HSK or BT diy spindle?

HSK is probably technically superior, but tooling is likely to be more expensive, and it is more difficult to make your own shanks. The latter point might not be too significant. It is possibly to buy a boring head adaptor in BT30 for about £15 and I have modified these several times to make specials. If the same exists for HSK then you don't need to worry about making shanks.

I made my own power drawbar for a BT30 spindle rather a long time ago.
www.cnczone.com/forums/vertical-mill-lat...-log/109301-cnc.html
image links etc have bitrotted in the meantime. You need "a few kN" of pull-stud force for BT30. I think that I have 2/1kN and it has been working very well, even with some very large cutters.
There is a video , but I am not sure that it adds much to this conversation.

In an old email I found a very simple analysis "Looking at the springmasters table, the D125621 is 700N @ .08mm and 1.5kN @ .18. Put simply you get 700N per parallel set and 0.1mm per serial set."

I think that I have 150 springs in threes, for 2.1kN and 6-7mm travel. But it was a a long time ago, so I could be rather out there.

I was lucky that the drawbar hole on my mill was unusually large, big enough to fit a tube full of suitable springs inside. If you are designing from scratch then you can make provision for whichever springs you settle on.

As far as I know _all_ ATCs use disc springs. I don't see any reason to buck that trend.

One thing to bear in mind is that you might need to eject your tool. My machine has _just_ enough cylinder force to compress the springs and release the tool, but if the taper sticks, there isn't enough left over to break the tool free.
I am seriously considering either building a tandem cylinder to replace the existing one (fun project , finding a cheap pneumatic intensifier (pressure doubler) or converting to oil-over-air hydraulic.
One limitation that I have is that I am unwilling to use anything but a "silent" compressor so I only have 100psi to play with. In any case, design your cylinder for rather more than the required spring force.

The 3D model of my drawbar is here: a360.co/3vUInVK

It seems that I used an 80mm air cylinder, which should give me 3.5kN of force @ 100psi and that seems slightly marginal when the spindle gets hot and the tools stick.

Also it seems that I went for sets of three springs, so that is 2.1kN design force. But note that is the "relaxed" load when clamping, and rather more is needed to reach the "released" position.

FWIW my power drawbar has been absolutely no trouble, it is far and away the most reliable thing I have ever made

Looking at upload.wikimedia.org/wikipedia/commons/4...IN_69893_hsk_63a.svg
I would say that a HSK clamp looks easier than a claw-style BT gripper. (And has a more definite eject)

If I was starting from scratch I think it would be with HSK. But I have about 40 BT30 holders and a BT30 spindle, so I am not about to change.

chris@cnc wrote:

how did you animate this svg? very good idea to explain

I didn't, that's a direct link from Wikipedia.

However, I have done other wacky things with SVG, this heating controller uses active SVG as the user interface in the web browser.
github.com/andypugh/ESP-Heat (the .svg is generated in C, rather than with an editor)

And my Lathe Macros screen uses a multi-layer SVG to display different images in different tabs with only one input file.

It's a very flexible format.

This thread prompted me to redesign my outer cylinder and add a secondary piston, to double the release force. This will allow me to reduce the compressor working pressure.

a360.co/3tCPsZn

The idea can be carried on for several layers.

Not yet designed is the drilling to take air from the top piston volume through the stem to the lower piston volume. And I need to decide how to vent air from below the top piston.