[ihavenofish]

The idea I had was like this. I like Hoss' carousel idea, but what I wanted was a manipulator between the spindle and it to do the moving. Basically, it pivots between the axis of the spindle and the axis of the slot in the carousel to move tools. I thought of a pneumatic cylinder on the end of it to move the tool up and down rather than use Z travel any more than necessary. The doodle attached shows the geometry between the carousel and the manipulator in an unworkable state, but that would be easy enough to re-arrange.

Jim

this is basically what im going for. but its hard to position the carousel where you can get a 360 degree rotation on the gripper. it needs to be 360 to be ably to grab the tool on the changer and spindle - whether its double grip or not.

[ihavenofish]

What makes the mounting difficult? As for eating a few inches of Z travel, I'm sure you can simply stretch the drawing you have of the existing column to make up the difference. I simply am not a fan of riser blocks as they are yet another joint and the only thing keeping them in place is a properly tensioned bolt...which introduces two new planes in shear.

you could over extend the rails for the travel needed i suppose. i dont want to make the column any taller due to weight.

mounting is hard because the assembly has to sit in front of the head, which means you need a big frame that runs back to the column clearing the box ways. its a large structure even outside the pivoting mechanics complexity.

it of course has many many advantages too on the electical and programming side and its fast.

hmmm, how bout a modification to the fanuc design.

static frame that holds fixed carousel on a 3" pneumatic driven slide. individual pivoting tool grippers at each of 8 stations. single pivot driver to flip down the tool gripper which is in front of it actuated by the carousel movement. drawbar release will be mechanical driven by the carousel motion. tool grippers will be spring loaded to compensate for drawbar release motion differences.

so the scenario is:

- raise head to top of travel
- rotate changer to empty slot
- actuate carousel
- rotate carousel to new tool
- retract carousel

done.

2 commands from the software runs the whole thing. if you wave a pneumatic drawbar release, one could use a mechanically actuated valve instead of having the carousel push on the drawbar.

8 easily acessible tools at the front of the machine - changale while the machine is running. no travel lost, no tools in the work area.

only "complex" logistical requirement is a ~20" long frame to mount the thing up.

the mechanics of the carousel pivots are complicated, but easily made on the cnc mill itself by the owner.

[cnc2]

double gripper and 8 tool carousel.

control aspects:

- extending carousel into position
- rotating the carousel
- dropping the gripper into position
- swinging gripper
- releasing drawbars on spindle and carousel
- dropping gripper to retract tools
- swinging gripper 180 degrees
- retracting gripper to insert tools
- swinging gripper to neutral position
- retracting gripper out of the way
- retracting carousel out of the work area

seems overly complicated to me from a control and programming standpoint. mechanically its not bad though.

sticking points i see are gettng the gripper to sit at 3 distinct heights. im not sure if thats easily doable with air. maybe a mini ball screw instead.

could leave the carousel fixed in position to remove some complexity.

hmmm

Your caroussel is cool & can be simplified...first it'll be at a fixed heigth say at the middle of the Z travel(at the top would be better for those machining high parts) then, you'll have to design a one piston pneumatic arm that makes the caroussel travel on a half circle path from the side of the column to under the mill head and back(a bit like the grim reaper lol) once under the mill head the caroussel will have to turn using one axis driven by a cheap stepper with worm gear for increased holding torque.

This design is simpler but requires more space behind the machine...like a second mill head in distance behind the column...hey, why not make it hand adjustable in height...you could choose between two or three preset positions before starting a job.

I don't have the skills for SolidWroks but i'll try to make a skech(with paint) on the top view you posted it might make it clearer.

cnc2.

[ihavenofish]

Your caroussel is cool & can be simplified...first it'll be at a fixed heigth say at the middle of the Z travel then, you'll have to design a one piston pneumatic arm that makes the caroussel travel on a half circle path from the side of the column to under the mill head and back(a bit like the grim ripper lol) once under the mill head the caroussel will have to turn using one axis driven by a cheap stepper with worm gear for increased holding torque.

I don't have the skills for SolidWroks but i'll try to make a skech(with paint) on the top view you posted it might make it clearer.

cnc2.

what you are talking about wont work. you cant have the changer move in half height into the work area. youd hit your work.

