Hardinge HNC Retrofit, The rough plan/idea. Comments & suggestions sought.

You may have paid too much.

Seriously, I'd get another CHNC that is mechanically complete. Use this one for spare parts. There's one on Ebay right now for about $1200

...

The brush servos on the original CHNC are good units.

...

I used a 7.5 3 phase and VFD on my refit. Worked well.

You're going to need a PLC to do this. Mach3 won't handle it. You'll need one output to raise turret, an output to spin turret, and five inputs - four encoder and one for turret down. i can show you logic if you get there.

This is not the best news group for your queries. Look up a web site called CNCzone.com

Karl

Consider that it might have been stripped for spare parts because it wasn't worth scraping the ways. Will you take my bet that it turns taper and faces cone? Do you want to use it for your hobby, or do you want it to be your hobby?

Jerry

...snip

I built an automated pyro-press (for making fountains) about two years ago. The process requires a minimum of one, and a maximum of four different tools to press each item. So the solution was basically the same as a turret holder for a lathe.

It locks the index with a small air cylinder that drives/extracts a self-centering pin from the indexing holes. The turret is free to move enough (when not being driven) so the pin can correct any minor mis-alignment as it drives in.

It revolves the turret with a simple ratchet/pawl affair driven by another small cylinder. It senses the tools with a built-in-place encoder that's simple to make and simple to extend for more tools/indexing positions.

The encoder consists of two (in this case for only four positions) IR reflective sensors mounted close enough to the turret body to "see" Grey-code striping painted with flat black on the bright aluminum turret body. When the correct count is seen by the sensors, the software drives the indexing pin.

The alignment of the code is non-critical, so long as the sensors are roughly centered in each bit when the turret stops after an indexing move. With the particular sensors I used, the beam is about 3mm dia. at a distance of 6mm, so the limit on my 5" diameter turret would be about 130 indexing positions, max. That would take (say, for 128 positions) seven sensors, the beams of which would also fit nicely in my 1" high turret.

It's also not necessary that the sensors be physically small enough to fit in a line from the top of the code stripes to the bottom -- no. Instead, the sensors may be staggered as necessary all around the periphery of the turret as their size requires, then the code stripes can be painted with that stagger.

I'd avoid straight binary for the encoding. Grey-code has the benefit that always only ONE bit changes per count change, so there's never any ambiguity seen by the computer as count transitions occur. If the software is right, it wouldn't matter if spurious codes were seen during a move, but Grey-code is just a 'cleaner' solution, and no harder to implement.

LLoyd

...

Lloyd,

I don't see the relevance of your post to restoring a lathe, but it's interesting nonetheless. Grey codes are ideal for quantizing a continuum, and I've long used them for that. Most uses of "Gray code" mean "reflected-binary Gray code", but there are others. Most software routines to convert reflected-binary Gray to unsigned binary work in O(n) operations, where n is the number of bits to be converted. I have a method for doing the conversion in O(log(n)). C code for this and other conversions is at

Jerry

Um... Jerry? The OP asked for suggestions on how he might control an indexable turret tool holder on the lathe he was considering to fit for CNC. I thought a discussion of a real-life simple method on another type of machine might be just a little relevent to the discussion.

And yeah... I misspelled Gray.

And BTW... NICE, tight little snippet of code. I'll use it with thanks.

LLoyd

Possibly, but I don't think so. There were two of the machines. The control on the machine that I got went bad and that is why it was stripped. Apparently it was holding

0.0001" before the control went bad. The ballscrews on the other machine have some backlash, and that is where "mine" are going. I have asked the guy for the ballscrews from the other machine after he pulls them, and he said I can have them. Not sure if they just need repacking or regrinding. New HNC ballscrews from Hardinge are about $1000 apiece, and a couple hundred for the support bearings. Probably a bargain, but way beyond my budget and __WAY__ beyond what my wife will let me spend.

Thanks.

-Ralph

It seems that you're in luck then. I lose my "bet".

Why didn't they just swap out the controllers?

I've removed backlash with a spring-loaded second nut. Maybe you can work something out. Oatmeal? :-)

I wish you the best of luck with your new hobby. :^)

Jerry

Hardinge lathes rarely suffer from worn ways. All that neat cutting oil keeps them in pretty good shape.

My 1986 machine has vitually no wear at all, and the ballscrews are sill zero backlash.

Scraping is not the thing for Hardinge dovetail beds either. Even in the worse cases the bedplate can be removed and either surface ground or exchanged for a reground one.

It will probably still do it.

The support bearings are no big deal. Did they also remove the blocks? The Z Axis nut housiing is an oilbath affair.

Wayne....

I would wait on the GREX. It does not support lathes with Mach YET. Soon, but not yet.

After the next firmware revision I may use it in my Supermax re-retrofit.

I am also using Pixie 100 boards from

to control the analog brush motor controllers. They give step/dir control to old drivers.

We'll see. I was told that it was holding those tolerances. Whether that proves to be true remains to be seen.

Good question. I don't know. These were ancient GE 550 CNC controls. Big, old, clunky. Too much work in comparison to ballscrew swap?

I have lots of oatmeal. Both, instant, 5 minute and full cook. I like grits and polenta too. I am curious how much repacking and/or regrinding ballscrews costs.

Thanks. I have re-retrofit a Bridgeport as well. Works great. Had an ancient Retrotek CNC control on it.

Next project after the Hardinge HNC is a Shizuoka AN-S CNC mill. Very rigid, big mill.

I am open to donations of machining centers and turning centers so I can avoid having to do all this retrofit work :-) Also open to donations of polenta and risotto.

Thanks!

I have put some pics online (clearly I have my work cut out for me :-)

Whole machine (covers not on):

\30.JPG

Carriage and turret (missing top plate):

\40.JPG

Spindle motor cabinet:

\41.JPG

Cutoff/parting tool slide:

\43.JPG

Back side of lathe:

\44.JPG

Back side of turret/cross slide:

\51.JPG

Solenoids?

\35.JPG

Nameplate:

\31.JPG

Would SWMBO be appeased if you grew some roses over it :-)

Mark Rand (another recent member of the "Why didn't people look after the Hardinge's that they sold to me" club) RTFM

Maybe. All that cast iron would help keep things green.

Ok. Since nobody seems to be forthcoming with documentation, I rough measured the dimensions for the ballscrews using a tape measure (will pull out calipers next time I go over to the machine):

Z ballscrew: diam