Getting the ShopBot up and running

[mikejohn]

I've moved away from this (http://www.talkshopbot.com/forum/cgi-bin/discus/show.cgi?tpc=7&post=23173#POST23173) thread, as the table and ShopBot are levelled (for now!).
I did say in that earlier thread I would remark on the Manual, but it has since occured to me that, as my 'Bot is a Hybrid Alpha/PRT96, there is nothing useful in pointing out the problems, as they will apply to no one else.
One thing I will say. If you have problems, read the manual, most of it is in there somewhere. Like all modern technology, it's finding that somewhere which is difficult.
One thing I did experience problems with was following the Assembly Manual section through to the section on 'Taking her out for a trial run...' without reading the rest of the manual, or rather the Quick start Tutorials, and parts of Users Guide, The Big Picture.
Connecting was a problem, solved so far only by using me 'biggy' Windows XP 2gyg + computer.
My smaller, older Pentium II will not connect....yet!
So the first command I tried, as per the manual, was (MZ)-1 no problem.
My second test (MY)3 or (MY)75 in my millimeter case sent the Y carriage crashing into its stops!
And before you ask yes I was in mm not inches.
Then I couldnt persuade the Shopbot Y back beyond the middle of the carriage.
Using (K) and the arrows to move around, it sounded like a sewing machine, any move commands sounded like an automatic car driver using a shift stick for the first time. Tried the settings on page 5 of calibrating the shopbot PRT, it did not like that! Tried the settings on page 117 of the user guide. Returned to the same problems. (Z2) 0,0 then use the arrows to move x away from zero, the measured distance of 1910mm instead of the 300mm indicated, (JX) 0 and it only came back halfway.
At that point I stopped, brought the computer home, and found something had eaten the BIOS and network card (co-incidental they said in the repair shop, nothing to do with operating the ShopBot software).
Back again today I found I could reset (UR)to all the 'original' settings.
Did this and cured the problems.
Then, still without a router bracket, I taped on a pencil and drew a circle

Now Brady said something in the other thread to the effect 'stop fussing and start cutting'. It aint that easy when doing it the first time. I understand experienced ShopBotters getting frustrated and bored by newbies showing there ignorance. Maybe there should be a thread for newby problems alone, then if your not inclined to get involved in basic stuff you can ignore those posts. As for me, I will always be sympathetic to the tenderfoots.
Now to solve the 'old' computer connection problem
............Mike

[bleeth]

Mike: Congratulations on your first project!! (a circle). We will let that qualify as "starting cutting". Ten minutes after I had my router buzzing I ran the shop-bot logo sample and then did a quick name of my company. Then I slammed together a couple of vector drawings from the vector-art disc and cut them with a v-bit. At this point my table was not surfaced, I hadn't "trammelled" my spindle, and had no dust system working either. But by gosh I just HAD to see how my new toy ran. It was quite a while till I had every thing as square and level and flat as I wanted but that little intro cutting gave such a sense of relief that I'm still very glad that I skipped ahead and cut something. The thing is that there are some of the items that you will be tweaking once you get operating for quite a while and I think Brady's and my point was akin to making sure you are seeing the forest as a whole while caring for the trees. Next week I expect to see a picture of your machine cutting a horse!! LOL

Dave

[mikejohn]

Dave
Thanks for the kind remarks.
Don't misunderstand me. I appreciate everybody who bothers to reply to any questions I ask. Mostly I know they are pretty basic. I hate for anyone to be upset by my naive perspective, but I do understand.
regards
Mike

[gerald_d]

Mike, I taught my kids to drive on an old airstrip where there were no obstacles for a long way.


I reckon that the manual should warn you to have the gantries/cars/slides far away from the stops before you are encouraged to take the first baby steps. There were some SB's that took off in the wrong direction, and I can imagine a metric-minded person saying MX,600 to an inch-minded SB......

Wow, you drew a circle! Well, did the ends join up? Was it the right diameter? Was it round (not ellitical)?

[mikejohn]

It was round, the ends joined up and it was the right diameter!
I didn't know whether to laugh or cry

How long does it take before you stop worrying as the 'Bot heads back to 0,0, wondering if it will stop?
I will sleep like a baby tonight (and it's a long time since I was one of those!)
..............Mike

[jsfrost]

Mike,
Congratulations again.
Yours is a better start than mine. Tried to aircut cut my circle under the spoilboard, and hit the stop just as the router nose disappeared.
Onsrud 1 Jim Nill in 20 seconds.

Confidance that it will stop comes quickly, but it's a long while before you can stop admiring the tool at work and do something usefull.

[gerald_d]

This 0,0 position has assumed a legendary magical status as being the bottom left home corner for x,y. But, in reality, we very seldom send the SB home. 0,0 can be defined anywhere on the table and it is rare that it is at the limit of travel of the x,y axes.

Remember that there is a third axis, the z-axis, and nobody talks of sending that "home", or to near the limit of its travel. So why do we only want to send 2 out 3 axes "home"?

Anyway, it only takes a few hours to get the confidence that the SB will go exactly where you tell it to. Irrespective of whether there is a clamp in the way.

[Alan Simpson (Unregistered Guest)]

Gerald has it perfect. Discovering that zero can be anywhere on the table and anywhere on the part gives amazing freedom. We routinely have multiple unrelated projects on our table. Moving zero will become routine.

