This is more of a build thread than a "look what I made today" sort of thing like I've posted before in here, so it'll span a bit of time.
Many moons ago... I bought a capstan lathe for 100 quid, it was a "get it in the van quick" deal in that incurable toolaholic way justifying it as I could get something out of it even if it was worn out completely.
I got it home and started to inspect and it was a odd beast indeed, the entire bed was unsupported apart from at the headstock end and of round ram construction. Gauges & test bars came out and moving the saddle towards the tailstock end caused it to droop significantly measureably and leaning on the end caused a dti to do entire rotations.
Some research showed it to be a button lathe, that is to knock out buttons of material very close to the chuck in their thousands without much accuracy more than about 2" from the chuck with strange collets that I have never been able to identify. Not much use outside that environment although really well made apart from the obvious structural issue. I confirmed this with some other owners of the same lathes and there was all sorts of ideas to make brackets up to brace the end of the bed etc, but by then I had stolen the suds pump and 3 phase switching + motor stuff for another machine, and was eyeing up the stand for yet another project machine build. No parts of the carcass wasted here...
So I had a rather substantial headstock assembly and other bits and bobs kicking around that didn't owe me anything, I planned to make a dividing head for my manual mill, then I got into cnc and did the bridgeport interact ->linux cnc conversion, and got the urge for a 4th axis (rotary spindle on bed) setup.
Sawing the headstock off the casting, machining it flat in the manual mill and adding a htd pulley + stepper to it yielded this.
I actually got this working like this :-
https://www.youtube.com/watch?feature=player_detailpage&v=efnqily6NlA
But using it, it tended to drift off position in heavier cuts, the stepper didnt have enough holding torque with the 2:1 drive ratio to the spindle despite claims of massive holding torque by the datasheets from the chinese manufacturer, you could literally grip the pulley and move it by hand against the braking torque. The stepper was a 1200oz/inch nema 34 unit driven by a dm880a driver fed with a true 70v, same as the main machine axis. I started to add a spindle brake with a mini moto disk brake actuated by a 12v linear actuator but then you can't do co-ordinated moves between all axis for true 3d cnc on all axis as you have the A axis locked by the brake between repositions.
Then John Stevenson made a comment thats exactly how to solve the issue, compound the drive giving it x times less speed but x times more holding torque from the same stepper. I can deal with a slower axis as I'm not doing production on it so tonight here we go. I dicked about pricing up harmonic drives, and it looked like I couldn't do it for less than about 350 quid for a decent quality drive. The chinese ones are rubbish with slop where they need to have none causing accuracy issues I've been told.
So I took a old bike wheel spindle, or swingarm spindle, not sure which, but it was 20mm and in the scrap box to be used. Cut it to length, then grind the end of it to 16.00 mm to match the 16.01mm hole I machined into a 15t htd pulley first on the lathe. I did the grinding in by mounting it in a 20mm 5c collet and putting a spin indexer on the magnet, then cranked the spindexer by hand to generate the rotation. Towards the end I cut the wheel square with the diamond to keep it all parallel. These spindexers are cheap and surprisingly accurate once you get the burrs and stuff out of them, and before now I've span them with a cordless drill on a bigger rotary grinding job and used them on the mill during setups too as I don't have a working division head.
And you can also use them as division indexers :-)
ebay link if anyone is curious, no connection just the same units everyone sells.
http://www.ebay.co.uk/itm/5C-Spin-I...arrive to finish it in its latest incarnation.
Many moons ago... I bought a capstan lathe for 100 quid, it was a "get it in the van quick" deal in that incurable toolaholic way justifying it as I could get something out of it even if it was worn out completely.
I got it home and started to inspect and it was a odd beast indeed, the entire bed was unsupported apart from at the headstock end and of round ram construction. Gauges & test bars came out and moving the saddle towards the tailstock end caused it to droop significantly measureably and leaning on the end caused a dti to do entire rotations.
Some research showed it to be a button lathe, that is to knock out buttons of material very close to the chuck in their thousands without much accuracy more than about 2" from the chuck with strange collets that I have never been able to identify. Not much use outside that environment although really well made apart from the obvious structural issue. I confirmed this with some other owners of the same lathes and there was all sorts of ideas to make brackets up to brace the end of the bed etc, but by then I had stolen the suds pump and 3 phase switching + motor stuff for another machine, and was eyeing up the stand for yet another project machine build. No parts of the carcass wasted here...
So I had a rather substantial headstock assembly and other bits and bobs kicking around that didn't owe me anything, I planned to make a dividing head for my manual mill, then I got into cnc and did the bridgeport interact ->linux cnc conversion, and got the urge for a 4th axis (rotary spindle on bed) setup.
Sawing the headstock off the casting, machining it flat in the manual mill and adding a htd pulley + stepper to it yielded this.
I actually got this working like this :-
https://www.youtube.com/watch?feature=player_detailpage&v=efnqily6NlA
But using it, it tended to drift off position in heavier cuts, the stepper didnt have enough holding torque with the 2:1 drive ratio to the spindle despite claims of massive holding torque by the datasheets from the chinese manufacturer, you could literally grip the pulley and move it by hand against the braking torque. The stepper was a 1200oz/inch nema 34 unit driven by a dm880a driver fed with a true 70v, same as the main machine axis. I started to add a spindle brake with a mini moto disk brake actuated by a 12v linear actuator but then you can't do co-ordinated moves between all axis for true 3d cnc on all axis as you have the A axis locked by the brake between repositions.
Then John Stevenson made a comment thats exactly how to solve the issue, compound the drive giving it x times less speed but x times more holding torque from the same stepper. I can deal with a slower axis as I'm not doing production on it so tonight here we go. I dicked about pricing up harmonic drives, and it looked like I couldn't do it for less than about 350 quid for a decent quality drive. The chinese ones are rubbish with slop where they need to have none causing accuracy issues I've been told.
So I took a old bike wheel spindle, or swingarm spindle, not sure which, but it was 20mm and in the scrap box to be used. Cut it to length, then grind the end of it to 16.00 mm to match the 16.01mm hole I machined into a 15t htd pulley first on the lathe. I did the grinding in by mounting it in a 20mm 5c collet and putting a spin indexer on the magnet, then cranked the spindexer by hand to generate the rotation. Towards the end I cut the wheel square with the diamond to keep it all parallel. These spindexers are cheap and surprisingly accurate once you get the burrs and stuff out of them, and before now I've span them with a cordless drill on a bigger rotary grinding job and used them on the mill during setups too as I don't have a working division head.
And you can also use them as division indexers :-)
ebay link if anyone is curious, no connection just the same units everyone sells.
http://www.ebay.co.uk/itm/5C-Spin-I...arrive to finish it in its latest incarnation.