[SCzEngrgGroup]

Any chance that this could be modified to work on the Torus?

I have no idea.... Most likely, no. It would require the PDB, and we have no plans to make that work on a Torus. I have no clue what mechanical differences there are, and there are several points that would be rather critical that I have no way of ensuring are the same.

I've got it working pretty nicely now. There are a number of fine-tuning changes I'll make in the next round of hardware - mostly tweaks on the tool receiver. I need to swap out the one-way air cylinder on the lock for a two-way, and get rid of the return spring. The spring return isn't quite cutting it, until I increase the spring tension to the point it causes other problems. The current aluminum mounting bracket for the transfer arm drive is too flexible, leading to some minor misalignments at times. I always intended the real brackets to be steel, but used aluminum for the prototypes just to make them easier to fab. I don't yet have the ATC "talking" to the PDB, so I'm triggering the PDB manually at the correct time. But other than these few minor issues, it's looking very good now. I even think once I get the PDB connected to it, I'll be able to shave a second or two off the toolchange time.

I need to rig up some better lighting for the shop, and someone to act as camera-man, so I can get a decent video, since I have my hands full operating the PDB when it's running back-to-back toolchanges.

Regards,
Ray L.

[SCzEngrgGroup]

Below is a (really poor quality) video of the ATC doing a half-dozen back-to-back toolchanges. Please note the ultra high-tech, carbon-based, organic power drawbar (i.e. - my finger) holding the tool in the spindle, as I don't yet have the ATC communicating with the PDB.

The only problem I'm seeing at any frequency at all is an occasional random failure locking the tool into the receiver. The return spring just doesn't have quite enough "Oomph". Fitting a two-way air cylinder on the lock, in place of the one-way cylinder and return spring, will resolve this problem.

VID 20141202 163428348 - YouTube

I REALLY need to get a decent movie light, a tripod, and real video camera.....

Regards,
Ray L.

[AUSTINMACHINING]

Below is a (really poor quality) video of the ATC doing a half-dozen back-to-back toolchanges. Please note the ultra high-tech, carbon-based, organic power drawbar (i.e. - my finger) holding the tool in the spindle, as I don't yet have the ATC communicating with the PDB.

The only problem I'm seeing at any frequency at all is an occasional random failure locking the tool into the receiver. The return spring just doesn't have quite enough "Oomph". Fitting a two-way air cylinder on the lock, in place of the one-way cylinder and return spring, will resolve this problem.

VID 20141202 163428348 - YouTube

I REALLY need to get a decent movie light, a tripod, and real video camera.....

Regards,
Ray L.

Looking great Ray. Alignment with the spindle looks challenging, but looks like you have it figured out.

[SCzEngrgGroup]

Looking great Ray. Alignment with the spindle looks challenging, but looks like you have it figured out.

Spindle side is not too bad, since the transfer arm is mounted to the head. Once it's set, it's set. Getting it to reliably line up with the carousel is a bit more challenging.

Regards,
Ray L.

[SCzEngrgGroup]

Here's a much better video, with the PDB being controlled by the ATC:

SAM 0411 - YouTube

Regards,
Ray L.

[amish_rabbi]

awesome!

[BTP]

Very cool! Load it up with tools and let it rip.

[Titaniumboy]

Ray,

Looking even better without the digital (finger) help.

What are the next steps and timeline now that the prototype appears to be dialed in and working 100%? I'm guessing some or all of the following steps will need to happen?

1. Novakon approval of prototype and authorization for 3-7 beta test units (by 1/1/15)
2. Order and delivery of components for beta test (1/7/15 - 2/15/15)
3. Fabrication of beta test ATC's (1/1/15 - 2/28/15)
4. Beta Test (2/1/15 - 3/31/15)
5. Fabrication of first production units (start 4/15/15)
5. First Production units shipped (5/1/15)

Am I even close?

Again, congratulations !

[SCzEngrgGroup]

Ray,

Looking even better without the digital (finger) help.

What are the next steps and timeline now that the prototype appears to be dialed in and working 100%? I'm guessing some or all of the following steps will need to happen?

1. Novakon approval of prototype and authorization for 3-7 beta test units (by 1/1/15)
2. Order and delivery of components for beta test (1/7/15 - 2/15/15)
3. Fabrication of beta test ATC's (1/1/15 - 2/28/15)
4. Beta Test (2/1/15 - 3/31/15)
5. Fabrication of first production units (start 4/15/15)
5. First Production units shipped (5/1/15)

Am I even close?

Again, congratulations !

