 100mms’2 and 10mms speed for a slow and steady reliable retract baseline then dial e accels back up. That’s how I work it anyways…hth.