Alan

[mikejohn]

Dave
I think I know what you mean about admiring the machine. Roxana claims I haven't loooked at her like that in a long time
.

Gerald
Do all of your cutting files start close to 0,0? Do you move your x,y to the desired postion and zero for each cutting set up?

Alan
Similar question to the one I posed to Gerald, do you 'zero' each position of the unrelated projects?

..............Mike

[gerald_d]

In our case, each and every job is re-zeroed. Therefore, your question, "Do all of your cutting files start close to 0,0?" is ambiguous to me. Each job, or file, has a 0,0 but this has no relation to the absolute "home" of the whole table.

[mikejohn]

That is what I meant, but put badly. Your 0,0 is related to the material, not to the table.
...........Mike

[gerald_d]

And that's why we've never felt the need for proximity switches on the table - they will probably paint us into a corner.


95% of the time we roughly position the center of the bit on a known corner of the material using the keyboard cursors, type Z2 , set the z=0 on the top surface of the material with the plate and then run the file.

[bleeth]

Mike: I cut my spoilboard along x any y 0 so I have a reliable "square" to register from. Although I do tend to use table home, I also cut a lot of full size sheet goods. I find it easier to just square my stock to my spoilboard and not worry about re-zeroing x & y. Even when cutting shapes out of lumber I tend to use the same system. It's a lot quicker and easier than aligning a plank in the center of the table and reaching that far.

A point that Gene and Gerald made previously about gear wear gave me a bit of a pause for thought and I've been thinking about a general rotation system of locations for carving smaller parts in different quadrants.

Dave

[richards]

Do yourself a favor and outline two corners of the material that you're cutting on the spoil board for reference (in case the material 'slips' a little during machining). Then, you can realign the material to the 'marks' and try again.

Also, drill a reference hole at a known spot in the waste portion of the material being cut. Then, if the shopbot looses steps, you can easily reposition the carriages back to a known starting point.

[ron brown]

I'd guess 90% or more of my files end with "J2,0,0". Many times I will register as Gerald does, write down the coordiantes so I can go back to them (thanks to the larger buffer and newer software these numbers can be found if I forget) if I am running multiples or have pieced together a project with multiple bits and such.

Ron

[beacon14]

I also have the edges of my spoil board machined for accurate reference, and I have the prox switches so I always know what the relationship of the bit is to the spoilboard. That doesn't mean I have to position every workpiece at 0,0. I'll often screw a small piece down on an empty spot on the table so as not to have to remove something else that I'll need again soon. Then I can measure to the edges of the spoilboard, use the J2 command with the quick measurements for the X and Y parameters. It's faster for me than having to manually move the bit to the corner of the workpiece and re-zero.
But I do use the lower left corner more often than not just because it's the most convenient corner to reach.

[Alan Simpson (Unregistered Guest)]

Gerald is doing a great job of speaking for me. We never have used limit switches. Each job has its own zero based on what is convenient. In some cases the z is based on the top of the material to be cut some times on the table top. X and y are usual the lower left of the material or in the center. I like to place it to where if the electricity fails or computer crashes I can relocate.

Really what you do is match the zero on the table to where you placed it in your toolpath program.

Alan

[mrdovey]

I have the Hall-Effect limit switches on my 'Bot and really like 'em.

I mounted a 2"x4" registration rail along the x-axis and cut it straight and square. I use it to ensure that everything I put on the table is parallel to the x-axis.

When I did the truing cut, I recorded where the true edge was (relative to the limit switch) so that I'd know where my work is.

I've built a number of work-holding fixtures and jigs since then; and have used the bot itself to cut an exact registration edge on each fixture. This approach has allowed me to set a workpiece into/onto a fixture or jig, tighten the clamps or turn on the vacuum and cut immediately (without manually positioning the spindle and squinting to see if I was where I wanted to be.)

Rather than using a Z2 command, I maintain an &X0 and &Y0 value for the home position of each fixture, and apply that offset to all coordinates used while cutting parts on that fixture.

One of the benefits is that I can always re-establish the exact same reference frame following a power or system failure.

Another benefit is that all parts can otherwise be coded as if they were at (0,0)...

Interesting that Ron finishes up with a move to (0,0) - I finish up with J2 48,24 to have the spindle as far away as possible from any humans who might be around.

Just goes to show that there's more than one way to skin a cat


...Morris

[gerald_d]

Morris, we also use jigs (or pallets) and our way of aligning them is...

1. roughly fix some scrap along the x and y axes (typically 2 blocks on each axis held with 4 g-clamps, or away from the table edges held down with screws or bolts)

2. cut reference faces on this scrap with a known cutter diameter and set x=0 and y=0 when the faces clean up.

3. lay the jig (pallet) against these reference faces. The jig's program does the rest - it knows that 0,0 is half_a_cutter_diameter off the corner of the jig.

With this method, we do not have to preserve a permanent reference edge or position on our table. That is also a liberating experience.