Well, it's working, but it's not working what I would call 100%. There are many things in the current hardware that are less than ideal. The basic design is sound, but does not represent a production design. I'd expect in the next few weeks, I'll start building a second prototype, with many small design changes. That will become the basis for a small beta build in a few months. I think production mid-next year is a realistic.

Regards,
Ray L.

[mklotz70]

Ray, since I haven't built an enclosure yet.....any chance we can get some pictures from the left side? I'm curious about where this will line up over the chip tray....inside/outside.....just so I can take it into consideration when the time comes to build the enclosure. Thanks!
Mike

[SCzEngrgGroup]

Ray, since I haven't built an enclosure yet.....any chance we can get some pictures from the left side? I'm curious about where this will line up over the chip tray....inside/outside.....just so I can take it into consideration when the time comes to build the enclosure. Thanks!
Mike

Mike,

It sits completely inside the chip trays.

Regards,
Ray L.

[SCzEngrgGroup]

Well, it's working, but it's not working what I would call 100%.

I'm happy to say I have to, at least partially, eat those words. I realized this morning I had not aligned the carousel quite as I had originally intended. I convinced myself that it was functionally the same, but it really wasn't. I re-aligned it this morning per my original intent, and it's now working even better. In fact, the only (very rare) failures I'm seeing are due to the collet sometimes pushing up into the spindle and closing before the toolholder is fully inserted. I expected this from the start, and in a few minutes will be testing a simple firmware change to the PDB that should completely eliminate the problem.

Regards,
Ray L.

[mklotz70]

Mike,

It sits completely inside the chip trays.

Regards,
Ray L.

Awesome Ray! Thanks!!

[SCzEngrgGroup]

Awesome Ray! Thanks!!

Mike,

The rear-most part of the carousel is about an inch forward of the rear wall of the side tray. So, unless you're making your enclosure out of 2" thick Plexiglas, you'll be fine. The carousel position is also adjustable over a wide range.

Regards,
Ray L.

[LeeWay]

You could fill it with water then and put fish inside. Talk about flood coolant.

[SCzEngrgGroup]

This afternoon I ditched the linear ball bearings, and made a couple of Delrin bushings to replace them. This has largely eliminated the slop in the lift, and the tilting of the transfer arm when under load. Unfortunately, it has also increased friction a bit, as I only have a 12mm reamer, which leaves the bushings just a hair too tight on the ground shafts. I'm ordering an adjustable reamer so I can enlarge them by about a thou, which should give a very nice, low-friction fit.

Regards,
Ray L.

[SCzEngrgGroup]

Things are looking really good now. I've gone through and re-tuned the servos, and made quite a few firmware updates. The operation is now smoother than ever, and I think the two-way air cylinder on the lock will get it working very near 100%. Hopefully that will show up on Monday. The second prototype should be pretty near perfect.

I did have to slow down the arm, until I can re-design the transfer arm fork, and get the lock on the fork, so it will stop tossing the tools across the room when it decelerates. I probably now hold the world record for "Most TTS Toolholders Dropped On The Floor In A 24-Hour Period". :-)

Regards,
Ray L.

[jid2]

You should make some aluminum "test holders" that are proper weight. Tool changer is looking good mounted on the machine, it looks happier than on the living room floor.

[kvom]

It might be a good idea to make the arm speed user-configurable to some degree.

[SCzEngrgGroup]

More good progress lately. I swapped out the lift motor for one with more "grunt" (the same one I use in the PDB). I will likely end up using the same motor in the arm drive as well. I also made more improvements to my home-made encoder design, using a different sensor and improved geometry. All is working really well now. Still waiting for the new air cylinder for the carousel lock, but once I have that, I think I'll be in very good shape.

The non-Tormach TTS holders are still proving problematic. They work fine in the carousel, as the tool receivers have a tapered entry, but they can be balky going into the spindle, as the collet has no taper at all. As long as alignment is perfect (most of the time) they work fine, but every once in a while the very short taper will catch on the corner of the collet, and cause a jam. At some point, I'll press plugs into the tops of the shanks, and re-cut the tapers to the proper dimension, and they should then work as well as the genuine Tormach ones.

Otherwise, I see no more significant issues at this point. With the new lift motor, the lift is stronger, and faster. The carousel and arm have been working very nicely for the last week, with only the lock causing any significant errors. The firmware has also made a lot of progress, getting a major clean-up over the last few days, and a near-total re-write of the UI code, making it more flexible, and maintainable. The only significant "missing pieces" now are reading error codes back from the PDB, which is not a big deal, and some improvements to the error recovery, which is already pretty good. With a little luck, I could be actually making chips for the first time using the ATC later this week, or early next week!

Regardsd,
Ray L.