Drill chuck for a mill
- Thread starter trlvn
- Start date
trlvn
Ultra Member
The stock for each wedge was 1/8 inch in thickness*. Before I started, I checked that the gap was less than 1/4" total. I had previously soaked the chuck with penetrating oil as there was some rust and it wouldn't move freely from closed to fully open.
With hand pressure, the wedges were well overlapped. Lacking a big-enough vise, I then used a C-clamp to press them together further. I added a little machinist clamp to the end of one of the wedges to help stabilize the floppy end of the C-clamp. At this point, one of the wedges was a lot further in than the other. Alas, I couldn't control the C-clamp very well. So I got out a big hammer!
I put one wedge down on an anvil and smacked the other wedge sharply with the hammer. The chuck popped off on about the third hit. One of the wedges now has a slight bow to it but it would be easy enough to make another one. I understand that commercial wedge kits use hardened and tempered steel.
Craig
* Same as in MrPete's video.
trlvn
Ultra Member
I picked up an R8-to-JT33 drill chuck arbor from the KBC Tools Mississauga store, yesterday. I thought I'd check the run out on the arbor before mounting my old Rohm 1/2" chuck on it. It is not very good. Near the small end of the jacobs taper, there is over 3 thous of run out although at the big end it is only about 1/2 thou. I can only check the big end of the R8 in my spindle but it is about 1/2 thou as well.
With my fingers, I can feel that the surface of the drill chuck arbor is not uniform. Sort of a lump on one side.
I called KBC and apparently I just order another one and return the bad one for their inspection. Hopefully the new one will be better.
Is there a good way to clean up inside the R8 taper on my mill? Like a bottle brush?
Craig
With my fingers, I can feel that the surface of the drill chuck arbor is not uniform. Sort of a lump on one side.
I called KBC and apparently I just order another one and return the bad one for their inspection. Hopefully the new one will be better.
Is there a good way to clean up inside the R8 taper on my mill? Like a bottle brush?
Craig
trlvn
Ultra Member
Ahh. Searching for that term, someone has shared a 3D-printable version on Thingiverse:
https://www.thingiverse.com/thing:2181640
As a bonus, someone else has also shared a design for a pair of nested knobs. If I understand correctly, the larger one engages the spindle and the smaller one drives the drawbar. Clever!
https://www.thingiverse.com/thing:2746893
I can see that a trip to the library's 3D printer station is in my future.
Craig
https://www.thingiverse.com/thing:2181640
As a bonus, someone else has also shared a design for a pair of nested knobs. If I understand correctly, the larger one engages the spindle and the smaller one drives the drawbar. Clever!
https://www.thingiverse.com/thing:2746893
I can see that a trip to the library's 3D printer station is in my future.
Craig
Tom Kitta
Ultra Member
I use ER32 collets for everything including drilling on my primary mill - mainly b/c it is a huge pain for lager mills without ATC to change tools between collet holder and drill chuck.
On smaller mill like yours people generally use R8 collets and thus every change in say endmil is a change in the tool.
Imagine if you got R8-ER32 holder for example with full set of ER32s - now changing between endmills could be done by just working on the collet chuck - no need to touch the draw bar at all. This is of interest if your draw bar end is 7ft of the ground.
On large mills it is a bit faster to use ATC vs ER32s chuck holder.
- With ATC you press a button top pop out old R8 collet - time around 5sec.
- locate new R8 and put endmill / drill etc in it - say 60 sec
- use ATC to put collet in - 5sec & done
- With ER32 chuck - you need to remove the old collet with a spanner wrench - about 30sec
- locate new Er32 and put in endmill - say 60sec
- tie it in - say 30sec
so ATC wins by 50sec.
Using draw bar on large mill:
- Find 40 taper tool holder and place end mill in it - 120 sec if you are prepared - frequently need to tighten weldon shank
- stand on something to undo the draw bar and remove old 40 taper - at least 60 sec, sometimes 120 sec if you need to move table over
- put in new tool holder in - at least 60 sec
So tool change without ATC on large mill can easily take 5 min.
On smaller mill like yours people generally use R8 collets and thus every change in say endmil is a change in the tool.
