Re: Re: Re: HARDINGE HNC RETROFIT

Douglas (comments mixed in)

I'd keep the motor power supply if you can tap the transformer for lower
voltage. There are a bunch of contactors in the small cabnet near the
power switch that are used for spindle direction. Then I think that
there is a setup (dc supply and relays) that controls spindle speed but I
don't remember much about it on this model.

>    From: doug king <evelle97530@...>
> Subject: Re: Re: HARDINGE HNC RETROFIT
>
>
> Thanks so much,Ray! I'll try and understand this info.Does the GE 550
> have essential items for the retro?Or are there just some nice goodies
> in there to keep around for future projects?

I'd keep the motor power supply if you can tap the transformer for lower
voltage. There are a bunch of contactors in the small cabnet near the
power switch that are used for spindle direction. Then I think that
there is a setup (dc supply and relays) that controls spindle speed but I
don't remember much about it on this model.

Other than that saving stuff is a question of how much space you have or
can get from SWMBO.

>Am I barking up the wrong
> tree,if I just want a cnc lathe that will thread and interpolate?

These are wonderful iron and the drive motors are rock solid units that
ElectroCraft built. I use the EMC and it does not have lathe threading
yet. Give us a bit and it should be there but it is not trivial and it
will not be built around step and direction but will use servo control
stuff. Matt Shaver has a retro to do of one of these, just like the one
you describe. It's taking up space in his basement right now. My three
units are a bit newer and use a dc spindle motor.

>Will
> this be cost prohibitive,compared to a removal of the turret and a bolt
> on top plate retro,like the Rock Solid?

I have no clue about the Rock Solid retrofit. I've never seen a need to
take away from the Hardinge design. There are a couple of week spots but
if you do the retrofit you are thinking about, you will replace the worst
of them.

> How is the turret itself
> controlled?

Okay! This is not exactly an easy answer and it relates to one of the
last remaining weak spots in the machine. The turret spins in a
horizontal plane. The tool holder plate is bolted to a large diameter
air cylinder which raises it up or forces it down. When it is forced
down, a set of six or eight teeth engage teeth in a plate that is bolted
to the cross slide. The weakest point here is the plate with alignment
teeth that is bolted to the cross slide. If your lathe has been crashed
very hard, this ring may be cracked where the front tooth sits. I've
known of some who have welded or silver soldered that ring where the
crack is. Some have run for years after even though Hardinge said that
kind of repair was a complete no no.

This makes the tool holder plate very rigid with repeatable location of
tools during cutting. You can easily get 0.0002 accuracy from it.

When the tool holder is up, a Clippard rotary air solenoid turns on a
tiny air motor under the front cover plate. This spins the turret. A
six or eight position rotary switch under the cover plate immediately
behind the turret feeds back tool position. This is a set of hall
switches that are tripped by a magnet on the rotating part.

When the turret nears the commanded tool, the control energizes a
solenoid that drives a hook arm to catch the turret and stop it in the
correct position to align the teeth. There is a shock absorber under
the last top cover that takes the stopping force. Rotation speed is
regulated by a small air valve between the Clippard valve and the motor.
You should be able to set the speed with this so that you can select any
tool position in less than a second.

All of this takes a good deal of tuning but once you get all the
mechanical dominoes in correct alignment the turret takes very little
maintenance.

> Thanks,Doug PS What do you mean by  there is no room on the
> x for feedback?

There are no feedback elements in the motors. They are simply servo
grade 3/4 hp 90 volt dc motors. The one on the cross slide is down under
the cross slide and between it and the frame of the lathe on the back.
When the cross slide is forward -- toward the operator -- there is less
than a 1/4 inch clearance.

> Isn't the encoder feedback?Surely we have to have one
> there.Right?

At this point the gong would have sounded and you would have been
replaced with the next contestant. There is no encoder. Nada! Zilch!

You've got to remember that this machine was a real design pioneer. When
it and the AB controlled successor were shown for the first time they
raised a lot of eyebrows. They were faster, more capable, and more
accurate than anything before them except a Swiss screw machine and for
short runs, that required frequent setups, they were as fast. Many shops
bought a lot of the turrets and kept whole tooling setups on the shelf
because you could change a complex part setup in ten minutes!

To get that level of speed and accuracy, Hardinge designed a feedback
package that ran the axis resolver and motor tach several times the speed
of the ballscrew. This little package cost a couple thousand dollars at
the time and Hardinge and the suppliers of the devices that were a part
of it insisted that it could only be setup by them. (several creative
people proved them wrong.)

You'll find these jewels behind the covers on the ends of the ballscrews.
The Z screw is visible from the operator position, and it's cover is to
the right at the back end of the machine. The x screw is centered under
the cross slide and it's cover is right over the chip shoot opposite the
operator position.

Postion feedback was by resolver. A resolver is basically a three phase
motor that is fed an ac signal and returns two ac signals. The returned
signals are triangulated (sin, cos) from the angle of the rotor and the
input signal. There are devices that can convert from resolver to
encoder but they are not cheap. Velocity feedback was by tach, a very
small very high quality brush type dc motor/generator.

These feedback packages were driven by a small turned extension on the
ballscrews. A bellows style coupling connects the feedback package to
this extension. These bellows are fragile and run into money quick.
There is a small slot (much to small I always thought) where you can
insert an allen wrench and loosen the set screws, there are two, 90
degrees apart, on each end of the bellows. The center shaft of the
feedback package has two bearings on it and a large, very fine toothed
drive gear between them. This shaft extend through the last bearing and
has a small hex cut in the back end. Do not use this to turn the axis
unless you want to trash the bellows.

The tach and resolver have gears that mate to this large center gear from
180 degrees apart. Backlash between each set of gears is taken up by an
eccentric ring on the face of the tach or resolver. Little clamps with
allen head screws hold these in place on the outer mounting plate.
Loosen the clamps a bit and rotate the tach or resolver to move the gears
closer together or farther apart.

Now if you want to keep the current dc axis drives (They are in the 550
cabnet) you will also have to keep the tach. But if you keep the tach
you will have a b%%$h of a time finding an encoder that will fit where
the old resolver was and has a shaft size for the resolver's gear.

If you want to use more recent drive amps like the gecko 3xx series you
can replace the whole drive package with a single encoder. You will just
have to figure out how to hold it in place.

> Have you got a working one of these babies?

I still service a GE 550 and it's Hardinge for a nearby shop. There are
some legacy parts that are easier to cut with it than they are to
reprogram for other machines.

Ray