[cnc2]

what you are talking about wont work. you cant have the changer move in half height into the work area. youd hit your work.

I've said adjustable height depending on the job !

Well, it might be simpler & can be made using two bent tubes sliding on each other and adjusted & secured by hand for two or three preset possible height positions...depending on the height of the job...to reduce Z travel & tool change time.

I've changed the way height is adjusted by installing the rotation axis at the top of the column & sliding the tubes on each other.

I've "canibalised" your SW views with MS-Paint but it should be understandable.

cnc2.

[cnc2]

sorry i edited after you posted...i didn't see your post
cnc2

[ihavenofish]

fanuc mod isnt ideal. im back to a drum on the side with a drum mounted double gripper.

trying to coordinate some of the motions mechanically so we dont need control over every aspect.


right now its:

- rotate drum
- drop tool in drum via cam
- swing gripper
- release drawbars
- drop gripper
- swing gripper
- retracts gripper
- tighten drawbar
- reset gripper to neutral

the drum is mounted on one side of the column, out at the side of the head. small tools (under 2") never in the work envelope. tool point to the side, making instalation easy even while the machine is running. pull stud retainers in the drum will be a hand pressure snap fit, so no powered drawbar is needed.

so we need to control:

power drawbar in the spindle - air.
rotating drum - small stepper.
rotating gripper - small stepper.
retracting gripper - air.

so under mach, is this considered 2 axes and 2 switches?

[ihavenofish]

i thought of a technical programming issue

with a dual gripper, the control software would have to track where the tools have moved, because with each change they are put back in a different spot.

so to remedy the issue, one would take the tool out of the spindle, move to the next slot, take that tool and put it back in the spindle.

slows the process down, but thats fine. you can use a single gripper though, which allows the tool drum to be moved to a more convenient location nearer the column. you need just enough clearance on the one side to let go of the tool. what im trying to figure out is a way to control the back and forth of the gripper with pneumatics between only 2 fixed positions... tricky one.

[cnc2]

Man, in my design you'd need one stepper(one axis) and two pneumatic actuators (two switches) one for the arm & one for unlocking the tool from spindle & the mechanichal part is simpler than yours
+ In my design ther's no gripper, the whole carroussel moves & the mill head has to move up & down to change the tool, a bit like the one of Hossmachine but with a robotic carroussel.


For the problem of tool arrangement in the carroussel i've read on the zone that some commercial machines used RFID to recognize the tool wherever position it is placed in the caroussel...i don't know if there's an open source solution for this kind of RFID readers.

cnc2.

[ihavenofish]

mounting arm for the drum and gripper. its an aluminium block bolted to the column, and 2 1/4" steel plates. easily made on the machine and cheap.

[Seb]

i thought of a technical programming issue

with a dual gripper, the control software would have to track where the tools have moved, because with each change they are put back in a different spot.

One of the lead hackers on the EMC2 project (an open-source machine controller for Linux), is busy adding support for tool changers where the tools move from spot to spot. Here is the machine that spurred the development, the first picture shows the tool changer clearly.

You know you want to drive an open-source machine with open-source control software

[ihavenofish]

One of the lead hackers on the EMC2 project (an open-source machine controller for Linux), is busy adding support for tool changers where the tools move from spot to spot. Here is the machine that spurred the development, the first picture shows the tool changer clearly.

You know you want to drive an open-source machine with open-source control software

sweet, thanks. thats basically exactly the changer ive arrived at.. except with only 8 tools.

[ihavenofish]

coming along. the changer could be moved up and down on the column the sit at the right height for whatever spindle one uses.

also, the same chager with different drum tubes could be used for bt/iso30, iso25, iso 20 and even hsk.

the rotation will be done by a belt drive to a short nema17 motor. on the top side of the mounting arm. im still working out a "dumb" gripper.

the drum holds 8 tools of less than 2.5" diametre however you coul dmount a 3" plus face mill if you left the 2 neighbour slots empty.

[ihavenofish]

[jalessi]

Very impressive so far, keep up the awesome work.


Jeff...

[ihavenofish]

there, thats final.

it will be a single gripper stepper controlled, pneumatic retraction. i did this to get the drum as far back and close to the column as possible.

now i simply need to design the details of the mechincal and electrical parts.

i wonder if the control oc can fit in the back cabinet, and a control panel with monitor on an arm can be mounted. specifically one that can be cut out on the mill - so 10" in one direction.

maybe someone else wants to design that?

