Boxes, Clamps and Thien Dust Collector! #83 - Brighton ON CA

[camilleg]

Originally Posted by Melissa

The rollers are not contacting equally along the length of the table. I have more wobble at one end than the other.

If the rails are straight but not coplanar ("propeller twist" as we say), then the plane of an untwisted gantry will agree with the rails more closely at some x positions than at others.

Thought experiment: Gantry 1 foot long along x dimension. 10 foot rails. Both rails straight. 3 of the 4 rail endpoints level, one endpoint a foot too high.
Parked at the "good" end, the gantry will wobble 1/10 of a foot. Parked at the "bad" end, although the vertical offsets at the gantry corners haven't changed, the wobble will change, because the rails are slightly farther apart and thus the V-wheels won't center on the rails the same as they did at the "good" end.

Does that theorizing match reality at all? I think it applies to a twisted gantry just the same as to an untwisted one.

[melissa]

Untwisting the gantry is proving to be a bigger challenge than I expected.

I upgraded to a 4-ton bottle jack, which had no problem twisting the gantry. Due to the large forces I expected, I double-clamped the close end of the gantry, just in case. I also installed a few extra C clamps to hold the pipe clamps, in the event things decided to go "sproing".

bottle_jack.jpgdouble_clamped_gantry.jpeg

Well... at 40mm of deflection, the jack slipped out (due to the twist of the gantry tube, which it was pushing against). Two of the pipe clamps jumped loose as well, fortunately without damaging the rails.
aftermath_1.jpgaftermath_2.jpeg

End result: the gantry is riding better, but still wobbles at one end of the table. I need to take careful notes as to the lateral wheel positions at all points along the X rails, to figure out exactly what's going on here.

[Kobus_Joubert]

I think the biggest problem there it the temperature...judging by your clothes ..
Good luck with the cold steel....We are moving into our WINTER and up here in Johannesburg it will be cold this winter.

[camilleg]

So how hot are J-burg autumns, then, if steel at ambient temperature is easy to bend?!

[Red_boards]

I got my gantry rolling smooth.
Moved table to get access to the other side and gantry is lifting on two corners at one end of the table.
Move table back level and all is good.
Unbelievable the amount of twist in a seemingly rigid table - and mine is only 3m long with some serious beams.

[melissa]

Originally Posted by Red_boards

Unbelievable the amount of twist in a seemingly rigid table - and mine is only 3m long with some serious beams.

I had to laugh when I read that!

And some people say that a table built of aluminum extrusion is stiff. I don't think so.

[melissa]

I finally figured out how to really measure the rails. The challenge is to measure the V profile where the wheels ride. A long straightedge on the top surface measures the top flat, which presumes a perfect V grind. Mine wasn't perfect .

With the gantry in place, I came up with this contraption. The dial indicator is screwed to a block of wood, which is then clamped inboard and outboard, and readings taken at regular intervals down the rail (I used the bolt locations, numbered from 1 to 29).

Important things to note:
- measurements are relative, not absolute
- measurements only pertain to one side of each rail
- with some trigonometry, the effective height of the rail can be calculated. This is left as an exercise for the reader .

Also, I'd be careful to drive the gantry slowly with the dial indicator engaged. I'm not sure what speed the ball-bearing tip is rated for .

measuring_rails_outboard.jpgmeasuring_rails_inboard.jpg

[melissa]

Fun with cable chains today. Or not.

I just learned that there's a limit to the "self-supporting" distance of a cable chain, and I was rather soundly exceeding the limit on mine . The hazards of building a really long table:

chain_sag_1.jpgchain_sag_2.jpg

The manufacturer's suggestion was to either install a glide tray, or upgrade to a stronger cable chain. Well, it was after 5pm and I couldn't get a price quote, so I went exploring other options.

Skipping all the bizarre ideas I batted around, this is what I came up with. Instead of raising the motor to clear the glide tray, I decided to lower the cable chain completely. Due to the length of travel, I also need a glide tray for the lower section, to guide the chain laterally. Here's the mockup with construction lumber, just to make sure it'll work.

chain_lowered_1.jpegchain_lowered_2.jpeg

chain_lowered_3.jpg

[KenC]

Hope this will give some preview of your set up.
DSC00349.JPG
DSC00350.JPG

[camilleg]

Any problems with buildup of static electricity, particularly for the "top half" of the chain sliding along, instead of merely draping onto, the tray? Ken's looks like nonconductive painted metal. Imagine one made of cat fur... after an hour of operation you could stick the MM to the wall.

[camilleg]

Ball bearing as a fifth wheel... clearly we're no longer in Road&Track acceleration test track territory!

[melissa]

All metal parts of the MechMate get bonded together with a ground wire. The tray will be connected to the base table. The gantry (and hence the cable chain) is grounded through the wire, which runs through the cable chain.

No ball bearings or wipers needed. It's an oscillating motion, not rotary.

[KenC]

The MM chassis/structure is grounded. I didn't use any wires to bond them, just make sure they have bare metal contact such as bolt through bolts, welded. My multimeter shows close circuit everywhere. No problems using my auto touch off concoction.
Static problem is only severe at the dust collector ducting. Simply wind some grounded cheap wire around outside of the dust collector flexible duct solves the problem.

