On Thursday 13 December 2018 20:48:35 Erik Christiansen wrote:

> On 11.05.18 14:28, andy pugh wrote:
> ...
>
> > Though, actually, what I would do is cut a slot in the top and rely
> > on two flats only to do the work.
>
> That one appeals a lot from the manufacturing viewpoint. It is quicker
> and probably as strong as the 4 grubscrews I used in the original
> taphats. It took me a while to visualise that not all of the
> through-hole is lost, just the AF width of the tap's drive square, so
> the (centre)drill in the chuck can still serve to guide the taphat -
> the purpose of the original.
>
> The grubscrews also avoid any backlash which might be an issue in
> rigid tapping. (Gene uses them under CNC, IIUC - something I haven't
> tried.)
>
> Erik
Yes, I wound up with the various "stations" of the hat make, scattered 
along the length of the g0704's table. 

The center hole is drilled with a stationary drill sticking up from a 
chuck mounted to the table, with the 7/8" brass slug in the r8.  Drilled 
all the way thru.

Once drilled, was transferred to an alu block jig with a slot cut in the 
side so a clamp on the unslit side held it in place, and a clamp over 
the slit squeezed the jig to lock the slug, and the 4mm screw slot 
machined and tapped so that about 2/3rds of the 4mm cap screws head 
projected above the side of the slug.

The slug was then equiped with the screw, and transferred to another 7/8" 
R8 with a slot cut in the side to accept that screw, which locks it from 
any rotational slippage, that r8 is chucked in a 3" lathe chuck mounted 
to a motorized table I made, and a piece of scrap alu brought up to hold 
the slug in the r8 while the table was rotated to drill and rigid tap 
the slugs 4 grub screw holes. The semi loose slug was supported against 
the cutting forces my a maple block of wood with half a 7/8 hole 
machined into the top, with a tool length contact of a piece of 
dbl-sided pcb super glued to the top face of the maple and hooked to the 
probe input to measure and set the length of the various drills and taps 
used, which of coarse varied with the size of the tap. Several tool 
changes of coarse but I could do from a 1.125" long cutoff done on the 
Sheldon, to a finished hat with the grub screws installed and the tap 
mounted in around 30 minutes apiece once I got going. The rotating table 
took up more space on the 704's table than anything else. So I've now a 
box with dividers makeing 10x pockets, and ATM about 55 taps all mounted 
and ready to go, all my taps up to  10mm or 3/8" are now mounted. The 1 
horse, pushed quite a bit past the 1hp mark, can't pull the 10mm or 3/8" 
in steel without peck looping, which I pause at the top to blow the 
swarf out and re-anoint with fresh cutting oil. So I've gotten into the 
habit of pecking everything above 6mm or 1/4", driving them deeper by as 
little as 1/4 turn per peck.

I have also found, with my rather eclectic collection of taps, that few 
are actually straight, so a sloppy hole in the hat allows it to wibble 
slightly, self centering in the hole so I get a tighter fitting thread. 
Some of these taps are bargain bin junk, re-sharpened over the years.

================== more comments you may find informative =============

I've added some hal code to measure the overshoot at the bottom of the 
hole, so I can be sure I don't hit the bottom of the hole and break the 
tap. EDM-ing broken taps out of a hole is a pita.

The g0704 can do a reversal in a couple hundred milliseconds, so I can 
run the tap at 400-600 revs, giving me the inertial mass to drive the 
bigger taps without the 5 to 15 horse motor of a bigger mill.

That hal code is much more important on the lathe, with its vfd and a 40 
lb chuck, reversal overshoots can easily exceed 3 turns! And the drive 
belts are yelping. 

And at higher speeds, the Sheldon's vfd comes out of 4 quadrant control 
so I actually have a limit3 in the path to profile the turnaround to 
something this cheap vfd can do by frequency control alone. Gated by  
some "its stopped" detection logic before it actually gets signalled to 
start in the other direction. Timers watching encoder, and when the 
pulses are slow enough, its considered stopped. Works nicely and things 
don't get commanded to do something it can't do.

Cheap vfd. No braking R's, or facilities to hook them up, so its a huge 
compromise between speed and amount of DC braking it can do. But I try 
to not exceed the motors nameplate FLA except for its dc braking, which 
is extremely transient, which seems to be about the best it can do and 
both it and the motor survive w/o throwing a fault.

I have found I can run for virtually unlimited time at its FLA and 9 or 
10 hz, with buckets of torque, running that way for half an hour and I 
can still lay my hand on the motor, uncomfortable of coarse but I can.

First vfd I ever programmed, so possibly it could be done better by a 
real expert at tuning a cheap clone. I don't claim to be one.

-- 
Cheers, Gene Heskett
--
"There are four boxes to be used in defense of liberty:
 soap, ballot, jury, and ammo. Please use in that order."
-Ed Howdershelt (Author)
Genes Web page <http://geneslinuxbox.net:6309/gene>


_______________________________________________
Emc-users mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/emc-users

Reply via email to