D1-4 backplate – or how to spend over 12 hours with thousands of dollars of machinery to save one hundred and thirty dollars
One of the reasons my lathe was kind of cheap was that when I bought it, none of the tooling it came with was immediately useful. There was a (partially complete) set of collets in a wooden box, but no spindle nose adapter for them. There was a tool post with only one extremely ugly, obviously home made tool holder. And there was a rather nice looking 3 jaw chuck with no backplate.
Before I had the lathe powered up, I traded the chuck to a buddy for two beat up direct mount D1-4 chucks – a 6” 3 jaw and a 8” 4 jaw. At the time I had no mill so this was a pretty sweet deal. The 3 jaw was especially hammered, but after grinding in the jaws with a thrown-together tool post grinder it seemed fine and served me well for the next 4 years…
Until a couple weeks ago. I mounted the chuck and watched as the work piece wobbled around. It seemed bad (?), so I removed the chuck and was shocked to find that the back had cracks around where the pull studs go. In hindsight this should have been unsurprising, given the cosmetic condition of the chuck… it should have been melted down into pots and pans years before I got it. But anyways.
uh oh!
so here I had a bit of a dilemma. After a bit of researching prices of chucks and backplates and despairing, I figured I could make a backplate at work for 0$ out of stolen material and “borrow” a chuck from my deckel dividing head so that I could spend the money on cat wine and cigarettes. I mean, how much work could it be?
so we start, as per tradition, with a bit of shoddy cad. Actually this took me way too much time (3hrs?). here’s the drawing to save you a bit of hassle if you want to make your own but I can’t guarantee this is 100% correct! Most of the dims are gleaned from various spec sheets online but some of them I had to do a bit of fake CMM / guesswork.
ye olde ersatz cmm - a centering microscope in the mill. actually this was not a great idea since the field of view of that microscope is like 5mm
again I can't guarantee these are 100% correct but at least it's a better starting point than I had. the backplate worked OK for me, the stud I've actually not made yet
Off to the saw. We start with a slice of 6” CRS I am stealing from my boss. Cut time with a blade missing 20% of its teeth: 5 mins?
screetch screetch screetch etc etc
And then off to the CNC lathe at work.
I took a test bore and checked it with a dial bore gauge, comped the tool offset and then let the machine do the final op. the taper finish is sublime. that’s a CNC magic trick – with flood coolant and good carbide inserts even crappy mild steel can have a brilliant finish. The face finish isn’t so good because the boring bar didn’t break the chip and dragged a ball across the entire face. This actually doesn’t matter since we will be grinding in that face later anyways.
start to finish, I was behind the lathe for something like 5 hours. between boring in the jaws and figuring out mazatrol, CNC was not a timesaver...
And back to my shop to test fit. We have something like 7 tenths of radial play.
don't look at the indicator too closely, this is a screen grab from a video and I didn't zero it
I guess while we’re here I should talk a bit about the geometry. Both the taper and the face need to make contact at the same time, which means the (tapered) bore diameter has to be pretty much bang on 2.500” at the spindle face of the back plate. This is impossible (?) to measure directly (you could use a CMM and probe two depths of the taper to find the diameter and angle, and then math it out to the face) but it’s quite simple to assess the fit on the lathe spindle. If the bore is too big, the face will make contact but not the taper, and there will be radial play. If the bore is too small, then there will be a gap between the backplate and the spindle face that you could feel out with either a feeler gauge or Prussian blue. obviously, if we’re gonna shoot one way or the other, it’s preferrable to have a slightly larger bore because removing material from the (flat) face is much easier than re-cutting the tapered bore.