Imagine if you got R8-ER32 holder for example with full set of ER32s - now changing between endmills could be done by just working on the collet chuck - no need to touch the draw bar at all. This is of interest if your draw bar end is 7ft of the ground.
On large mills it is a bit faster to use ATC vs ER32s chuck holder.
- With ATC you press a button top pop out old R8 collet - time around 5sec.
- locate new R8 and put endmill / drill etc in it - say 60 sec
- use ATC to put collet in - 5sec & done
- With ER32 chuck - you need to remove the old collet with a spanner wrench - about 30sec
- locate new Er32 and put in endmill - say 60sec
- tie it in - say 30sec
so ATC wins by 50sec.
Using draw bar on large mill:
- Find 40 taper tool holder and place end mill in it - 120 sec if you are prepared - frequently need to tighten weldon shank
- stand on something to undo the draw bar and remove old 40 taper - at least 60 sec, sometimes 120 sec if you need to move table over
- put in new tool holder in - at least 60 sec
So tool change without ATC on large mill can easily take 5 min.
Crankit
Well-Known Member
I bought a used high quality Japanese keyless chuck that was on a 40MT and boy was that thing stuck on! Finally took heat and a can of super cold spray with the JT wedges to get that sucker off.
Cheers
Wayne
trlvn
Ultra Member
These KBC arbors are crap! I picked up a replacement and the second appears to be worse than the first. The dial test indicator jumps around erratically as I rotate the spindle. There appears to be a 2 thou valley in one spot and a 4 thou high ridge in another. This is all on the jacobs taper portion of the arbor. I can measure the portion that protrudes from the R8 spindle and it shows less than 1 thou of TIR.
I'm going to systematically measure it again. Surely it is not that hard to grind such arbors to final size, right? I expect that the arbor should add less than one thou of run out. Well less!
Unfortunately, BusyBee does not even list an R8 to JT33 arbor. That means I'm stuck waiting for a mail order and an even more annoying return or exchange process if I get another one that is poorly made.
Thanks for letting me vent.
Craig
I'm going to systematically measure it again. Surely it is not that hard to grind such arbors to final size, right? I expect that the arbor should add less than one thou of run out. Well less!
Unfortunately, BusyBee does not even list an R8 to JT33 arbor. That means I'm stuck waiting for a mail order and an even more annoying return or exchange process if I get another one that is poorly made.
Thanks for letting me vent.
Craig
Brent H
Ultra Member
Hey Craig,
That sucks! You can try an assembly of the arbour and chuck and see if together they are poor?
I just picked up three old jacobs chucks at a “treasure” store for $15. The one has a straight shank so when I get home I’ll see if it is any good. The other two are small pin type chucks.
Brent
That sucks! You can try an assembly of the arbour and chuck and see if together they are poor?
I just picked up three old jacobs chucks at a “treasure” store for $15. The one has a straight shank so when I get home I’ll see if it is any good. The other two are small pin type chucks.
Brent
Tom Kitta
Ultra Member
Well I got NMTB40 to jackobs something taper. When I placed a chuck onto that sucker it wobbled. So its not just your bad luck - it seems the little Jacobs taper for some unknown reason is hard for the Chinese to grind properly. Which is strange - Jacobs sizing is universal AFAIK.
Tom Kitta
Ultra Member
Well I got NMTB40 to jackobs something taper. When I placed a chuck onto that sucker it wobbled. So its not just your bad luck - it seems the little Jacobs taper for some unknown reason is hard for the Chinese to grind properly. Which is strange - Jacobs sizing is universal AFAIK.
trlvn
Ultra Member
Thank you! I'm going to order this arbor and see if it is any better. It is actually cheaper.
Craig
Accusize is redistributor for mostly Chinese stuff. You might be OK or it may well be KBC's cousin. Typical coin toss.
One thing you might want to check - sometimes the set screw in your R8 quill can be set a bit deep. You might not notice any difference on some tools because the slot is sufficiently deep & screw is shy of contacting. But other tooling shanks can have a slightly shallower slot. If the screw rubs it could potentially affect the tapered portion from fully seating. You would probably feel this as being stickier inserting & releasing. I rather doubt it, but something to tuck away in the noggin.
