Dual NC controls within a single machine

[Jim Shaper]

I want to add live tooling to a machine that wasn't configured to ever have it, but I'm not excited about ripping the Fanuc 3Td control off it right now because it does offer a level of functionality over computer based controls (IMO). The only real issue is that my spindle is not indexable via the Fanuc interface (that I'm aware of). I can assign different functions to the various connection ports for bar feeder, and what not, so working the live tooling via standard programmed code wouldn't be the issue.

So right now I'm wondering how plausible it would be to drive the servo drives with two different sets of commands where I could essentially piggyback a mach3 or what have you ontop of the Fanuc control. Preventing the signals from traveling upstream to conflicting sources should be easy enough with diodes, but how does the NC respond to an axis which hasn't been ordered to move - moving? The only real interaction would be to locate the spindle via it's own encoder (or one added to the jack shaft which drives it) from outside the 3T control. The rest could be done within the 3T to do the moves needed while the tooling was activated and the mechanical spindle brake was engaged.

The other option is to upgrade the control to a version which handled live tooling. I'm aware of another machine of very similar vintage which had live tooling on the same 3T control. I haven't seen a single 3T control intact for sale, and I'm not sure what would happen (parameters wise) if I started swapping boards. I also have no clue what boards would need swapping, or where to find that information.

I'd like to keep this as uncomplicated as possible, and hopefully that will also allow it to remain a cheap (or at least reasonable) expenditure. I'll devise my own live tooling.

Am I crazy?

[PixMan]

Yes, you are crazy.

What functionality are you referring to that a 3T has over more-modern controls? Surely it can't do all the canned cycles and high-end G-codes that an 18Ti or similar control could handle. Is there any such thing as a "3TT" control? That would essentially be a "dual control", as it was on a Star RNC16B that I used to run with it's "0TT" control.

[Jim Shaper]

I'm referring to the control panel having more functionality than having to mouse through a software application. I also don't like how mach3 is open loop between the hardware and control. The cncbrain intrigues me, but I would still prefer a console with real buttons (and I'm not a computer programmer, so writing code to make that happen is out of my league).

Canned cycles don't impress me, since it's just code in the end that I can program and "can" myself (the machine will handle DNC via rs232). On top of that, my parts are dead simple and don't need any funky code wizardry.

I'm not aware of what you're calling a dual control, as anything that's packaged with the functionality of running live tooling and what probably boils down to the spindle accuracy of rigid tapping contains everything I'm looking for in a controller. The issue I'm facing is having two controllers running the same drives at the same time - piggybacked. If I added a control with the resolution needed to locate the spindle radially, I could then run the 3T in X and Z to accomplish my machining functions.

Mentioning the 18Ti is relevant, but buying one would cost more than my Nakamura did. I've got a possible line on some newer fanuc controls (in the 5K range - which is what I paid after tax and delivery for the machine), but they're stuck in limbo right now so I can't plan on that purchase.

[Jim Shaper]

I should clarify something: Rigid tapping would have the resolution needed, but the control still needs to be able to index the spindle in a degrees format.

Being able to drill a pattern in a disk's face is the end goal here. I could do it in two steps, but I don't yet have a cnc mill. So since it shouldn't be all that difficult to accomplish within my motorized rotary 4th ( ) I'd like to just do it in a single chucking.

I'm also not opposed to spending a bit of cash as long as it gets me where I want to be in the end. The machine is very well maintained and worth bringing the control side up into a current or at least more modern configuration.

[Al_The_Man]

What you could do is the way many machines achieve C axis control is to add a C axis servo that drives the spindle direct for 360deg orientation.
In normal spindle mode, the axis is turned off.
In your case, it could be individually controlled and separate from the 3T, the only thing is you have to remember to engage it only when the spindle is stopped and disenge when the C axis is done with.
If the Motor is tightly couple by timing belt etc, then the motor encoder could be used for orientation.
Al.

[Jim Shaper]

The spindle encoder is coupled by timing belt.

