There have been a number of threads about table flatness. Some have indicated that sag in the support rods is really the problem. It certainly appears that is the case.
I've set up my table so the 4 corners are within .002 of flat, and a straight edge shows less than .004" variation from flat over the table. However, with a dial indicator on the carriage I measure about .036 high in the center.
To further verify that this is the result of sag in the support rods, I made a frame of 2x4s on edge measuring 24" square and clamped it to the table. Cross members divided it into 4 smaller squares. It was glued up on the top of my cabinet saw so the bottom side was flat. I clamped it to the table and made a couple of skim cuts over the entire frame, with the last cut being .015".
I then used a straight edge and feeler gauges to check the flatness of the resulting surface. In the middle, it was about .040" low, assuming the 4 corners are in a flat plane. At the center along each of the edges it was about .020" low.
Conclusion: The router drops about .040" in the center of the table relative to its position in the corners. For the least variation from true flat movement over a cut surface, this means the best location for the material will be at the center of the table with X0, Y0 at the center of the piece.
If depth of cut accuracy is extremely critical, one could either take a skim cut on the material first, or use Aspire to put the cut on a compensating surface.
Table flatness is NOT the problem.
Moderators: al wolford, sbk, Bob, Kayvon
Table flatness is NOT the problem.
- Attachments
-
Re: Table flatness is NOT the problem.
I have just taken my Shark apart for the umptieth time, this time to replace all plastic parts with alloy.
My HD Shark is less than 12 months old.
I had noticed the router was flexing from side to side when it was cutting for example a circle.
This is what I found in the sides of the gantry
Despite the 1/4" reinforcement plate at the bottom part and the U channel just a bit higher, the sides are bent.
Just sitting on a flat surface it show a bend of 5/8". The U channels are bent as well.
It beats me why NWA didn't reinforce the side all the way
The top of the gantry also is reinforced with U channels but they do not help much when when screws are at the edge of the plastic instead of in the middle of it
Wolffie
My HD Shark is less than 12 months old.
I had noticed the router was flexing from side to side when it was cutting for example a circle.
This is what I found in the sides of the gantry
Just sitting on a flat surface it show a bend of 5/8". The U channels are bent as well.
It beats me why NWA didn't reinforce the side all the way
The top of the gantry also is reinforced with U channels but they do not help much when when screws are at the edge of the plastic instead of in the middle of it
Wolffie
Re: Table flatness is NOT the problem.
I have not even taken mine apart yet. this throws a wrench into it. I was planning on it. let me know what you doWolffie1 wrote:I have just taken my Shark apart for the umptieth time, this time to replace all plastic parts with alloy.
My HD Shark is less than 12 months old.
I had noticed the router was flexing from side to side when it was cutting for example a circle.
This is what I found in the sides of the gantry Despite the 1/4" reinforcement plate at the bottom part and the U channel just a bit higher, the sides are bent.
Just sitting on a flat surface it show a bend of 5/8". The U channels are bent as well.
It beats me why NWA didn't reinforce the side all the way
The top of the gantry also is reinforced with U channels but they do not help much when when screws are at the edge of the plastic instead of in the middle of it
Wolffie
Re: Table flatness is NOT the problem.
If you are working with the pro or pro plus there is an issue of bearing rod flex that causes the z axis to sag down in the center of its travel. This is primarily due to the use of unsupported bearing rods in both the x and y axis. The end result is the router sags low in the center and no amount of fiddling with the table, or even surfacing it with the shark, will get it right. If you surface plane a sacrificial top, it will be dished down in the center and your work will not conform to it unless you pull it down with vacuum. (not practical)
A solution is to convert your machine to fully supported bearings as I did. The files for the parts required are available for free, just PM me and I will get them to you.
The process is documented by Winreboot, who did the x axis conversion and was kind enough to post pictures.
Here is the thread: http://www.cncsharktalk.com/viewtopic.php?f=9&t=1342
The cost runs about 300 dollars per axis , for the x or the Y. The x being the more critical.
I have received feedback from several folks that did the conversion and the seem to be happy with the result, but I would rather they speak for their own experience with it.
The HD models try to solve the issue by supporting the rods in the center. This is not as rigid as a fully supported situation which binds the x bering rods to the backplate along its entire length, forming a "C" beam type of arrangement which resists sag and rotation.
A lot of good info can get buried in the forum because of the ever growing volume of posts. Most of these issues had been experienced by others already and a lot of good info is here if you dig for it.
Hope this helps.
Bob
A solution is to convert your machine to fully supported bearings as I did. The files for the parts required are available for free, just PM me and I will get them to you.
The process is documented by Winreboot, who did the x axis conversion and was kind enough to post pictures.