Another thing is make sure your drawbar is fully sucking the arbor in. I've had one that was tapped shallow for some offshore reason & the arbor taper was just barely in contact with the spindle taper so had runout & vibration. I've heard others run into this & have to add a spacer under the drawbar nut or else slightly reduce the drawbar length if that was the culprit.
But more than likely you have a Friday 4:25PM grind
One thing you might want to check - sometimes the set screw in your R8 quill can be set a bit deep. You might not notice any difference on some tools because the slot is sufficiently deep & screw is shy of contacting. But other tooling shanks can have a slightly shallower slot. If the screw rubs it could potentially affect the tapered portion from fully seating. You would probably feel this as being stickier inserting & releasing. I rather doubt it, but something to tuck away in the noggin.
Another thing is make sure your drawbar is fully sucking the arbor in. I've had one that was tapped shallow for some offshore reason & the arbor taper was just barely in contact with the spindle taper so had runout & vibration. I've heard others run into this & have to add a spacer under the drawbar nut or else slightly reduce the drawbar length if that was the culprit.
But more than likely you have a Friday 4:25PM grind
Attachments
Last edited:
trlvn
Ultra Member
@PeterT Thanks for the suggestions. I'll compare the slot depth on the latest arbor with some others.
I'm pretty sure the arbor is being fully seated. It is firmly wedged in the taper when I try to remove it. Like the others, I have to loosen the drawbar a couple of turns and whack it with a brass hammer to get it to release.
Fingers crossed that the arbor from Accusize is a bit better.
BTW, among some other stuff that I bought, I got a homemade R8-to-J6 arbor that I'd almost forgotten about. I'm going to throw it in the spindle and see what kind of runout it has. The guy I bought it from has been an enthusiastic model maker for years--turns and grinds his own crankshafts for internal combustion engines. My guess is that any runout on it will be from the spindle and not the arbor!
Craig
I'm pretty sure the arbor is being fully seated. It is firmly wedged in the taper when I try to remove it. Like the others, I have to loosen the drawbar a couple of turns and whack it with a brass hammer to get it to release.
Fingers crossed that the arbor from Accusize is a bit better.
BTW, among some other stuff that I bought, I got a homemade R8-to-J6 arbor that I'd almost forgotten about. I'm going to throw it in the spindle and see what kind of runout it has. The guy I bought it from has been an enthusiastic model maker for years--turns and grinds his own crankshafts for internal combustion engines. My guess is that any runout on it will be from the spindle and not the arbor!
Craig
Tom Kitta
Ultra Member
I routinely match Chinese quality and frequently exceed it so I would not be shocked if quality home made stuff was superior to the cheap China stuff.
One has to understand that the price at the factory door is 1/4 of the price on Amazon or so - thus they need to make these things for around 4.50 CAD max including profit for the factory.
One has to understand that the price at the factory door is 1/4 of the price on Amazon or so - thus they need to make these things for around 4.50 CAD max including profit for the factory.
trlvn
Ultra Member
If it is worth doing, it is worth overdoing. Right?
I was getting what almost seemed to be erratic readings with the dial test indicator trying to measure the runout on the KBC arbor (1-507-305). So I marked the spindle pulley in 8 roughly equal segments. On the clock, that would be 12:00, 1:30, 3:00, 4:30, etc.
On the arbor, I noticed that readings near the big end of the jacobs taper seemed to be very different from those near the small end. So I marked an arbitrary spot near the big end, near the small end and roughly in the middle. I used a "Verdict" dial test indicator marked in 0.0005 graduations and interpolated to the nearest 0.0025.
I did 3 sets of measurements removing the arbor and reinstalling it before each set.
My measurements are in the attached pdf. BTW, I have a crappy fine adjust on the mag base so in the raw data if I got the starting point (12:00) within 1 thou, I would start the measurements from there. What matters is the increase or decease in the value as I rotated the spindle. If the arbor was perfect, it would never change from the starting value. I also recorded the measurement when I got back to 12:00 as a sanity check--if I was within 1/2 thou of the starting measurement, I figured I hadn't screwed up too badly.