(This means we can have multiple jigs on one table, with multiple 0,0's)

[mikejohn]

Gerald
How do you hold the material to the jig?
How do you hold the jig to the table?
Is there a standard table size for the PRT96 (2440 x 1220)?
.................Mike

[gerald_d]

Material to jig - typically with bolts that come up through the bottom of the jig, and nuts/washers/battens down from the top. Also with toggle clamps (http://images.google.com/images?sourceid=navclient&ie=UTF-8&rls=GGLD%2CGGLD:2004-43%2CGGLD:en&q=toggle%2Bclamp)

Jig to table - if it is near an edge of the table, then the first choice is g-clamps. Alternatively, screws or bolt&nuts into/through the table.

Standard table size - I doubt it. Why do you ask?

[ron brown]

Morris,

To clarify a point. I use 0,0 when I am going to make a small run or carve letters where I might want to change something - like Z height.

When I run multiple parts after the file is "proven" I do clear the machine from the area for parts change.

To fix jigs to the table I sometimes use dowels or pins of a certian diameter for location and threaded inserts if it is a semi-permanant jig or a jig I will use many times. These jigs usually employ toggle clamps or have vacuum ports. Pins or machined blocks are used for XY register and the jig is held down by other devices such as screws or bolts/inserts. I have had zero plates on these same jigs.

As we evolve, our thinking should change. I know as I see things done I put my failures into the mix to see if I can get a new failure.....

Ron

[stickman]

Hey Ron,

Do you have a few pictures of these jigs, that employ the zero plates.

Jay

[mikejohn]

Gerald,
My end cross pieces which are C channel, with the 'open' side facing outwards are exactly 2440mm apart (as per the Dorrington Mark II).
If I make the table this length, I can clamp to steel at the ends, rather than MDF. I thought the width might be slightly larger (40mm wider) than the maximum y travel.
Plent of room to clamp along the sides.

I have a general question.
The x, y Unit scaling values are correct (7.21839), the bot moves the distance I tell it to.
The z axis was wrong. After a number of tests and measurements, it came out close to an error of 2.187. I applied this to the 'wrong' z unit scaling value, 78.7402, and got a value of 36.0037.
Testing this seems to give the correct movement, but sounds a bit rough when moving.
Does anyone have the precise value, closest to 36.0037 for a mm set up?
Thanks as usual.

..............Mike

[mikejohn]

I've just looked up Geralds list here (http://www.talkshopbot.com/forum/cgi-bin/discus/show.cgi?tpc=312&post=18437#POST18437) and find a value of 916.733 which, divided by 25.4 gives me 36.0919. Is this the value I should be using?
.............Mike

[gerald_d]

Mike, it is awful to clamp to a bit of tapered channel. Rather run the MDF about 50mm past and clamp to that.

Whether inches or metric does not change the unit values (the factor 25.4 does not apply).

Suggest that you follow-up the rough z (and strange unit value) with SB-support. But, the 36.0037 looks like 5 times your 7.21839, which is also an unusual value??? Are you sure that your screen is displaying mm and that you are measuring mm?

[mikejohn]

Gerald
I have got the ShopBot running in DOS, SB software version 2.39 with the 'old' computer, which is in the workshop. (If I went to the workshop in my undies I would get arrested, its 6kms away!) I do, however have the settings I used when I was using this computer with SB windows version 3.+.
If I (VU) here I get x and y at 28.8735 but z is 36.0918.
It says somewhere in the manual (but I cant find it) that you have to be in the correct system, Imperial or metric, before setting the figures, but if you set inches when in imperial it will convert when going into metric. I got the connection to work by restoring original values.
The 'strange' unit values came from the software converting. But the Z was wrong.
I will try the 36.0918 in the morning. It is 5 times, but the values given in inches in page 5 of Calibrating the ShopBot PRT in the manual z (916.7328) is 5 times x/y(183.3466). The motors appear the same between x/y and z, don't know about the gearboxes. x/y pinions have 25 teeth, z has 20.
I will take your advice (whenever don't I
) on the table length.

From my first post at the top of this thread "And before you ask yes I was in mm not inches."


............Mike

[gerald_d]

It does look like your unit values are 25.4 times smaller than what I have in my table. I am 23km from my shop, but still in my overalls.

Looks like you must divide the values in this post (http://www.talkshopbot.com/forum/cgi-bin/discus/show.cgi?tpc=1843&post=17376#POST17376) by 25.4.

A big g-clamp inside a channel doesn't work either.

[gerald_d]

This will then be the table of possible Unit Values:
Inchmillimeter 63.6622.5064 70.73552.7849 76.39443.0077 84.88263.3418 91.57333.6052 101.8594.0102 114.5924.5115 127.3245.0128 141.4715.5697 152.7896.0153 183.3477.2184 229.1839.0230 254.64810.0255 305.57812.0306 366.69314.4367 407.43716.0408 458.36618.0459 509.29620.0510 565.88422.2789 611.15524.0612 733.38628.8735 814.87432.0817 916.73336.0919 1018.5940.1020 1131.7744.5579 1466.7757.7469 1833.4772.1839 2037.1880.2039 2933.55115.4941 4074.37160.4083

[mikejohn]

Normal fast response from support, well done Frank!

He says in his reply:

"I checked my software with the units set to metric and the value is 36.0918, so using Gerald's list and the calculator got you close enough. If you are comparing the sound of the Z to the X and Y then it will sound rougher due to the different gearing and is not a problem. If you are concerned and can get a phone to the shop I can listen to see if it sounds normal."