in this case we have 9 tenths (0.0009") of radial play total, which means that the radius of the bore is too big by 0.00045". the taper angle is 1:8 so, to for both faces to contact at the same time, 8 x the radial clearance needs to be ground off the spindle side (=0.0036" = 3.6 thou). hopefully this makes sense, I ought to draw a diagram but I'm too lazy
Drilling and tapping etc… I do this in a "polar coordinates type way" with the dividing head because it’s super simple with my mill, and I think it's faster to only change the rotary axis to do each feature of a hole at a time. The only exception being the set of holes for the ¼-20 lock screws, since I don’t have an indexing stop at 16° I do all the ops for one position at a time to avoid having errors in concentricity. Also because the counterbores overlap, I counterbore with an endmill in a collet before drilling to tap size, otherwise the drill would go sideways through a thin wall (upon further thought, maybe this should have let me know my drawing was off). Total time for this: about 2 hours
funny enough, that chuck holding the backplate is the one I was stealing for the lathe
Test assembly. I stole the camlock pins off of the old chuck and found a mistake in my drawing – the ¼-20 lock screws were too close. Whoops! hence the angle is now 16° from the 15° I had originally guessed. I actually don't know if this is true, I just ended up turning down the heads of the bolts lol.
Then back to the lathe. Slap some indicators on it and we have 0.002” radial runout (?) and 0.001” on the face. Not the worst? A facing cut trues up the face (or more likely with my beat up lathe, makes it concave or convex) and an OD turn brings it to the chuck size, and also creates a huge mess and sends blue chips directly into my sandals and socks. Time for turning, facing and cleanup: probably an hour
Oh ya – I took apart the chuck at some point and it was filled with some turd coloured schmoo. Maybe this is a coolant thing – I had a rotary table that had a rotten worm and wheel, which I presume was from coolant. Fortunately there’s no real damage but to clean it out was another hour or more faffing around with the parts cleaner outside (in the cold) and still it stank up the entire shop when I brought it inside.
cleaning an old machine tool filled with some sort of disgusting paste - a hobby machining classic
At this point I couldn’t do anything with it until I went back to work (no surface grinder at my shop), so I used this as an excuse to play with the big roll engraving attachment for my pantograph. We’ll put a “1” next to camlock stud #1 so that I know how to orient the chuck on the lathe spindle. The roll engraving attachment weighs a bullion pounds and setting it up takes half a century, so engraving this 1 takes me about an hour.
While I’m here I engrave some more rubbish on the perimeter – why not. Here we see the effects of one of the pivots of the roll engraving attachment being worn out and wobbly – the engraved text comes out wobbly as well. I shouldn’t be surprised that there’s some play in this guy since it’s hammered enough that it’s missing all its paint. I ended up fixing this with a shim which was no big deal...
fun fact: when I bought this panto it was under a tarp in a field
Now to the surface grinder at work to get the fit perfect. This was my first time using a surface grinder so I’ll refrain from any real commentary…
hydraulic auto feed so we just dial in the dept of cut and let er rip eh
For the last pass on the spindle face I did the rotato potato technique (dunno what it’s actually called) which is supposed to give very flat parts. Dunno why (if I had to guess, it’s because any error from the Y axis of the machine is absent since that axis is not used); I just used it because the pattern it gives is pretty.
a 1/4" wide wheel meant that this was alot of rotations. don't ask me why I didn't change to a wider wheel, I'm pretty stupid. oh also those pull studs I just stole off the busted old chuck
And the finishing touch is to turn a short shoulder that matches the ID of the back face of the chuck. A quick chamfer of the holes and stoning of the face and the chuck can be bolted up finally.
And the final result… well it’s disappointing. 0.004” of TIR with a dowel pin in the chuck. I feel like I did my part - so like any good craftsman I'm gonna blame the 70 year old chuck. I’m thinking about grinding in the jaws but just this weekend the power feed on the lathe stopped working (DC motor – it could be a number of issues) so I will have to tackle that first... the fun never stops eh
the finished product. well at least it looks nice??
So was it worth it? Not really? but now that I have this process (and CNC lathe setup and program) figured out of the way next time I will make a set true chuck or something. I just need to find 400$ for a nice ish chuck first...
