re:Levelling large table
Posted by
ballendo@y...
on 2000-12-22 16:57:32 UTC
Doug,
Which do you have more of; time? or money?
Both?!? See Smokes' reply...
You may find that .625 cast alum has a 'tighter' tolerance for
flatness (from the mfr) than .500 . So using it will get you closer
to start with.
What level of 'dead nuts' are you REALLY after? do you mean to accept
the basic flatness of the plate alum and just hope not to inroduce
ADDITIONAL inaccuracies? Or do you REALLY mean 'DEAD NUTS FLAT'? If
so, you're back to Smokes' reply , or the use of a flycut spoilboard
or table.
Surely you realise that a half-inch thick 60x120 plate will not be
anywhere NEAR flat.
Is this a bridge or Gantry machine? Big difference in the solutions
available!
You will still have to assure that the spindle axis IS perpendicular
to X and Y, but 'floating' a dial indicator over the machine travel
area (while held in the Z axis) will allow you to adjust
jackscrews... Don't need a collimator.
3)Agreed. But it does provide the quickest path and allows for
construction variations. This is why I asked if you had more time or
more money.
Essentially, you are looking for a way to 'set up' a planar surface.
Whether it's a bridge or a gantry, you will need to attain a 'flat'
table. Boatbuilders and masons have used string or wire
for 'straight', which is the FIRST part of FLAT! Thin wire stretched
across the table at many points (held above the surface by feeler
gauge stock thick enough to 'bridge' the high points) will give you a
reference to get the table parts straight. It won't be flat yet...
If a gantry-style machine you can then 'shim' the table into
alignment and then fasten it the frame to hold the alignment, using
the indicator in the Z to check and assure your results...
In a bridge design, the table will have to be kept flat within itself
or by the use of a sub structure which also moves with the table. I
HAVE seen many cases where a $20k+ machine relied on the flatness
of .500 tooling plate to define and support the machine geometry,
usually with some help from a small welded steel frame! This can
work, but WILL not be (or stay) FLAT! Other times , ONLY
the 'bearings' or 'rails' (of the linear bearing system) were bolted
to the .500 plate! IT was 'supposed' to be flat, right?!? NOT!!!
Hope this helps.
Ballendo
P.S. BTW, a 'leveling' jackscrew 'should' be designed as an
ADJUSTMENT, NOT a SUPPORT (IMO). For example: You have a welded tube
base with angle iron welded to the base assembly to support the
table. Slots are milled in the 'standing' leg of the angle to allow
for adjustment with a second angle iron which will actually SUPPORT
the table. This second angle has threaded holes which match the first
angle iron slots... Drilled and tapped holes are put into the base
tubine 'next to' the angle iron such that the heads of bolts will
press against the upper angle; adjusting it. ONCE the adjustment is
done, the bolts in the slotted holes are tightened. This means the
load of the table is CARRIED by the 2 angle iron pieces, and only
SUPPORTED by the jack screws...
|____________________Table_______
|______ |
[ ]| |_
| | | | |
| | |-|-|-Bolts thru slots here
Jackscrew-| | |_| |______
____| |___|________|_______Base______
| | | |
| |-tapped hole in base
This Ascii Art is basic, but hopefully will get the idea across. In
reality the angle might be much thicker in relation to the jackscrew.
And the jackscrew might act on/under the 'leg' of the upper angle,
rather than the 'flat'
The point is that the jackscrew should not be relied upon for
SUPPORT, IMO.
P.S.S. There is a man (company?) selling plans for a large router.
The plans are not cheap, and the machine is expected to cost nearly
$20-40k (depends on size/heads/etc.) BUT, he is comparing this to the
cost of a commercial CNC machine of similar capabilities which will
start at about $40K, making this approach viable for some. Sorry, I
don't have the url just now... Looked well executed,IMO.
Which do you have more of; time? or money?
Both?!? See Smokes' reply...
You may find that .625 cast alum has a 'tighter' tolerance for
flatness (from the mfr) than .500 . So using it will get you closer
to start with.
What level of 'dead nuts' are you REALLY after? do you mean to accept
the basic flatness of the plate alum and just hope not to inroduce
ADDITIONAL inaccuracies? Or do you REALLY mean 'DEAD NUTS FLAT'? If
so, you're back to Smokes' reply , or the use of a flycut spoilboard
or table.
Surely you realise that a half-inch thick 60x120 plate will not be
anywhere NEAR flat.
Is this a bridge or Gantry machine? Big difference in the solutions
available!
>We are starting to draw up our router table and need a way to get1)2) Really wish I knew if this was a bridge or gantry?
>the top dead-nuts level. The plan is to use 60"X120"X.5" cast
>aluminum tool plate for the table and a square tubing steel weldment
>for the base.
