[dkp_design]

I have some A2Z end mill holders, and live the tool offsets. Greatly speeded up my work with multiple changes. But I am a little leery of the runout of a setscrew style 3/8" and smaller holders? Does anyone make a ER16 collet style holder that one would clamp tool in with the collet, then it screws onto spindle like the A2Z ones? Or is my worry of runout unfounded? I am not doing real precision work, but there seems to me more chatter and less tool life with the A2Z 3/8" running a .250 4 flt end mill.

[MetalZilla]

Unfortunately, I know of no such ER holders that will work for you.

[GBowes]

I have found the runout too great with the A2Z adapters.
Have you thought about something like the following?

ER16 10mm Straight Shank L100MM Chuck CNC Milling Lathe Tool Workholding L27 | eBay

They claim 0.0006" TIR, but we must remember the source. I have been thinking about getting an ER11 version and building a light duty, high speed spindle for my Taig.
If you have a Taig Mill, you should be able to fit this all the way into an ER16 3/8" collet with the excess shaft inside the spindle. By adding a stop that the shaft touches inside or above the spindle, you could eliminate the variation that is likely when tightening your existing ER16.

Graham

[dkp_design]

Hmm that might work Graham. Or thread the end, use a drawbar like design with a ER16 collet on the shaft down at the neck. Would always pull the shaft up, against the er16 collet, bottoming it out at effectively same depth

[awetmore]

Two ideas:

1) You can get exactly what you are looking for on the old non-ER16 spindle. Taig sells ER16 holders for that spindle. I used that method for about a year and it worked great and was nicer than using the A2Z holders because you get the benefits of wide clamping ranges from ER16. It was nice getting repeatable Z. The downside is that they stick out a lot more (so you lose effective Z) and runout is probably a bit higher.

2) You can locktite or glue small sleeves on the shaft of your cutting tools to get repeatable insertion into the collet for each one. They don't need to be anything too complicated.

[dkp_design]

Well if runout is higher, than thats a step backwards. I'm trying to get better run out with a resettable holder.

[awetmore]

Anything that extends the tool farther from the headstock bearings is going to increase runout. If you want to keep identical runout the only options are to either put a sleeve on the tool itself or make a repeatable stop that works like a drawbar.

[rlockwood]

I have some A2Z end mill holders, and live the tool offsets. Greatly speeded up my work with multiple changes. But I am a little leery of the runout of a setscrew style 3/8" and smaller holders? Does anyone make a ER16 collet style holder that one would clamp tool in with the collet, then it screws onto spindle like the A2Z ones? Or is my worry of runout unfounded? I am not doing real precision work, but there seems to me more chatter and less tool life with the A2Z 3/8" running a .250 4 flt end mill.

I'd designed this system here awhile back, and figure it would work well. I too have a taig, and would much rather have a dozen er-16 tool holders than the a2z holders.

I'd bet someone with a cnc lathe could turn these out for pretty cheap as well. I dont have one, and am doubtful there's too much of a market for them.

At any rate, its just an er-16 taper inside of an er-16 taper, with one end ID threaded for a "draw-bar" and the other OD threaded for an m22x1.5 collet nut. The whole thing can be machined in one operation, so the tapers should be quite concentric.

It might be advantageous to place some keyed slots on the outside, to make loading tools outside of the spindle easier. There's also the inherent cost of er16 collet nuts, which are fairly expensive by themselves.

[Jeff-Birt]

The problem with trying make/use a tool holder like the OP is after is that you wind up with a lot of run out. Not only do you multiply the run-out the further you move from the spindle but you also add more run-out with the second ER type collet.

The solid tool holder like A2Z makes can be used to allow a somewhat quick change with preset tool lengths. They had some problems with run-out last year but I think they have that under control now and have added a much better quality control system. With any solid tool holder like this if your bit shanks are not accurate, i.e. you buy the cheap import bits from Ebay, you may have fitment problems. If the shank is too small it will cause the bit to be offset to one side (more than normal).

For many operations you can get by with using depth setting rings. When I machine my end panels I touch off the engraving bit on each panel (first bit used), that way the engraving depth is very precise. All of the other machining steps are through the part so being off a few thousandths of an inch is not a problem. The depth setting rings on the 1/8" shank tools I use is with a few though from bit to bit so I can just swap from the engraving bit to the milling bit without resetting the Z axis.

Using a tool setting script also makes things easy. When I change tools I put the new bit in, run the bit down close to the part, place a wire with an alligator clip on the bit, and then set a small piece of single sided PCB stock under the bit. Then all I need to do is press a button in Mach and it runs the bit down until it hits the PCB stock and resets the Z axis automatically. This makes bit changes really quick.

