[Hirudin]

When I was modifying my R8 TTS collet in an attempt to reduce runout I noticed that the relief groove at the bottom of the shank of TTS tool holders was causing wear on the R8 collet. To fix that issue I cut a chamfer into the mouth of the R8 collet. This had a side-benefit of making it easier to insert non-tapered TTS tool holders. Something like this could be helpful for the ATC as well.

Link to that thread for those interested: http://www.cnczone.com/forums/novako...ce-runout.html

[SCzEngrgGroup]

When I was modifying my R8 TTS collet in an attempt to reduce runout I noticed that the relief groove at the bottom of the shank of TTS tool holders was causing wear on the R8 collet. To fix that issue I cut a chamfer into the mouth of the R8 collet. This had a side-benefit of making it easier to insert non-tapered TTS tool holders. Something like this could be helpful for the ATC as well.

Link to that thread for those interested: http://www.cnczone.com/forums/novako...ce-runout.html

Yeah, that's one of MANY things about the whole TTS design that convince me they really didn't put much thought into it. I've really grown to dislike TTS. A nice 30-taper would be vastly superior. Even a few tweaks to TTS, like rtapering the collet, and inverting the ATC groove profile (putting the beveled edge on the TOP, rather than the BOTTOM) would make it work enormously better. But, I think I am pretty much forced to make it work with TTS as it is. I have no problems with real TTS holders, just some of the knock-offs, which have almost no taper at all.

Regards,
Ray L.

[SCzEngrgGroup]

This is starting to get exciting! The last couple of days, I've made major progress in beating the last few tiny gremlins out of the system. I've made a number of small, but significant, design "tweaks" - in particular, a re-design of the little piece that actually locks the toolholder into the tool receiver, and a new transfer arm fork - along with quite a few more firmware enhancements, and I now believe the whole mechanism to be operating at virtually 100% reliability. Rebuilt with all the final "production" parts (which, sadly, will mean making a complete, new set of parts), and all the final sensing capabilities (some parts, like the lock, are still running "open-loop"), I am very confident the whole mechanism, in production form, will be extremely reliable and robust!

With the new transfer arm fork, it appears even the dodgy no-name TTS holders with their minimal tapers are no longer a problem, and I can now run the transfer arm at full-speed, without need for an explicit lock mechanism on the fork - the spring force alone in the fork provides more than enough "grip" to ensure that even the heaviest tools cannot possibly come out "in-transit". The production fork will be better still.

I hope within the next few weeks to bring up the Ethernet interface that will be used to communicate with the PC. Ethernet will provide a very simple, flexible, reliable interface. The Mach3 plug-in will be able to communicate easily with the ATC (and through the ATC to the PDB), and I will also be able to have stand-alone Windows configuration, control, diagnostic/test and data-logging applications that can run along-side Mach3. This will aid in customer support as well, as it will even make it possible for me to monitor and control the ATC remotely, if necessary, to debug any really difficult problems. It will also enable controlling the PDB from the PC, as well as sending data, such as spindle true speed and power, back to the PC for on-screen display.

Regards,
Ray L.

[BTP]

On most VMC's with side mount tool changers that use a switch arm similar to what you are doing, you can "flag" heavy tools so that they don't get launched from the arm. It just slows the tool change arm speed to about 50%.

[SCzEngrgGroup]

On most VMC's with side mount tool changers that use a switch arm similar to what you are doing, you can "flag" heavy tools so that they don't get launched from the arm. It just slows the tool change arm speed to about 50%.

That's an interesting approach, though I don't think it will be necessary here. If it is, it's easy enough to implement in that way. Very heavy tools are really not TTS-friendly to begin with, and I believe the arm will no safely handle even my 2.5" face-mill.

Regards,
Ray L.

[Titaniumboy]

Ray,

Congratulations on the continued progress. It is looking good.
I assume you could use your ATC with a generic power drawbar (pneumatic and bellevilled), it is just that there wouldn't be any feedback from the power drawbar. What would your ATC do in the event that that the power drawbar failed to either release or grab the tool properly during the handoff? Would the ATC notice the extra force - or lack thereof - during a botched tool load and e-stop the mill?

[SCzEngrgGroup]

Ray,

Congratulations on the continued progress. It is looking good.
I assume you could use your ATC with a generic power drawbar (pneumatic and bellevilled), it is just that there wouldn't be any feedback from the power drawbar. What would your ATC do in the event that that the power drawbar failed to either release or grab the tool properly during the handoff? Would the ATC notice the extra force - or lack thereof - during a botched tool load and e-stop the mill?