I was surprised at how much variation there was between trials. I thought that R8 tapers would seat in a very repeatable fashion. Trial 1 and Trial 3 are pretty similar. Maybe I screwed up something in Trial 2--not sure.
Nonetheless, the measurements showed that the arbor is a fine approximation of a potato: lumpy!! Especially the big half of the taper. Hills and valleys. More than 3 thous of variation from the low spots to the high spots.
Anyway, I've returned the arbor to KBC. The guy at the counter asked a few questions to be sure he understood my measurements and then refunded my money with no hassle.
And I was wrong about the homemade arbor (R8 to JT6). I did a quick test on it and the runout exceeded the 10 thou range of the DTI I was using!!
Craig
I was getting what almost seemed to be erratic readings with the dial test indicator trying to measure the runout on the KBC arbor (1-507-305). So I marked the spindle pulley in 8 roughly equal segments. On the clock, that would be 12:00, 1:30, 3:00, 4:30, etc.
On the arbor, I noticed that readings near the big end of the jacobs taper seemed to be very different from those near the small end. So I marked an arbitrary spot near the big end, near the small end and roughly in the middle. I used a "Verdict" dial test indicator marked in 0.0005 graduations and interpolated to the nearest 0.0025.
I did 3 sets of measurements removing the arbor and reinstalling it before each set.
My measurements are in the attached pdf. BTW, I have a crappy fine adjust on the mag base so in the raw data if I got the starting point (12:00) within 1 thou, I would start the measurements from there. What matters is the increase or decease in the value as I rotated the spindle. If the arbor was perfect, it would never change from the starting value. I also recorded the measurement when I got back to 12:00 as a sanity check--if I was within 1/2 thou of the starting measurement, I figured I hadn't screwed up too badly.
I was surprised at how much variation there was between trials. I thought that R8 tapers would seat in a very repeatable fashion. Trial 1 and Trial 3 are pretty similar. Maybe I screwed up something in Trial 2--not sure.
Nonetheless, the measurements showed that the arbor is a fine approximation of a potato: lumpy!! Especially the big half of the taper. Hills and valleys. More than 3 thous of variation from the low spots to the high spots.
Anyway, I've returned the arbor to KBC. The guy at the counter asked a few questions to be sure he understood my measurements and then refunded my money with no hassle.
And I was wrong about the homemade arbor (R8 to JT6). I did a quick test on it and the runout exceeded the 10 thou range of the DTI I was using!!
Craig
Last edited:
trlvn
Ultra Member
One other small thing about KBC. I ordered the original arbor last Sunday and picked it up Monday. Later Monday, after finding the runout on the arbor, I noticed that the web price was more than $5 less than I paid. When I exchanged it for the second arbor on Wednesday, the counter guy said that it was part of the group of products that went onto a weekly special starting Monday. He offered no apology about the fact that I didn't get the 20% discount when I picked up the product on Monday.
That left a bit of a bad taste in my mouth. If it was on special when I picked it up on Monday, I think I should have got the special price. I've been involved in enough mail order systems to know that good companies treat their customers better than that.
FWIW,
Craig
That left a bit of a bad taste in my mouth. If it was on special when I picked it up on Monday, I think I should have got the special price. I've been involved in enough mail order systems to know that good companies treat their customers better than that.
FWIW,
Craig
Do the ends of your arbor still have center drills like what they may have used to set it up originally when grinding? It would be interesting if you could support it on 2 centers & rotate if the ground tapered surfaces are egged. Of course there could be other distortions that might be giving you runout but I would expect their grinding would be the very last step they would do after hardening etc. & presumably stay that way.
I have a similar dud, shiny potato syndrome. I also have an older gen USA Jacobs = bang smack on (but was more $$ at the time). Some of the Eastern Europe stuff can be quite good, but the price is up there too.
I have a similar dud, shiny potato syndrome. I also have an older gen USA Jacobs = bang smack on (but was more $$ at the time). Some of the Eastern Europe stuff can be quite good, but the price is up there too.