The other tactic to keep in mind is to minimize bit changes as much as possible. On one end panel I machine I was using a 1/16" bit because I need the small radius it provided for one feature. Given the thickness of the panel I had to make two passes which was very time consuming. Adding a tool change to machine all but the one special feature with a larger tool was a possibility but that would take time as well. What I wound up doing was switching over to a 2mm (0.0787") bit, which was stout enough to cut through the panel in one pass and then 'over cut' the corners on the special feature. 'Over cutting' is a feature that was added to CamBam a while back. In corners it moves the tool out at a 45 degree angle ever so slightly (radius of bit). This relieves the corner giving the effect of a smaller radius bit. Visually comparing a panel cut with the 1/16" bit and the 2mm bit it is hard to see the over cut in the corners. Just this simple change saved a tool change or need for a two pass operation and saved 6 minutes per part! I have another product I think I can use the over cut on to save a tool change as well.

The main points are to not over complicate tool changing. Often you can eliminate the need for a tool change by rethinking how your part is deigned or how it will be machined. You can also use a tool changing script combined with an input to make tool changes faster and more accurate.

[rlockwood]

The problem with trying make/use a tool holder like the OP is after is that you wind up with a lot of run out. Not only do you multiply the run-out the further you move from the spindle but you also add more run-out with the second ER type collet.

The solid tool holder like A2Z makes can be used to allow a somewhat quick change with preset tool lengths. They had some problems with run-out last year but I think they have that under control now and have added a much better quality control system. With any solid tool holder like this if your bit shanks are not accurate, i.e. you buy the cheap import bits from Ebay, you may have fitment problems. If the shank is too small it will cause the bit to be offset to one side (more than normal).

For many operations you can get by with using depth setting rings. When I machine my end panels I touch off the engraving bit on each panel (first bit used), that way the engraving depth is very precise. All of the other machining steps are through the part so being off a few thousandths of an inch is not a problem. The depth setting rings on the 1/8" shank tools I use is with a few though from bit to bit so I can just swap from the engraving bit to the milling bit without resetting the Z axis.

Using a tool setting script also makes things easy. When I change tools I put the new bit in, run the bit down close to the part, place a wire with an alligator clip on the bit, and then set a small piece of single sided PCB stock under the bit. Then all I need to do is press a button in Mach and it runs the bit down until it hits the PCB stock and resets the Z axis automatically. This makes bit changes really quick.

The other tactic to keep in mind is to minimize bit changes as much as possible. On one end panel I machine I was using a 1/16" bit because I need the small radius it provided for one feature. Given the thickness of the panel I had to make two passes which was very time consuming. Adding a tool change to machine all but the one special feature with a larger tool was a possibility but that would take time as well. What I wound up doing was switching over to a 2mm (0.0787") bit, which was stout enough to cut through the panel in one pass and then 'over cut' the corners on the special feature. 'Over cutting' is a feature that was added to CamBam a while back. In corners it moves the tool out at a 45 degree angle ever so slightly (radius of bit). This relieves the corner giving the effect of a smaller radius bit. Visually comparing a panel cut with the 1/16" bit and the 2mm bit it is hard to see the over cut in the corners. Just this simple change saved a tool change or need for a two pass operation and saved 6 minutes per part! I have another product I think I can use the over cut on to save a tool change as well.

The main points are to not over complicate tool changing. Often you can eliminate the need for a tool change by rethinking how your part is deigned or how it will be machined. You can also use a tool changing script combined with an input to make tool changes faster and more accurate.

While I agree with many of your points, IMO they still dont eliminate the need for repeatable tool changes--

Further, the a2z setup basically ignore the repeatability of a tapered interface, so there is likely room for improvement by comparison as far as runout is concerned.

Its obvious that a loss of runout and rigidity is inherent to anything that extends the tool holder. However, both of these can be minimized and are common practice in virtually every CNC machine shop operating today. The point to be made is, the minute loss in accuracy is a sacrifice most are willing to make-- the benefits outweight the drawbacks.

I do concur with your sentiment about minimizing tool changes, it can often be a time saver. However the opposite is often true as well. I could offer many an instance where by adding a tool to handle the roughing, and another to do finishing work, has offset the toolchange time ten fold.

I also dont fancy my own design very much. Realistically, it gives you repeatable tool changes, but they're still going to be cumbersome. A tiny little drawbar would be the ticket..