Thanks again Gerald, once again to the rescue


................Mike

[mikejohn]

Today I made sawdust

Now this may not mean a lot to you, but to me it was all my Birthdays and Christmas's wrapped up in one.
OK, it was only on a scrap piece of ply, and it was only a 6 letter name 35cm x 10 cm, and it was with a small single speed Black and Decker router fixed to the carriage with two hose clamps but it worked.
Mind you, there's no way I'm putting a picture of my result on the Forum! (You know how mothers love their new born babies, but the rest of us find them incredibly ugly?)

...............Mike

[gerald_d]

Happy Birthday!

A six-letter name? That's very kind of you - thanks!

But with a Black & Decker?? Yuck

[mikejohn]

Gerald.
The name was 'Albert" The person who bought our first Rocking Horse named it 'Albert' so the first in any line we make is now called 'Albert'. First pull-along is Albert the Basset, first automaton M.Albert (said with a French accent) first paper automaton Alberto the Clown, now first thing off the ShopBot a sign saying Albert.

Good day for me, not so good for you from what I read elsewhere.
..........Mike

[mrdovey]

Congratulations, Mike! It is indeed a thrill to watch that first cut...

[bleeth]

Try "supercallifagilisticexpiallidocious"
(With a horse relief under it).

Or to quote one of our most beloved old yanks when he was young and making an Italian flic-
"Now we begin"

(Wait till something tricky comes out-then you'll really feel hot.)

[mikejohn]

Dave,
I would try "supercallifagilisticexpiallidocious" but I only have the 96 bot, not the 120


As for Horse relief, they relieve themselves everywhere, (but we do do a good line in little bags of Rocking Horse manure at craft shows!)

..............Mike

[gerald_d]

Mahatma Gandhi, as you know, walked barefoot most of the time, which produced an impressive set of calluses on his feet. He also ate very little, which made him rather frail and with his odd diet, he suffered from bad breath. This made him ........ A super calloused fragile mystic hexed by halitosis.

[robtown]

You'll pay for that one... ;)

[dvanr]

Ah Gerald,

It seems Gandhi and you are a bit like minded.

"There is more to life than simply increasing its speed." Gandhi

[mikejohn]

This is the beasty actually cutting into a piece of scrap Sycamore. Its the baby Black and Decker fixed speed router hose-clamped to the z axis. The bit is a cheapy 60 degree bit.
I know I have a long way to go, bit it shows something is working.
I now have the bracket to mount the big router,and will post pictures as I have used a different technique to Geralds mount.

..........Mike

3181

[gerald_d]

Good stuff! That V-Carving is mesmerising, ain't it?


Why is the "h" out of line?

[mikejohn]

The carving started at the "h", went to the "T" and then continued on round. I dont know why its out of line. I was going to fit the new router bracket today, and try again, but we went to congratulate a friend on a new baby, and I don't think operating the ShopBot after three glasses of wine is a good idea.


..........Mike

[sales@fretworks.co.uk]

Mike

Congratulations, the wait is over...now you can start a "NEW" learning curve and make some money.

BEST WISHES

Paul

[mikejohn]

To all newbies Don't try to cut without starting the router


I have a question. I am cutting through 20mm pine with a 6mm bit in 2 passes at 15000rpm 25mmsec. It works fine. What I would like to do is offset my second cut 2mm from the first.


3182
How will the bit react to having to cut away the red part, as well as the 10mm deep second cut?
Do I have to change any settings?
Am I likely to break the bit?
...............Mike

[gerald_d]

I wouldn't recommend that. Maybe 1mm or .5mm, but 2mm feels like too much. Wouldn't change any settings.

With 6mm bits cutting wood, we find that we mostly wind the router up to max. speed (>22 000) and move around 35 - 40 mm/sec. Maybe we will get a shorter bit life than at 15000 rpm, but the cut quality seems to be better.

Sometimes, if the cut must be deeper than the cutting edge length (CEL), we do that 2-3mm offset that you show above, then cut away the red, then plunge deeper in. This is to give clearance to the non-cutting part of the cutter in the scrap part of the job.

[gerald_d]

1. Don't try to cut without starting the router
2. Don't leave the router running in the wood if it is not moving. (friction causes heat causes fire)

[mikejohn]

If I can move at 35-40 mm/sec then I can make the extra cut of the red bit, then plunge for the second time.
when you read my mail, you will see I don't know how to do this!

...............Mike

[mikejohn]

I have to admit it was with a V-bit only 1mm into the surface. It did alert me to the danger though.
.............Mike

[gerald_d]

You are an AutoCAD man.....

Draw Boundary, copy this first boundary to a z=deeper position, you now have a "stack" (aero language) of boundaries. Look at the Properties of each boundary and set that boundary's Elevation so that it matches the cutting depth. In Vector, select the boundary's top to bottom. (We use 3D views in both ACAD and Vector).

Basic logic is to solve these problems in AutoCad, first and foremost.

Supper calls!

[mikejohn]

Gerald
I am now making files in a routine manner.
It took some figuring out, but I am at last there.
Years of reading this Forum still didn't prepare me for the amount of re-learning I had to do.
But I can now read .spb files quite easily!
Like most things, its obvious when you know how!
..............Mike

[gerald_d]

Mike, it is only 2 weeks since you started this thread - you are making it look too easy!