Here is the thread: http://www.cncsharktalk.com/viewtopic.php?f=9&t=1342
The cost runs about 300 dollars per axis , for the x or the Y. The x being the more critical.
I have received feedback from several folks that did the conversion and the seem to be happy with the result, but I would rather they speak for their own experience with it.
The HD models try to solve the issue by supporting the rods in the center. This is not as rigid as a fully supported situation which binds the x bering rods to the backplate along its entire length, forming a "C" beam type of arrangement which resists sag and rotation.
A lot of good info can get buried in the forum because of the ever growing volume of posts. Most of these issues had been experienced by others already and a lot of good info is here if you dig for it.
Hope this helps.
Bob
Re: Table flatness is NOT the problem.
FixitMike
I beg to differ with you. And I would also like to say that the shark MUST be on a flat LEVEL base. It must be level and supported in all 4 cornes and also in the outside center under the Y rails. With the shark properly leveled and supported you should be able to get a good reading in all four couner and also on the sides. You will still will most likly have a bad reading (as your post shows) in the very center of the table because the gantry leans forward when the router is in the center. The following link is how I went about setting up my Shark Pro HD. Belive me I spent some time to under stand all the "issue" and how to correct them with out rebuilding my machine.
http://www.cncsharktalk.com/viewtopic.php?f=15&t=2297
The Y rails must be level and flat end to end. The X rail will lean forward when the router is in the center of the table BOTH can be delt with.
And if you still have the HDPE support holding the deck You will also have a problem.
Take the time to look at this It should help you and save a lot of time tearing your hair out. You need to address all the issue and understand how to fix them one at a time.
I beg to differ with you. And I would also like to say that the shark MUST be on a flat LEVEL base. It must be level and supported in all 4 cornes and also in the outside center under the Y rails. With the shark properly leveled and supported you should be able to get a good reading in all four couner and also on the sides. You will still will most likly have a bad reading (as your post shows) in the very center of the table because the gantry leans forward when the router is in the center. The following link is how I went about setting up my Shark Pro HD. Belive me I spent some time to under stand all the "issue" and how to correct them with out rebuilding my machine.
http://www.cncsharktalk.com/viewtopic.php?f=15&t=2297
The Y rails must be level and flat end to end. The X rail will lean forward when the router is in the center of the table BOTH can be delt with.
And if you still have the HDPE support holding the deck You will also have a problem.
Take the time to look at this It should help you and save a lot of time tearing your hair out. You need to address all the issue and understand how to fix them one at a time.
drueth
Shark Pro Plus HD
new to CNC 12/2012
Shark Pro Plus HD
new to CNC 12/2012
Re: Table flatness is NOT the problem.
I think FixitMike has a non HD Machine (as this was posted in the pro plus area.)
So my comments were for the non HD machines.
The HD machines differ greatly in how the bearing rods are supported compared to the older pro and pro plus.
In the pro and plus the bearing rods just float in the center and so they flex a good deal under the weight of the machine.
In the HD they are center supported, but if the center supports are not exactly right they can pull the rod out of straight and cause a different set of issues. The HD series rods are more rigid, but not necessarily perfectly straight.
I think drueth is talking about the HD machines. Although a flat surface to place the machine on is desirable / necessary for all models.
In the HD models, the Y axis rods are clamped down in the center, but if they are too high or low in the center compared to the ends, the gantry will tilt forward or back. This would have to do with the accuracy of the machining of the plastic parts when manufactured as well as whether the machine was fastened down to a flat table top.
With the non HD versions, the rods are only held only at the ends and should be naturally straight with no load, but they sag out of straight with weight applied to them. The sag/flex issue is much more severe in the non HD machines, and the solutions / corrections are different, depending on what model you have.
Bob
So my comments were for the non HD machines.
The HD machines differ greatly in how the bearing rods are supported compared to the older pro and pro plus.
In the pro and plus the bearing rods just float in the center and so they flex a good deal under the weight of the machine.
In the HD they are center supported, but if the center supports are not exactly right they can pull the rod out of straight and cause a different set of issues. The HD series rods are more rigid, but not necessarily perfectly straight.
I think drueth is talking about the HD machines. Although a flat surface to place the machine on is desirable / necessary for all models.
In the HD models, the Y axis rods are clamped down in the center, but if they are too high or low in the center compared to the ends, the gantry will tilt forward or back. This would have to do with the accuracy of the machining of the plastic parts when manufactured as well as whether the machine was fastened down to a flat table top.
With the non HD versions, the rods are only held only at the ends and should be naturally straight with no load, but they sag out of straight with weight applied to them. The sag/flex issue is much more severe in the non HD machines, and the solutions / corrections are different, depending on what model you have.