The only real problem I have with the current control is that there's no provision for indexing the spindle. I'm also not sure what adding that provision to the 3T entails.

[Al_The_Man]

Originally Posted by Jim Shaper The only real problem I have with the current control is that there's no provision for indexing the spindle.

This is why I alluded to a stand-alone C axis add-on, this would operate separate from the 3T, similar to a stand alone rotary table used on a mill table would be.

Originally Posted by Jim Shaper I'm also not sure what adding that provision to the 3T entails.

The only problem of not being fully integrated, it demands some operator intervention for indexing during normal part programming.
I gathered that this is what you were looking for?

Al.

[Jim Shaper]

Ideally, it would be programmable much like a rotab. I'd be content for the time being with having it require some operator participation.

It's one of those things where it would almost make sense to have it be controlled by a whole different controller, but then you end up with the same problem of needing to index the turret and that's probably the biggest hold-up for me just buying the cncbrain and trying it.

Being able to switch back and fourth between controls would be acceptable, but how does the 3T respond to the drives being run by something else?

This is all to facilitate accurate placement of 16+ holes per part - so it's not just a matter of me being lazy (but that's also part of it lol).

[Al_The_Man]

Right now the spindle has an encoder for CSF and Tapping etc, these functions usually gear the Z or X axis off the spindle encoder so the spindle does not have to have precise positioning or rpm as the other axis follow the encoder.
With C axis control, the spindle has to act like a servo, and many are not made or set up for this, hence the separate C axis servo. The 3T spindle does not really care at this point if the chuck is being turned as it is not being activated.
The controller for this can be one of your choice, I little while ago I fitted one such similar application and I used a Galil stand alone 1500 unit to drive a servo motor, I used a Maple systems display for operator input, but it depends on how comfortable you are with programming such units, something like Mach is already setup, if you preffered to go that route.
The 3T is an older control and the options for integrating any other options are limited.
Al.

[fizzissist]

We had an old Wells Index (760?) with a Bandit control that went south about 10 years ago. Bought parts from AHHA and did an extensive retrofit, built a computer to control it, and ended up with a pretty nice little system with a 4th axis. I built the 4th axis stuff to be interchangeable with a big rotary table and a 5C horizontal dividing head...which could be set at varying angles..

The idea here is that the AHHA control, and I'm sure many others too, have the ability to trigger from external controls via M-code or 5V signals. If the output of the Fanuc 3T has the ability to send a programmable or user definable signal out, then an inexpensive 3rd party control could use that as input.

What I'm thinking is that your G-code has the triggering M-code command embedded just after the execution of a spindle lock (presuming a C axis?) or whatever, and the Fanuc waits for a command back from the 3rd party control after it's completed it's task.

I'm only familiar with the 6TB control, and that was from years ago....but I'd think that there's some interfacing that could be done to do what you want. The Fanuc was designed to operate many different outboard devices....maybe you could even use a plain ole M7 to trip a relay?

This is essentially what I had wired up with ours to run a plasma torch set up on the mill....which we never actually did, but it was all wired up.

....just a thought

[Jim Shaper]

Al, we seem to be going around in circles here. I understand what you're saying about using a servo to locate the spindle, but I fail to see why this needs to be done with something other than the fanuc 10 servo that's already there. Once the encoder is tapped for the second control, there's no reason to need another motor to rotate the spindle.

Using relays to switch the control signals in or out from the two controls sounds like a good way to trigger a feedback fault and get the NC to shut down. But, what happens if you drive the axis with signals that originate outside of it? Will it fault then?

The only reason this can't be achieved from within the 3T is that it lacks the programing option to do so.

As you can tell, I don't want to add another motor to the spindle drive, because if I went that route, I'd be better served making a stand alone rotary 4th for my mill.

[Al_The_Man]

I must admit you have lost me.
What is the Fanuc 10 servo's function/axis right now?
You either have C axis function now or you do not.
If you do not, and C axis function is what your are looking as per post #4 then the present spindle set up will not cut it.
Al.