Cutting shallow, wide slot in hardwood

[Tom Gardner]

[Tom Gardner]

sliding jig on linear bearings. The operator slides the jig across the bit and the block drops in a barrel. It takes about 4-5 seconds. Not too bad, but I'm worried about repetitive motion injuries and the routers only last 4 months.

My production requirements are only 800 per day but I can only dedicate so many man-hours so I need 300 per hour. Cycle time has to include material handling and pee breaks.

My biggest concern is the possibility of repetitive motion injury so, I'd rather spend the money to automate rather than taking the chance of somebody getting hurt.

[Tom Gardner]

The blocks are 7" long and I buy them. I can't make them cheap enough anymore so I farmed it out to an Amish company, they do a great job! I only have to do 600-800/day but I can't dedicate 8 man-hours to it, only 3-4.

See:

this shows the parts and how they fit.

[Prometheus]

A dado stack or molding cutterhead would do the job quickly. Even if the stack isn't wide enough, it'd be pretty easy to bolt a sacrifical fence to the miter gauge (or make a crosscut sled) with two blocks screwed to it as stops at 6.25" from the center line on each side (assuming a 3/4" dado stack) Run the first cut with the block tight to one side, then slide the block over to the other side and finish it off. Simple, cheap, and quick- and you're not going to burn out a table saw every few months.

If the cut is rough from the dado stack, but needs to be smooth, easiest bet might be to use the above method, then clean the groove with a 1.5" sanding belt. Easy to make a jig for that, too- just get a standard industrial deburring sander and put a couple of blocks on a mount behind the belt to make sure the depth remains consistant, and a couple of stops on either side to ensure that the corners don't get knocked off when aligning the pieces. Same deal as above works here, too- if it's cheaper or easier to get 1" belts, get them and make the side stops a little oversized to allow the operator to slide the piece back and forth. Probably a good idea to make the sanding jig out of steel- it costs a little more, but will hold up a lot better to industrial-type use.

Should be able to make 800-900 pieces really quickly that way, unless your employees are masters at wasting time. Nice thing about what you're doing there is that you can just make a single jig for each step and lock it in- no excuses for boneheads to mess up a whole order of them.

As far as the repetitive motion goes, I don't think you're going to avoid that entirely unless you make a robot to do it. Which isn't a terrible idea either- a guy could do that mechanically for a couple of grand, and then set it to run and walk away. There, the router is probably the best bet, but you could get a heavy-duty motor and just mount a collet to it- no need to go out and buy a router with any bells and whistles if you are just going to mount it in a carriage.

Of course, if these are a real pain the butt for you and you'd like to outsource,

[Morris Dovey]

[ Paragraphs reordered ]

| My production requirements are only 800 per day but I can only | dedicate so many man-hours so I need 300 per hour. Cycle time has | to include material handling and pee breaks.

Ok - this essential info was missing - and is necessary to have a starting point.

| I do it now with a 3 hp router swinging a 1.5" bit. The block fits | in a sliding jig on linear bearings. The operator slides the jig | across the bit and the block drops in a barrel. It takes about 4-5 | seconds. Not too bad, but I'm worried about repetitive motion | injuries and the routers only last 4 months.

I suggest hitting E-Bay until you find a good price on a 5-10 hp industrial spindle and a VFD. If the spindle hasn't been abused, it should last a very long time. I use a 5 hp Colombo spindle and Delta (not the tool co) VFD for CNC routing and expect that they'll outlast me.

It would not be difficult to move your sliding jig with one or a pair of micro-steppers and set up the jig to clamp while in motion and release at extremes of movement. That would allow the operator to drop the block into the jig and press a pair of buttons (one for each hand), then reach for the next block while the cut is being made. You should be able to feed the block past the cutter at 3-4 in/sec (180-240 ft/min). Once the cut block has been dropped, the jig can be retrieved at a still higher speed. You'd need a PC (an old, recycled

386 would probably do) and a stepper controller to drive the motors. Using steppers makes it easy to control feed speed and acceleration/deceleration of the fixture.

If you follow the link below, you can see what the controller would look like. Your machine, of course, would only be a single-axis machine and only 1/3 as complex as the JBot shown on the web page.

| My biggest concern is the possibility of repetitive motion injury | so, I'd rather spend the money to automate rather than taking the | chance of somebody getting hurt.

Well, the operator will need to move - and the movements will be repetitive... unless you set up a system to pick blocks off a pallet or out of a box. :-)

Actually, I think it'd be fun to build something like this! E-mail me if you'd like help.

-- Morris Dovey DeSoto Solar DeSoto, Iowa USA

[Morris Dovey]

| If you follow the link below, you can see what the controller would | look like. (corrected)

-- Morris Dovey DeSoto Solar DeSoto, Iowa USA

[Roger Shoaf]

Have you asked your Amish supplier how much they would charge to supply the blocks already to your spec?