Bob
Re: Table flatness is NOT the problem.
I originally posted this here: http://www.cncsharktalk.com/viewtopic.p ... 0&start=10
This is my calculation for the Pro plus. As you can see, the bulk of the deflection comes from the Y axis support rods.
This is my calculation for the Pro plus. As you can see, the bulk of the deflection comes from the Y axis support rods.
BillK wrote:Thanks for the responses and your findings.
As I previously stated, I am a Mechanical Engineer, so I took some measurements, made some assumptions, and did some calculations. Here's what I got.
Measurements: The two round steel bars that go across the gantry (X direction) are 31 inches long and are 20mm (0.788 inch) diameter. The two round bars in the base are 29 inches long and also 20mm diameter.
Assumptions: The Gantry package weighed in at 48 pounds. Subtracting out the packaging, control box, cables, software boxes, clamps, etc. Assume the Gantry assembly with router installed weighs 40 pounds. Assume the router carriage with motor, shafting, router, etc. weighs 10 pounds. Assume the threaded rod carrys no load. Assume at the center both steel shafts share the load equally.
I did standard deflection calculation for the shafts with the assumed loads and under their own distributed load.
So here is what you should get, or close to it.
Across the front or back in the X direction a perfectly flat table you will see a rise in your indicator of 0.0084", which is really the deflection of the bars at the center. As previously reported, I indicated .010" after making the install of the aluminum square and adding the brass shim.
At the very center of the table, due to deflections of the four steel shafts under load, you should see a maximum deflection of 0.0285" if your table is perfectly flat.
Remember, the aluminum track will also deflect, so Mike's measurement of 0.025" at the middle completely corresponds to this.
If your table then maps out at about 0.030", there is nothing you can do to get it flatter, it is flat, your seeing normal deflection. If your table shows .080" like Mike's or .090" like mine you can get it to .030" making the improvements like we did.
Two things that did my table in were the use of the plastic clamping parts as previously reported, and what appears to be a machining error that made the right side of the table about 0.035 too low. I think Next Wave can control this better and deliver tables that consistently map out to about 0.040" total which would give them a reasonable tolerance.
Thanks.
BillK
Re: Table flatness is NOT the problem.
sacrificial top. what is the best material to use on the shark hd. I want a 24x24 maybe 1 inch
Re: Table flatness is NOT the problem.
When I looked into this, I found a lot of folks using a good MDF. I've been using MDF for a few years and am overall very pleased with it, and the flexibility/ease in jig configuration by having an MDF base. I drill holes and use inserts to rapidly position clamps/jigs. It's MDF and sacrificial, so no worries about drilling where I need it.
As my shop does not have HVAC - central heating and mother nature's cooling, I do have to pay attention to swelling sometimes with high humidity (can get into the 0.01 or more range). A quick run to 'flatten' the table cures this for me. I still use shims for work that requires greater than 0.03 accuracy (ie lithographs).
I did attach it using plastic stove bolts, so I don't have to be concerned about running into one of the scarf board hold downs when doing a cut through or re-planing of the entire surface. I have learned to sweep over the board using a metal detector to ensure I didn't leave one of the inserts in the table that I'm not using a clamp with...
As my shop does not have HVAC - central heating and mother nature's cooling, I do have to pay attention to swelling sometimes with high humidity (can get into the 0.01 or more range). A quick run to 'flatten' the table cures this for me. I still use shims for work that requires greater than 0.03 accuracy (ie lithographs).
I did attach it using plastic stove bolts, so I don't have to be concerned about running into one of the scarf board hold downs when doing a cut through or re-planing of the entire surface. I have learned to sweep over the board using a metal detector to ensure I didn't leave one of the inserts in the table that I'm not using a clamp with...
Re: Table flatness is NOT the problem.
I sort of like that ideajeb2cav wrote:When I looked into this, I found a lot of folks using a good MDF. I've been using MDF for a few years and am overall very pleased with it, and the flexibility/ease in jig configuration by having an MDF base. I drill holes and use inserts to rapidly position clamps/jigs. It's MDF and sacrificial, so no worries about drilling where I need it.
As my shop does not have HVAC - central heating and mother nature's cooling, I do have to pay attention to swelling sometimes with high humidity (can get into the 0.01 or more range). A quick run to 'flatten' the table cures this for me. I still use shims for work that requires greater than 0.03 accuracy (ie lithographs).
I did attach it using plastic stove bolts, so I don't have to be concerned about running into one of the scarf board hold downs when doing a cut through or re-planing of the entire surface. I have learned to sweep over the board using a metal detector to ensure I didn't leave one of the inserts in the table that I'm not using a clamp with...