[mikejohn]

Gerald
I just looked at that date, and I am truly astonished. It feels like forever! I know Ive aged 5 years in that time!
I don't know if you can see this well.

3183
It's the file to cut a single 2200mm x 220mm x 20mm pine board to be held on a pallet (or jig).
Using a 6mm bit, there are 50 holes, 38 of which are countersunk. Using two passes 10mm deep, a move speed of 40mm/sec and 24500 rpm on the router. The 8 large circles you see are just markers where clamps will go.
14 minutes to cut in simulation. This is fine for me.
Now I have a technique, its going very fast. I do stop and check each file in simulation
As I'm cutting through I have zeroed from the table. I will air cut with zero 3mm above the table, and put yogurt pots where the clamps will go, to see how close the bracket comes.

.........Mike

[gerald_d]

Why use yoghurt tubs? You can measure the router bracket footprint, see how far your bit protrudes from it, measure the curve of the g-clamp and draw it all in AutoCad. (We never have this issue with the die-grinder).

I may have been a tad optimistic with the speeds for your pine which is a lot closer-grained than ours. Our stuff (http://www.thorpe-timber.co.za/products.htm) grows way too fast, and the rings are typically 3-5mm apart.

Did you see my mail reply re bits?

[mikejohn]

Gerald
I'm chasing down the bits, thanks.
Dropping the speed to 30mm doesn't greatly increase the time, so I'll start there.
The router is good, but big.
I think I will try a die grinder, once I'm cutting.I like the thin neck.
I have drawn the clearance of the footprint in AutoCad, just needed the confidence of not knocking over the yogurt pots!
Is 8 clamping positions sufficient over 2 meters?
...........Mike

[mrdovey]

I haven't had difficulty with larger (~4mm) offsets between adjacent cuts with (North American) Southern Yellow Pine.

I was intrigued by Mike's clamping strategy - and will offer reassurance that with eight clamps the board probably won't go far


Here are a couple of pix from this afternoon's work on a panel saw showing how I clamp boards using T-tracks and shop-built screw clamps. This clamping approach might work well for you, too.


3184
3185

These were 2"x4" fir boards being cut with a 1" cove cutter fed at 1.5"/sec and spindle running at 18000RPM.

[mikejohn]

Morris
Thanks a lot. You have just given me an answer to my problem.
At this time I have 3 main tasks for the ShopBot.
One is using 220 mm (9") wide, 20mm (4/5") thick pine board.
If I have a jig with one fixed long rail 20mm high going the full length of the table, and 4 or 5 sliding (T-track) stops on the other, I can screw into the board I am cutting sideways, through the rail and stops, locking the sliding stops well away from the cutting area.
As I said above, it takes about 15 minutes to drill and cut 1 board. I can easily get 4 'pallets' onto the table, clamped at each end.
A great solution which has probably been spoken of here before, and I haven't understood it.
Now If I have the pallets like this across the table ____A________B________C________D____ if I cut first A (15 mins) then B (15 mins), then when C is cutting remove and replace the A pallet, when D is cutting remove and replace B,and so on. Providing I slide the pallets out, don't lift them, what problems might I face?
If necessary it would be possible to create a holding piece at the table end that would prevent the pallet from either lifting or moving sideways during removal/replacement. This would give me 15 minutes to take off the cut boards, remount new boards, and hopefully remove the cut pieces as well.
It sounds good in theory, but one thing I am finding, theory and practice are not necessary the same thing with 'Getting the ShopBot up and running'

..............Mike

[mikejohn]

Morris
Just read it again.
You say shop-built screw clamps. Does that mean as you tighten the clamp it screws into the wood you are cutting?
........Mike

[gerald_d]

Mike, consider for a minute what happens if you clamp a plank in a bench-vice and then try to rip it down its length, cutting between the vice jaws....

Sideways clamping can only work if there is solid material left directly between the pressure points after cutting. Once you install T-Track into your table, or any other metal device, you have a table surface that is unforgiving of a cutter that wants to visit The Deep.

Our approach is to use a T-shaped batten (approx 45 wide across the top, 35 wide at the bottom) between the planks. The top of the T holds the planks down, the leg of the T spaces them. Then we have bolts up through the table, through the T's, with nuts on top. Sometimes a spring around the bolt between the table and the T. Planks are slipped in from the end, nuts are loosened, tightend with a cordless drill. The bolt holes are permanently in the table - an sbp file drills them when a new spoilboard is laid. The T-battens are consumable - often (intentionally) hit by the cutter. In your sample pattern, we would try to get the straight cuts to the inside of the plank, and the curvy bits to the outside - so that there is more left to clamp on.

Are 8 clamps enough? Probably an overkill - except for a rush job where you are pushing too high a speed with a very blunt up-cutter.

[mrdovey]

(Scratching head...)

What I think you described sounds like it should work ok.

The fixture pictured above is a piece of plywood with a fixed rail (just as you described) against which the workpiece is clamped using a sacrificial board ("caul") and clamps that "ride" and lock to the T-tracks. My preference is to not have any kind of metal in contact with the workpiece (IOW: "near the cutting tool"). My clamp is nothing more than a drilled and threaded block of aluminum with a fat (3/8") socket head cap (machine) screw that presses against the caul - which, in turn, squeezes the workpiece against the fixed rail. I've found that a "squeezer" every foot or so is more than enough clamping force for even my most agressive cutting.