[Mark Jerde]

Gentlemen, we are in the presence of genius. Good question, Roger!

-- Mark

[Tom Gardner]

Well, kinda'! The idea was brought up but dismissed due to the fact that the slot has to be indexed to one end of the block. The tolerance is +/-

1/16" and if the block is but in the brush machine the wrong way, and the block was over or undersized, by even the 1/16", the pattern would interfere with the mounting holes. The clamps on the brush machine leave a witness mark that orients it in all the rest of the operations. However, I intend to revisit it now that you bring it up. If the Amish can mark the orientation, it'll work. Good idea, thanks!

[Tom Gardner]

You're right, he's right!

[Tom Gardner]

credit though!

[Roger Shoaf]

You might be able to facilitate this with your Amish vendor by offering to supply him with a jig of some sort. +- 1/16" is a whopping big tolerance, The craftsmanship of the Amish wood workers ought to be able to give you at least +- 1/32".

[Rick M]

"Morris Dovey" wrote

Snip of some very good advice

Hey Morris,

This job is easily handled by a servo motor system. The one I'm thinking about is a stepper motor with an integral controller. This controller can be PC controlled, or you can program it to perform the very functions you describe. With four digital input/output lines, and one analog input line, you can (a) start sequence with two switches in series. This drops a shield, forces a clamp to close, start the spindle, and after a brief (programmed) delay, starts moving. When it hits the end-limit switch, it shuts off the spindle, releases the clamp (allowing the part to drop), and after another brief (programmed) delay, returns the carriage to the home position (set by the "home" switch). The analog input port can be used to set the travel speed, and you have the last output to either serve as a clamp release, or other function you may desire.

I used 3-stack 34 size motors from Intelligent Motion Systems

for the assembly line battery tester; I was moving 45-60 pounds of fixtures on two axis and needed the torque. We used acme screw rod and nut to move, and used turned, ground, polished rod on linear bearings for low friction directional control.

Hope this helps ... with this approach, it's more mechanical than electrical. You will, of course, have to provide a power supply for the motor, and will need an RS-422/485 interface (available from IMS as well) to program and/or control the motor. Nice thing about the 422 comms, you can daisychain a pile of motors on one comms line (I had 14 steppers running at once in direct comms mode with no issues).

Regards,

Rick

[Morris Dovey]

Rick M (in 13PTg.77157$ [e-mail]) said:

| "Morris Dovey" wrote | | Snip of some very good advice || || It would not be difficult to move your sliding jig with one or a || pair of micro-steppers and set up the jig to clamp while in motion || and release at extremes of movement. That would allow the operator || to drop the block into the jig and press a pair of buttons (one || for each hand), then reach for the next block while the cut is || being made. You should be able to feed the block past the cutter || at 3-4 in/sec (180-240 ft/min). Once the cut block has been || dropped, the jig can be retrieved at a still higher speed. You'd || need a PC (an old, recycled 386 would probably do) and a stepper || controller to drive the motors. Using steppers makes it easy to || control feed speed and acceleration/deceleration of the fixture. | | Hey Morris, | | This job is easily handled by a servo motor system. The one I'm | thinking about is a stepper motor with an integral controller. This | controller can be PC controlled, or you can program it to perform | the very functions you describe. With four digital input/output | lines, and one analog input line, you can (a) start sequence with | two switches in series. This drops a shield, forces a clamp to | close, start the spindle, and after a brief (programmed) delay, | starts moving. When it hits the end-limit switch, it shuts off the | spindle, releases the clamp (allowing the part to drop), and after | another brief (programmed) delay, returns the carriage to the home | position (set by the "home" switch). The analog input port can be | used to set the travel speed, and you have the last output to | either serve as a clamp release, or other function you may desire.

Sounds good. I think I'd leave the spindle running and only power it down if the operator failed to initiate a cycle within some timeout period - say ten or fifteen seconds.

| I used 3-stack 34 size motors from Intelligent Motion Systems |

for the assembly | line battery tester; I was moving 45-60 pounds of fixtures on two | axis and needed the torque. We used acme screw rod and nut to move, | and used turned, ground, polished rod on linear bearings for low | friction directional control.

Also sounds good to me. I used essentially the same linear motion control approach with the JBot; but found a threadless lead screw and follower (DAGS: Rohlix) that I think I may like better.

| Hope this helps ... with this approach, it's more mechanical than | electrical. You will, of course, have to provide a power supply for | the motor, and will need an RS-422/485 interface (available from | IMS as well) to program and/or control the motor. Nice thing about | the 422 comms, you can daisychain a pile of motors on one comms | line (I had 14 steppers running at once in direct comms mode with | no issues).

Does help. Even nicer aspect of the 422 I/F is that it's pretty much immune to shop electrical noise over long runs.

I think perhaps you should be offering Tom a proposal. :-)

-- Morris Dovey DeSoto Solar DeSoto, Iowa USA