[TOTALLYRC]

Hi Fish,
I wouldn't be overly concerned about using an axis to control the tool changer.

I am setting up a click plc to control my toolchanger and running it over serial modbus. Not as hard as it sounds.

For something as slow as the carousel, the plc could probably be setup to run the stepper motor just by pulsing the step input of the drive and with some creative programing you can get it to check it position via switches. 3 position switches gets you 8 tool locations and we can say that all three switches on is also home. When you first start up,if need be, it can be setup to find home so it knows where it is.
$69 + $14 for the cable gets you the basic plc with some I/O and it is expanable from there. The software is free.

On my tool changer the carousel in air driven and uses a geneva mechanism, which is kind of choppy. I am looking to go to either a stepper or servo but I don't want to use an axis for this so I am looking innto driving the stepper drive directly from the plc.

Your machine looks like it will be a great addition to the shop, depending of course what the finances look at the time it is available.

Since my oldest is looking into some machinist training, I think I will need a second mill and a bigger workshop.

Keep up the good work.

Mike

[ihavenofish]

Hi Fish,
I wouldn't be overly concerned about using an axis to control the tool changer.

I am setting up a click plc to control my toolchanger and running it over serial modbus. Not as hard as it sounds.

For something as slow as the carousel, the plc could probably be setup to run the stepper motor just by pulsing the step input of the drive and with some creative programing you can get it to check it position via switches. 3 position switches gets you 8 tool locations and we can say that all three switches on is also home. When you first start up,if need be, it can be setup to find home so it knows where it is.
$69 + $14 for the cable gets you the basic plc with some I/O and it is expanable from there. The software is free.

On my tool changer the carousel in air driven and uses a geneva mechanism, which is kind of choppy. I am looking to go to either a stepper or servo but I don't want to use an axis for this so I am looking innto driving the stepper drive directly from the plc.

Your machine looks like it will be a great addition to the shop, depending of course what the finances look at the time it is available.

Since my oldest is looking into some machinist training, I think I will need a second mill and a bigger workshop.

Keep up the good work.

Mike

i need 2 motors with defined positions. i was fairly certain there could be a way to do it without an actual axis in mach. i like the idea of the separate pulse driver simply controled by switches.

you could probably put one switch on the carousel and 8 "triggers". the basically stream pulses and stop when you have hit the right amount of triggers. use a separate home switch for startup index.

the gripper could run the same way possibly, but it has to switch directions so its slightly more complex.

got any info on a board that does this? i guess mach has to say "go" and the board returns the switch triggers and mach says stop when it feels like it. correct?

[ihavenofish]

width.

got to thinking about enclusures.

this machine has a 10" wide saddle, and 20" travel. 10+20+20 = 50" working width, plus at least 7" for the drive motor. call it 60" or 5 feet

thats how wide your cabinet/enclosure space would need to be on the inside. not terrible, bur consider this:

invert the table. now you have a 41" wide cradle, for 20" travel, and the motor can be inverted under the cradle, so no extra room needed.

thats a LOT of saved space. but we now need 41" box ways - 5" longer than most ground stock will come. we ca get by easily with a 30" cradle casting, so thats not an issue really. we can widen the saddle too by 2-4", giving more support to the table. theres more friction on the table ways now due to length - about 2x, but thats still fairly insignificant.

what say the masses? saved shop space vs hard to find longer way stock?

[TOTALLYRC]

i need 2 motors with defined positions. i was fairly certain there could be a way to do it without an actual axis in mach. i like the idea of the separate pulse driver simply controled by switches.

you could probably put one switch on the carousel and 8 "triggers". the basically stream pulses and stop when you have hit the right amount of triggers. use a separate home switch for startup index.

the gripper could run the same way possibly, but it has to switch directions so its slightly more complex.

got any info on a board that does this? i guess mach has to say "go" and the board returns the switch triggers and mach says stop when it feels like it. correct?

I am doing something along these lines for my current 24 position tool changer so let me get back to you when I have done some more work.

There are several tool change macros for rotory tool changers and they can be modified to do what needs to be done.