I don't think I'd do the pallet rotation you've described. Instead of jumping up to move a pallet every fifteen minutes, I think I'd prefer to load 'em all up and leave the 'Bot alone for an hour while it did 'em all. Then I'd reload all four at once and restart. It seems safest to me to leave the 'Bot alone while it's running. With any decent clamping setup, the changeover/restart time should be less than five minutes - and your productivity will be better with only a quarter of the interruptions.

My inclination would be to build a single pallet (or two if I needed to produce a lot of the part) to hold four of the 2200 x 220 x 20 blanks. That should minimize alignment hassles and simplify setup.

Everybody seems to do this stuff differently. I'm basically lazy - and cowardly in my dealings with any tool that has more horsepower than I. I like to make work easy - and I don't think that courage should ever need to be a factor around whirly sharp things.

[gerald_d]

I think that Morris and I have cross-posted - Morris was not aware of my post when he replied to Mike.

[mrdovey]

Gerald makes a good point - and I failed to mention that the T-tracks on my pallet do not extend under the workpiece. Their ends are covered by the caul in the photos.

I'm not so sure about the vise/ripping analogy. There needs to be enough "meat" left for the "leavings" to be held; but somehow that's never been a problem for me.

I really like the 'T' shaped hold down idea. I think I'll try using that with a pallet (rather than drilling my table). Thanks!

[mikejohn]

Gerald and Morris,

The planks are thicknessed (planed and sanded) to 20mm. The widths are cut to 220mm. The length will be what I want it to be, up to the maximum table length.
So how about this.

3186
I divide the width of the table into four work spaces. The T-rail hold downs marked with 'A' are fixed in there position, but can of course be tightened and slackened to allow the timber to be taken on and off the table.
The half T-rails marked 'B' have slots for the bolts, and can be moved in and out (towards or away from the planks) to take account of width differences in the sawn planks.


3187

As the position "A's" are fixed to the table, those plank edges are aligned to the ShopBot, regardless of any width differences with the planks, so a single file to cut all four planks is possible, taking on board Morris's "I don't think I'd do the pallet rotation you've described. Instead of jumping up to move a pallet every fifteen minutes, I think I'd prefer to load 'em all up and leave the 'Bot alone for an hour while it did 'em all.
The two lines marked 'C' are the outer limits of the bit in the y direction.
Gerald, do I need the springs? I imagine once I have the board under the T-rail sliding it in or out won't be helped by springs.

Is there a simple way to hold the bolt heads under the table?

I 'air-cut' the file shown above this morning. Apart from a couple of places where I obviously had a minute gap between lines, and the Z axis jogged up and down it 'cut' very well. Actually slightly faster than the simulation time.

Thanks as usual
....................Mike

[gerald_d]

Mike, we tend to use "cup-square (http://www.mcarthur-group.com/images/fix/cup_square_hex_mix.jpg)" bolts from under the table, and don't loosen the nuts far enough for them to fall out. For our MDF pallets we have bolts with big square heads welded to them - they are held into recesses with hot-melt glue.

Half-Tees are not a great idea as they twist over. We balance the outer T's with scrap of the same thickness as the plank. Springs are a luxury - they make compressed air cleaning easier if the battens are lifted off the table.

[fleinbach]

Gerald,

I've never heard the term “cup-square” before. Over here they are called “carriage bolts” I read somewhere before they were used in the early days for building the old wooden carriages.

Here's a link to a company that was manufacturing them in 1865

http://www.lamson-sessions.com/OurHistory.html

[mikejohn]

Gerald
In the lower picture,if I have the distance greater than 220 mm (say 225 mm)in order to allow for planks of differing sawn widths, how do I ensure the plank won't move laterally? Or will the cramping pressure of the T-rails hold it? Or do I need some sort of wedge?
I would like to avoid precisely cutting the board, to the exact mm, if I can.
............Mike

[mrdovey]

Mike...

I like what you've drawn. It looks as if it might work well with a Propel nut inserted into the botom of a pallet, with a button-head cap screw through the T (inserted from the top.)

Hmm. I actually already have a palette almost like this - it'd just never ocurred to me to use the T-shaped retainers. I'll try to remember to take a couple of photos today to post tonight.

Gerald, I like the way you think!

[gerald_d]

Frank, we always used to call them them carriage bolts, but in the last few years the guys at the sales counters give us a blank look on that one. Their faces light up with "cup-square", so I supposed that is the official new name. Many companies still make them today.

Mike. The "cramping" pressure (plus friction) stops the lateral movement. We typically use 8mm lightly tightened. Over-tightening might split the grain of the T if the head of the T is too wide. That's why we keep around 45mm max, with a large steel washer (penny, fender, or body washer in local parlance).

Morris, simplicity is the ultimate sophistication.
If you srew a bolt into a jig-nut from the top, you better be sure that sawdust hasn't collected in the nut. Otherwise the bolt will push the sawdust out and lift the jig off the table.

[mikejohn]

Gerald
If I use your 45 mm top, 35mm base T, I can cut the board as you suggest, leaving this waste.