[Gerald D]

Ken, if your ground runs through your v-rollers, just think how they weld together if you pinch the power cable to the router.

[KenC]

Luckily nothing of such happened... It takes more then a few 10s of amps & it will only occur when there is a major short circuit somewhere.... & anyone who intentionally direct a live cable to my MM will have to walk over my dead body

Forgot to mentioned that the spindle is also grounded individually via its power lead... that also help to reduce the chance of welded bearing... All I'm trying to say is if you practice reasonable electrical installation practices you already prevented disaster from happening.

[melissa]

I spoke to a technical rep from Murrplastik yesterday about the glide trays. Apparently they're expensive and difficult to source (probably special order from Germany).

I opted for the local solution, and contacted a sheet-metal supplier to have some steel cut and bent to the appropriate shape. 15 minutes for them with a shear and a hydraulic brake, and this is what I got:

steel_glide_tray.jpg

It still needs a coat of paint, but that will happen after welding. I'm planning to bend some custom supports for the tray, due to its position being way different than the plans. And with steel supports, I may as well weld them instead of bolting them to the tray.

[KenC]

when the cable is installed with reasonable alignment, it will wind & unwind in a fairly straight line, the trick is at the fix end bracket alignment.
Using a trout may cause more problem in reality, the cable may just kink up at the side while you are not looking... My solution is to remove the side wall all together, cable trays are cheap, 4" wide tray complete with cover is less the USD3 per 8' length for . cheaper then cutting a piece of MDF board of the same size... & its painted...

[JamesJ]

Ken, do you know of any US suppliers for the tray you described?

[KenC]

No clue of any US supplier.
These are very common electrical trays, available in every single electrical supply stores in Malaysia. Most probably you can find some if you ask your local friendly electrical supply stores.

[JamesJ]

Thanks.

[melissa]

X racks are welded and installed. I clamped wood blocks down the main beams to hold the rack in position, before adhering it to the rail with VHB tape.
x_rack_installation.jpeg

I test-fitted my 30T pinions, and it looks like I'll have a bit of interference with the end of the gantry. I'll prop up the gantry, one end at a time, and grind a small recess to improve clearance a bit.
pinion_gantry_interference.jpg

[Gerald D]

Is your rail 28mm high? Is the tape 2mm thick? Is the rack 0.5" high? Are you using the C size v-rollers? And your gantry/car end plates are the standard laser cutting?

If all of those correct, then there should be a small clearance....or else many other people have not mentioned the problem.

[melissa]

Yes, all of those things are correct.

I just measured my pinions, the hub is about 34.5mm diameter, with a wall thickness around 10.5mm.

It may be that there was actually a small clearance, but I was worried that the pinion might not seat fully in the rack, in the event that it touched the gantry. It was a simple enough adjustment with the angle grinder, and a bit of touch-up paint.

[Gerald D]

Melissa, how old are your laser cuttings? I suspect you have used the "classic" parts and not the later "Mamba" parts.

Taking part 10 20 451 P as an example, the vertical distance from the center of the roller down to the bottom edge of the plate was 62mm originally. I became aware of the potential interference problem with 0.5" rack (compared to the 16mm the metric guys use) and thus modified the plate profile to a dimension of 53mm, which should give everyone plenty of clearance now. (I think that even the 3/8" rack will work now)

[Gerald D]

PS. The hub diameter of the pinion is not a factor. The hubs can go as big as the root diameter of the gear teeth. Basically that means the the gantry metalwork shouldn't hang below the tips of the rack teeth.

[domino11]

Gerald,
His cuttings are the latest Mamba Set. I never cut any of the classic parts.

[Gerald D]

Originally Posted by Gerald D

Melissa, how old are your laser cuttings? I suspect you have used the "classic" parts and not the later "Mamba" parts.

Taking part 10 20 451 P as an example, the vertical distance from the center of the roller down to the bottom edge of the plate was 62mm originally. I became aware of the potential interference problem with 0.5" rack (compared to the 16mm the metric guys use) and thus modified the plate profile to a dimension of 53mm, which should give everyone plenty of clearance now. (I think that even the 3/8" rack will work now)

I am home at my PC now and see that I was talking nonsense above

The distance between bearing center and bottom edge of plate is 61.8mm. In theory, there is only 0.4mm clearance with 1/2" rack, 2mm tape and 28mm rail height.

[melissa]

Originally Posted by Gerald D

Basically that means the the gantry metalwork shouldn't hang below the tips of the rack teeth.

Indeed it doesn't. I had a closer look today, it appears the bottom of 1020451 is exactly even with the bottom of the rack.

That puts the hub closer to 1020451 than I was comfortable with, hence the grinding now. I didn't want to have the machine already running, and then find out there was a problem.

[Gerald D]

Grinding a bit away is fine for you, but the drawing should be changed so that nobody has to grind.

[camilleg]

Better grinding the gantry than grinding of, um, teeth