On Saturday 16 February 2019 20:01:38 Gene Heskett wrote:
> On Saturday 16 February 2019 14:28:33 Chris Albertson wrote:
> > Thinking about this as an electrical engineer, I'd say don't use
> > DC. DC current is as you say blocked by paint and oil films. Use
> > AC. I think a low radio frequency. Then the DC insulators would
> > act like capacitors and pass AC. even while blocking DC. To detect
> > contact use an AC voltage sensor, typically a diode and small
> > capacitor.
>
> The disadvantage there is the rc charge time. This means among other
> things a relatively leasure velocity to the first trip, and a much
> slower 2nd approach, where as in using the dc circuit, one usually
> uses a rotating with the spindle probe so even if it has some
> runnout, the contact surface is recorded an by the first contact
> discharging a .1 ufcapacitor, which takes long enough to recharge that
> the contacts logic zero, is captured and reported to LCNC instantly
> even if by the time the servo thread actually reads it, the contact
> has been lost again for .9 milliseconds. The rf circuit, cannot
> possibly respond in that time frame unless the closing velocities are
> also very slow. The DC method is inherently the faster method, but
> does require a very low ohmage connection in order to fully discharge
> the cap on a 10 microsecond contact. And we definitely DON'T have a
> low resistance circuit on this machine.
>
> I have one of this old tony's contacts about 75% made, but out of a
> brass tube instead of a steel probe and I'll take the machine out of
> it by grounding the workpiece, and wiring the tube straight to the
> probe input, using the same old cap for storage. The brass is long
> enough it can hit and be bent 1/4" without damage as it will just
> spring back.
>
> But its been a long day today. I'll figure out something thats
> hopefully repeatable.
>
Got that done, almost worked when using brass against the edge of this
alu panel, so I thought I'd get fancy, and found a steel sewing pit with
the rounded backend sewing pin and soldered it into the tip of the tube.
Disaster, I can see it sliding along on jerks as I run it back and forth
touching the edge of the alu, but the only place its making actual
contact is where the brushed finish is damaged, anyplace else needs a
10lb push with my finger to break thru the aloxide and actually make a
contact. Would probably work with most anything metallic except the alu.
To do this would need a 1kv supply, limited to 1 microamp so as not to
eat up the probe too fast and measure the voltage electrostaticly.
Last week there were at least a dozen guys selling imitation Reneshaws
for 69$/copy. But I'll be darned if I can find one on fleabay today.
> And I've got to figure what to do Monday as my lady adds another year
> to her journey on this ball of rock and water, making it to her 79nth.
>
> Something she'll appreciate without any physical effort. With copd,
> there's not much of that left. Sigh...
>
> That said, I think the DC method can be made to work with a separately
> wired probe, with both the ground on the workpiece and the hot on the
> probe wired independently from the machine. Run the probe cable as a
> shielded wire in the cable chain. At least this chain can be opened
> to add more wire, something none of the other cable chain I've bought
> can do. Nice!
>
> > I don't know if this is done commercially but the AC method should
> > in theory by MUCH more reliable. You can even use very long
> > cables if you use strong filters tuned to the frequency. Notice
> > how well the current passes from an AM radio station to your radio
> > even over a miles-long air gap.
> >
> > My guess is that a 100 KHz signal would go right through paint.
> >
> > On Sat, Feb 16, 2019 at 9:46 AM Gene Heskett <[email protected]>
>
> wrote:
> > > Greetings all;
> > >
> > > I've just found that because everything it painted before
> > > assembly, apparently including the inside of the spindle motor
> > > mount, that a ground to the bed frame can be anywhere up to 2 or
> > > more thousand ohms to almost anything else on the 6040, and
> > > apparently even includes the spindle bearings as part of the first
> > > 50 or so ohms.
> > >
> > > The net result is that using the workpiece as one contact, the the
> > > tool in the spindle as the other for the alignment function is
> > > fraught with enough variables I could break a tool against the
> > > edge of the workpiece, even damaging the workpiece, before a
> > > contact is detected. Since there isn't Z room enough for one of
> > > those $65 spindle mounted contact detectors, and it would take at
> > > least ten feet of ground braid strung thru the cable chains to
> > > arrive at a decent ground on the motor housing, which wouldn't
> > > solve the problem entirely because of the oil film in the spindle
> > > bearings, how the heck do I arrive at a reliable connection that
> > > only responds to a contact between the tool and the workpiece?
> > >
> > > A flying ground lead one could bring up and clip onto the tool
> > > would probably work, but sure resembles something Robe Goldburg
> > > would dream up as it would need to be long enough to reach the
> > > tool regardless of where it is on the table.
> > >
> > > That, or using a much higher voltage limited to a few microamps so
> > > as not to constitute a shock hazard. But basically use it to
> > > measure the air gap. I could make that work even before a physical
> > > contact was made but thats not a tasty idea in the long view
> > > either.
> > >
> > > Any other ideas out there? Hopefully something that doesn't
> > > involve changing tools to use.
> > >
> > > Thanks all.
> > >
> > > 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
>
> Cheers, Gene Heskett
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