3188

Even where indicated with the two green circles I would not touch the T.
If I went to 55mm tops/35mm base, this would allow me some variation on board width, with just 4mm shaved off the T at the green circles.
Can you see any problems?
Around 30% of my ShopBot time, per annum, will be taken up by cutting/drilling boards of this size, so I guess I need to get it right


Thanks
Mike

[gerald_d]

Don't see any problems Mike, but don't make anything too permanent at the beginning - you need to get the feel of it all first. When we have "plank" jobs, we have no control over the width (or thickness) that our clients supply to us - your planks sound a lot more standardized.

Morris, I am sure that you can you see the hassles with a dust collector foot in this setup.

[mikejohn]

Gerald
I plane and sand the planks to the thickness I want, my supplier cuts them to width.
I intend to create my T hold-downs out of beech, on the Shopbot.
How many (8mm?) bolts per full length hold-down?
Now, if I'm cutting 4mm ply, can I simply turn the T over and clamp it down?

...........Mike

[gerald_d]

Mike, the best tool for making T's is a table saw - no debate about that. (It saves me from figuring out how to hold it on the SB table). Bolts are probably at 600 centers (watch out for cross-members under the table). You can cut small notches in the top of the T to hold the ply when upside down, or you can shorten the T to hold both ply and planks - remember how to skin a cat? You might even find a useful alu profile that can work as a T.......

Start with cheap pine and get a feel of all of this first.

[fleinbach]

Mike,

If you want to remain on the safe side here is a link to a company that supplies nylon fasteners including "cup-square" carriage bolts. http://www.nylonalloys.co.uk/homepage.htm

[mikejohn]

Gerald
I buy pine dried to 8% for $185 per cubic meter, cut to 220mm x 25mm x 4 m.

3189
So this is the latest plan

If I use aluminium I won't be able to scallop the edges where necessary.

I thought, If I plane the timber to 35mm high x slightly over 55mm wide, the length of the table, I can then clamp it and drill my five 8mm holes. Then I can bolt the timber to the table, ShopBot cut the T shape to fit the drawing. Do one side, flip over the rail, do the other side. The reason for the 'slightly over' 55mm is in case I don't clamp it exactly square to the z axis. Cutting it will square it up.

By removing the second and fourth hold downs from the table I can fix 495mm wide plywood. As I determine this width, it can fit the hold downs exactly.

............Mike

[gerald_d]

The proof of the pudding shall be in the eating.....

[sales@fretworks.co.uk]

Quote...Mike, the best tool for making T's is a table saw - no debate about that...

That would only be done with adequate guarding.

Working safely and within the law. In the United Kingdom whenever you are employing anybody, even your family and even if they are only working part time you are required to work within the law. The 'law ' means the Health and Safety at Work Act (HASAWA) guided by the associated Codes of Practice (COP's). Overseas visitors will find similar regulations in their countries but one countries rules will probably have the same objects - to save life and limbs.
If you are only working on your own in the back of your garage, it is prudent to work within these recommendations because these laws and their recommendations are there to protect you. Generally it is your fingers, lungs, eyes and ears that suffer most. If you are employing someone, and they suffer, it will be your pocket if not liberty that may suffer also.


Paul

[les_linton]

Frank, Morris, Gerald and Mike,

I too have been watching this post for the past couple of weeks.

Just getting started here also and while I have seen some sort of variations in my head for these clamping solutions, your posts with pictures, questions, concerns and warnings have sure brought the logic of what we are all trying to do to light.

I think that this would be a good day to work on some of these clamping issues, so a couple of quick questions.

Morris, you say that you have a 2 x 4 registration board along the x axis. Are you mounting that to the frame cross pieces and squaring it up with the bit mounted in the spindle/router?

If using the aluminum t-trax, is there any real issue about using them across the y axis at several locations? I would think that this would give you greater flexability for mounting stock to the machine as well as other fixtures.

Thanks guys..

Les

[gerald_d]

That was quite a sermon Paul, and I am wondering what I said to deserve it? Did I suggest anything unlawful?

[mikejohn]

In Romania I have a neighbour whom we call 'fingers', mainly because he's short of a few on each hand
.
You walk through his workshop with your heart in your mouth. Guards? What guards. The guys working his planer thicknesser climb over and around mountains of shavings, clearing it when it blocks by poking a stick into the still-running machine. To watch them shoving timber through the table saw or jointer, unguarded, no push-sticks often, is extremely worrying.
I've tried to discuss safety with him, but he just smiles, saying it was like this back to his fathers day.No ear defenders, dust masks, dust extraction. Floor piled high with finished and unfinished parts. Health and safety? If they come around, slip them a little something, they go away.In Jan 2007 Romania joins the EU. I fear many are in for some nast suprises.

Hit the old edit button fast there
. I think, Gerald, this was what Paul was trying to say, rather than querying your technique

Les
It is becoming clearer to me that, if you have a restricted use for the ShopBot,as I have, holding the material is more straight forward than if you are cutting different files and materials each and every day.

................Mike

[sales@fretworks.co.uk]

Gerald, No I am not but I think our Health & Safety are pretty draconian in their regulations.........http://www.hse.gov.uk/pubns/wis16.pdf

Here in the UK you can now be fined for drinking water from a bottle or eating a bar of chocolate while driving, you can imagine what they are trying to do in a workshop full of machines.

Regards

Paul

[sales@fretworks.co.uk]

Hey Gerald, you edited you post.