One of the reasons my lathe was kind of cheap was that when I bought it, none of the tooling it came with was immediately useful. There was a (partially complete) set of collets in a wooden box, but no spindle nose adapter for them. There was a tool post with only one extremely ugly, obviously home made tool holder. And there was a rather nice looking 3 jaw chuck with no backplate.
Before I had the lathe powered up, I traded the chuck to a buddy for two beat up direct mount D1-4 chucks – a 6” 3 jaw and a 8” 4 jaw. At the time I had no mill so this was a pretty sweet deal. The 3 jaw was especially hammered, but after grinding in the jaws with a thrown-together tool post grinder it seemed fine and served me well for the next 4 years…
Until a couple weeks ago. I mounted the chuck and watched as the work piece wobbled around. It seemed bad (?), so I removed the chuck and was shocked to find that the back had cracks around where the pull studs go. In hindsight this should have been unsurprising, given the cosmetic condition of the chuck… it should have been melted down into pots and pans years before I got it. But anyways.
uh oh!
so here I had a bit of a dilemma. After a bit of researching prices of chucks and backplates and despairing, I figured I could make a backplate at work for 0$ out of stolen material and “borrow” a chuck from my deckel dividing head so that I could spend the money on cat wine and cigarettes. I mean, how much work could it be?
so we start, as per tradition, with a bit of shoddy cad. Actually this took me way too much time (3hrs?). here’s the drawing to save you a bit of hassle if you want to make your own but I can’t guarantee this is 100% correct! Most of the dims are gleaned from various spec sheets online but some of them I had to do a bit of fake CMM / guesswork.
ye olde ersatz cmm - a centering microscope in the mill. actually this was not a great idea since the field of view of that microscope is like 5mm
again I can't guarantee these are 100% correct but at least it's a better starting point than I had. the backplate worked OK for me, the stud I've actually not made yet
Off to the saw. We start with a slice of 6” CRS I am stealing from my boss. Cut time with a blade missing 20% of its teeth: 5 mins?
screetch screetch screetch etc etc
And then off to the CNC lathe at work.
I took a test bore and checked it with a dial bore gauge, comped the tool offset and then let the machine do the final op. the taper finish is sublime. that’s a CNC magic trick – with flood coolant and good carbide inserts even crappy mild steel can have a brilliant finish. The face finish isn’t so good because the boring bar didn’t break the chip and dragged a ball across the entire face. This actually doesn’t matter since we will be grinding in that face later anyways.
start to finish, I was behind the lathe for something like 5 hours. between boring in the jaws and figuring out mazatrol, CNC was not a timesaver...
And back to my shop to test fit. We have something like 7 tenths of radial play.
don't look at the indicator too closely, this is a screen grab from a video and I didn't zero it
I guess while we’re here I should talk a bit about the geometry. Both the taper and the face need to make contact at the same time, which means the (tapered) bore diameter has to be pretty much bang on 2.500” at the spindle face of the back plate. This is impossible (?) to measure directly (you could use a CMM and probe two depths of the taper to find the diameter and angle, and then math it out to the face) but it’s quite simple to assess the fit on the lathe spindle. If the bore is too big, the face will make contact but not the taper, and there will be radial play. If the bore is too small, then there will be a gap between the backplate and the spindle face that you could feel out with either a feeler gauge or Prussian blue. obviously, if we’re gonna shoot one way or the other, it’s preferrable to have a slightly larger bore because removing material from the (flat) face is much easier than re-cutting the tapered bore.