You will still have to assure that the spindle axis IS perpendicular
to X and Y, but 'floating' a dial indicator over the machine travel
area (while held in the Z axis) will allow you to adjust
jackscrews... Don't need a collimator.
3)Agreed. But it does provide the quickest path and allows for
construction variations. This is why I asked if you had more time or
more money.
Essentially, you are looking for a way to 'set up' a planar surface.
Whether it's a bridge or a gantry, you will need to attain a 'flat'
table. Boatbuilders and masons have used string or wire
for 'straight', which is the FIRST part of FLAT! Thin wire stretched
across the table at many points (held above the surface by feeler
gauge stock thick enough to 'bridge' the high points) will give you a
reference to get the table parts straight. It won't be flat yet...
If a gantry-style machine you can then 'shim' the table into
alignment and then fasten it the frame to hold the alignment, using
the indicator in the Z to check and assure your results...
In a bridge design, the table will have to be kept flat within itself
or by the use of a sub structure which also moves with the table. I
HAVE seen many cases where a $20k+ machine relied on the flatness
of .500 tooling plate to define and support the machine geometry,
usually with some help from a small welded steel frame! This can
work, but WILL not be (or stay) FLAT! Other times , ONLY
the 'bearings' or 'rails' (of the linear bearing system) were bolted
to the .500 plate! IT was 'supposed' to be flat, right?!? NOT!!!
Hope this helps.
Ballendo
P.S. BTW, a 'leveling' jackscrew 'should' be designed as an
ADJUSTMENT, NOT a SUPPORT (IMO). For example: You have a welded tube
base with angle iron welded to the base assembly to support the
table. Slots are milled in the 'standing' leg of the angle to allow
for adjustment with a second angle iron which will actually SUPPORT
the table. This second angle has threaded holes which match the first
angle iron slots... Drilled and tapped holes are put into the base
tubine 'next to' the angle iron such that the heads of bolts will
press against the upper angle; adjusting it. ONCE the adjustment is
done, the bolts in the slotted holes are tightened. This means the
load of the table is CARRIED by the 2 angle iron pieces, and only
SUPPORTED by the jack screws...
|____________________Table_______
|______ |
[ ]| |_
| | | | |
| | |-|-|-Bolts thru slots here
Jackscrew-| | |_| |______
____| |___|________|_______Base______
| | | |
| |-tapped hole in base
This Ascii Art is basic, but hopefully will get the idea across. In
reality the angle might be much thicker in relation to the jackscrew.
And the jackscrew might act on/under the 'leg' of the upper angle,
rather than the 'flat'
The point is that the jackscrew should not be relied upon for
SUPPORT, IMO.
P.S.S. There is a man (company?) selling plans for a large router.
The plans are not cheap, and the machine is expected to cost nearly
$20-40k (depends on size/heads/etc.) BUT, he is comparing this to the
cost of a commercial CNC machine of similar capabilities which will
start at about $40K, making this approach viable for some. Sorry, I
don't have the url just now... Looked well executed,IMO.
Discussion Thread
Doug Harrison
2000-12-22 13:07:25 UTC
Levelling large table
Bob Campbell
2000-12-22 13:35:18 UTC
Re: [CAD_CAM_EDM_DRO] Levelling large table
Smoke
2000-12-22 14:13:59 UTC
Re: [CAD_CAM_EDM_DRO] Levelling large table
ballendo@y...
2000-12-22 16:57:32 UTC
re:Levelling large table
R. T. Robbins
2000-12-22 17:41:33 UTC
Re: [CAD_CAM_EDM_DRO] Levelling large table
Doug Harrison
2000-12-22 18:25:33 UTC
Re: [CAD_CAM_EDM_DRO] re:Levelling large table
Smoke
2000-12-22 20:57:15 UTC
Re: [CAD_CAM_EDM_DRO] re:Levelling large table
dave engvall
2000-12-28 18:20:31 UTC
Re: [CAD_CAM_EDM_DRO] Levelling large table
Joe Fahy
2001-01-14 13:57:03 UTC
Re: [CAD_CAM_EDM_DRO] re:Levelling large table
Doug Harrison
2001-01-15 10:32:52 UTC
Re: [CAD_CAM_EDM_DRO] re:Levelling large table
Joe Fahy
2001-01-15 12:46:15 UTC
Re: [CAD_CAM_EDM_DRO] re:Levelling large table
kem@b...
2001-01-16 05:44:31 UTC
Re: re:Levelling large table
Joe Fahy
2001-01-16 12:44:45 UTC
Re: [CAD_CAM_EDM_DRO] Re: re:Levelling large table