No it wasn't a sermon....and no it is not unlawfull (yet).

I recently got pulled up by the HSE for having a stack of timber near a door with one piece pulled out to hold the door open.. as Mike said I did think of slipping him something !! but I left it in the stack of timber.


Regards


Paul

[mrdovey]

First a note to Les...

My 2" x 4" table registration rail (reference edge) is fastened only to the replacable MDF and plywood top. I've posted some photos (http://www.iedu.com/DeSoto/pix.html) on my web site (http://www.iedu.com/DeSoto/). I do square it up after installation and record its position (programatically).

Yesterday I promised some photos of one of my pallets - and while this has only a single row of threaded holes to which T-strips might be attached, this might be a step in a useful direction:

3190
Here's a close up that makes it easier to spot the row of threaded holes:

3191
I turned the pallet around so you can see how the Propel nuts have been set into counterbores:

3192
And for anyone lurking who isn't familiar with Propel nuts, here's what they look like before installation (rust added after purchase):

3193

I've generally tried to avoid routing my T-tracks. The 'Bot is perfectly capable of routing Aluminum [Aluminium elsewhere?] but T-track is too pricey for me to waste.

...Morris

[mrdovey]

An addendum concerning my use of T-tracks: When I use these handy clamping accessories, I install them with the top of the track about 0.01" (0.25mm) below the surface of the pallet.

I'm fairly diligent about checking my programs before actually cutting. Since I got my 'Bot two years ago, I've managed to not ever cut either clamps or spoilboard - and I design some of my workholders to be (intentionally) carved up in the cutting process.

...Morris

[robert_cheal]

Back when I had my ShopBot on order I searched the forum for information about the different hold down techniques. For me holding down various parts seemed to be one of the major challenges. One technique that I picked up on the forum for solid wood and I still use often is to leave the board a little thicker than I need, and then machine the parts to just shy of the board thickness after that I run them through a small thickness planer and the parts drop out. This has worked well for some fairly small parts. The caution would be to a plan for keeping the small pieces from the layout waste from dancing around into the planer blades as the board goes through.

As expressed by others cleaning up tabs on a part with a flush cutting router in a table is really quite fast. I like to create the placement of tabs when the part is drawn in AutoCad.
Then create my tool path in Vector 9 (I have not felt the need for Parts Wizard yet), I would not consider Vector for drawing the part but it works great for getting the cutting sequence worked out.

When I first started out I followed the lead of others and placed aluminum T-tracks at 12” o.c. on the “Y” axis. I then placed continuous fence along the “X” axis, this fence has horizontal screws at 6” o.c. The tips partially extending (about 2mm) through the fence pointing toward the “Y” axis, the screws can be backed out of the way when needed.
I then created a similar adjustable fence that could be locked down along the T-tracks in “Y” axis. To hold a board down I press it down and tap it into the screws of the “X” fence then bring the adjustable fence against the board, tighten it down to the T-track and tap the fence and board tight against the fixed fence. The extended screw tips hold the part in and down flat on the table. This has worked out for me when running specialty moldings for the most part the challenge is when the finished board to be machined is badly warped. The next phase of this hold down method has evolved to where the screw tips have been replaced with a steel pin (sharpen point) that retreats out of the way by way of a spring, the pins can be pressed into the board to be held down and then released with a cam lever, the cam levers are machined from ½” Baltic Birch. Each fence is about 1” thick x 4” wide. The idea to use this if for example you are running multiples of a 3” molding you set the fences at 3” apart, lock the fence down, 3” boards are placed into the 3” slot and as the board is pressed down flat to the table the cam levers from each side are turned and the pins extend into the board to keep it tight and flat. Then when the board is machined you quickly release each cam lever and the board ready to come out.

I realize there are some limitations and some obvious precautions for the variety of hold down techniques. Every one comes up with solutions that work for them, I am still trying to figure out how I would create a vacuum system to meet the needs of the different things that I cut on the ShopBot, fixtures with quick referencing, change over, etc,. But it sure is fun learning and humbling at times especially when I look back to figure out why and what I have just tried to say. That’s the beauty of being a victim of my own perception.

Robert Cheal

[bleeth]

If you put the SB dust shoe on the far left car rail as it is designed to do with a spindle then if you are cutting close to the far left corner of the table the shoe can come uncomfortably close and even hit the leg of the SB designed table. The best remedy for this that I can see is to affix a stand off block between the L bracket that attaches to the car rail and the C bracket that holds the dust chute so the chute is more centered between the rails.
I don't know if this is a problem with the router system as I do believe that the parts are somewhat different and I haven't looked for this when around someones table with a router.

It's kind of amazing to me that in over a year of ownership today was the first time that I had cutting to do close to 1,48 and found out the hard way. I got "lucky" in that the toolpath was almost done cutting a series of parts and this location was on the last part but spending the next couple of hours resquaring the carriage rails again was not a great Saturday.

Dave

[mrdovey]

Dave...

I must have a special talent because I did that almost immediately I had my 'Bot assembled.


My solution was to route the dust under the Y-car in the +X direction - just far enough that I could attach a 4" hose to the DC.

If you haven't seen this lash-up, take a look at my dust collection web page (http://www.iedu.com/DeSoto/dust_collection.html).

...Morris