in this case we have 9 tenths (0.0009") of radial play total, which means that the radius of the bore is too big by 0.00045". the taper angle is 1:8 so, to for both faces to contact at the same time, 8 x the radial clearance needs to be ground off the spindle side (=0.0036" = 3.6 thou). hopefully this makes sense, I ought to draw a diagram but I'm too lazy
Drilling and tapping etc… I do this in a "polar coordinates type way" with the dividing head because it’s super simple with my mill, and I think it's faster to only change the rotary axis to do each feature of a hole at a time. The only exception being the set of holes for the ¼-20 lock screws, since I don’t have an indexing stop at 16° I do all the ops for one position at a time to avoid having errors in concentricity. Also because the counterbores overlap, I counterbore with an endmill in a collet before drilling to tap size, otherwise the drill would go sideways through a thin wall (upon further thought, maybe this should have let me know my drawing was off). Total time for this: about 2 hours
funny enough, that chuck holding the backplate is the one I was stealing for the lathe
Test assembly. I stole the camlock pins off of the old chuck and found a mistake in my drawing – the ¼-20 lock screws were too close. Whoops! hence the angle is now 16° from the 15° I had originally guessed. I actually don't know if this is true, I just ended up turning down the heads of the bolts lol.
Then back to the lathe. Slap some indicators on it and we have 0.002” radial runout (?) and 0.001” on the face. Not the worst? A facing cut trues up the face (or more likely with my beat up lathe, makes it concave or convex) and an OD turn brings it to the chuck size, and also creates a huge mess and sends blue chips directly into my sandals and socks. Time for turning, facing and cleanup: probably an hour
Oh ya – I took apart the chuck at some point and it was filled with some turd coloured schmoo. Maybe this is a coolant thing – I had a rotary table that had a rotten worm and wheel, which I presume was from coolant. Fortunately there’s no real damage but to clean it out was another hour or more faffing around with the parts cleaner outside (in the cold) and still it stank up the entire shop when I brought it inside.
cleaning an old machine tool filled with some sort of disgusting paste - a hobby machining classic
At this point I couldn’t do anything with it until I went back to work (no surface grinder at my shop), so I used this as an excuse to play with the big roll engraving attachment for my pantograph. We’ll put a “1” next to camlock stud #1 so that I know how to orient the chuck on the lathe spindle. The roll engraving attachment weighs a bullion pounds and setting it up takes half a century, so engraving this 1 takes me about an hour.
While I’m here I engrave some more rubbish on the perimeter – why not. Here we see the effects of one of the pivots of the roll engraving attachment being worn out and wobbly – the engraved text comes out wobbly as well. I shouldn’t be surprised that there’s some play in this guy since it’s hammered enough that it’s missing all its paint. I ended up fixing this with a shim which was no big deal...
fun fact: when I bought this panto it was under a tarp in a field
Now to the surface grinder at work to get the fit perfect. This was my first time using a surface grinder so I’ll refrain from any real commentary…
hydraulic auto feed so we just dial in the dept of cut and let er rip eh
For the last pass on the spindle face I did the rotato potato technique (dunno what it’s actually called) which is supposed to give very flat parts. Dunno why (if I had to guess, it’s because any error from the Y axis of the machine is absent since that axis is not used); I just used it because the pattern it gives is pretty.
a 1/4" wide wheel meant that this was alot of rotations. don't ask me why I didn't change to a wider wheel, I'm pretty stupid. oh also those pull studs I just stole off the busted old chuck
And the finishing touch is to turn a short shoulder that matches the ID of the back face of the chuck. A quick chamfer of the holes and stoning of the face and the chuck can be bolted up finally.
And the final result… well it’s disappointing. 0.004” of TIR with a dowel pin in the chuck. I feel like I did my part - so like any good craftsman I'm gonna blame the 70 year old chuck. I’m thinking about grinding in the jaws but just this weekend the power feed on the lathe stopped working (DC motor – it could be a number of issues) so I will have to tackle that first... the fun never stops eh
the finished product. well at least it looks nice??
So was it worth it? Not really? but now that I have this process (and CNC lathe setup and program) figured out of the way next time I will make a set true chuck or something. I just need to find 400$ for a nice ish chuck first...