No, this ATC will work only with my PDB - It expects to get a confirmation back from the PDB that the commanded operation was successful. It would be possible to add an "adaptor" to make another PDB "look" like mine.

Regards,
Ray L.

[SCzEngrgGroup]

No work getting done on the machine today - we're in the middle of a big storm since last night. Power has been out since before dawn, and cable/internet were out most of the morning. We're on generator right now, but it doesn't supply my shop, so I'm house-bound.

I'm using the time to complete the production design of the transfer arm drive, which is now very nearly done. I've beefed it up quite a bit, changing the 12mm linear rails to 3/4", and extending the linear bearings to 3" length, to minimize "tilting" of the transfer arm under heavy load. I've integrated nice, clean, simple limit and home switches for the lift and arm. I also extended the lift travel by 2", so the "home" position will now be back against the column, and 2" up, putting the bottom of the transfer arm just about even with the bottom of the lowest point of the head casting, where it should present essentially no risk of interference when the machine is in operation. Nice thing is, the new design is, overall, simpler, and easier to fabricate.

Last night I got the new double-acting air cylinder mounted on the prototype, but didn't have time to test it out. As soon as the power comes back on, I'll get that out of the way. I also received the Turcite rod I plan to use for the lift bearings and leadscrew "nut" on the new transfer arm drive. With luck, I should be able to get the new drive built before the end of the year. I don't expect it to take more than probably two days make all the parts, though I'll need to order in some raw materials and and will need to get a few parts before I can assemble it.

Regards,
Ray L.

[SCzEngrgGroup]

Attachment 260668

The picture pretty much sums up how I feel today - With the last few minor tweaks, I believe the ATC is now 100% functional and, for all practical purposes, 100% reliable. I need to re-make the tool receiver lock arms and cams one last time, before I can test that assertion, but I am very confident. The carousel will do random seeks until the cows come how without error. I have the new lock arm and cam on one tool receiver, and I can swap tools in and out of the spindle to the one receiver until the cows come home without error. The whole system is now operating very smoothly and consistently. As soon as I get all the tool receivers retro-fitted (hopefully just a few days), and figure out how to make a temporary housing to protect the motors and encoders, I should be able to run the ATC for real, making chips, and run a true reliability. i.e. - 8 hours of constant operation, which would be about 1400 toolchanges. My only (minor) concern at this point is how the carousel damper will be affected by temperature, but I am sure that problem if it even exists, is easily solvable.

Regards,
Ray L.

[kvom]

What happens AFTER those cows come home?

[tmarks11]

What happens AFTER those cows come home?

I have seen Ray's house... he doesn't have much pasture land up in those hills. I think the cows would be very unhappy.

[SCzEngrgGroup]

Today I managed to make all the new tool receiver lock cams, despite some problems with Mach3 (I made the parts on the Pulsar, since the Pro is still in pieces....). Mach3 decided to, all by itself, switch the IJ Mode to absolute, which produced very strange results. Since these parts are very small (about 3/4" round) ather than the usual "crop circles", it just did all kinds of bizarre things, that were not obviously caused by IJ mode.

I hope tomorrow to get all the levers made, and have a 100% functional ATC!

I also received something I found rather funny. I ordered an adjustable reamer from Shars. Here's the box it came in:

Attachment 260892

Regards,
Ray L.

[daniellyall]

good old changlish for ya

[ram48]

Beat me to it..................................I wonder I wonder if that is the mandarin or Canton dialect

[SCzEngrgGroup]

good old changlish for ya

Yeah, but the only excuse for that these days is major laziness! With the Internet, you can easily get anything translated better than that!

Regards,
Ray L.

[LeeWay]

That is assuming you know how to read and google. Not everyone does. Check the Mach version on the Pulsar. The one that I found that works right is 3.043.066.

[G59]

They use Baidu as their Google over there. Maybe that has something to do with poor translations?

[daniellyall]

or its a hint

[SCzEngrgGroup]

That is assuming you know how to read and google. Not everyone does. Check the Mach version on the Pulsar. The one that I found that works right is 3.043.066.

I am running 3.043.066, which, BTW, many people on the ArtSoft forum recommend NOT using, as it has many problems.

Another problem I'm seeing, that I don't recall seeing before, is that each time it starts a new operation (a new pocket, contour, etc.), it pauses for as much as 10 seconds, apparently "chewing" on the G-code. I can't see a thing in the G-code that would explain this either, yet it does it every time, even before drilling and reaming operations!

Regards,
Ray L.

[G59]

Only a few more days Ray, and you'll have the Dynomotion going again. For now, just do like everyone else who's running Mach3 and put up with it. That version of Mach3 is not looked upon as too stable from what their forums say.