Needs Pictures: 0
Picture(s) thanks: 0
Results 1 to 15 of 19
-
19th August 2020, 06:07 PM #1SENIOR MEMBER
- Join Date
- Aug 2006
- Location
- Canberra - West Belco
- Age
- 63
- Posts
- 646
Another Hi Tech Router table - project
This hobby - that hobby ... far too many hobbies but back into the woodworking one for the time being and it's new Router Table time after reading the two high tech threads and a build 3D printer experience the table will be similar to Mikes in this thread by MandJ Automated Router table
along with Chris's work in this thread New Router Table Build
Thanks to both for inspiration and hopefully I can get hold of the firmware being used.
So at the planning stage about all i have decided is:
Use a 2.2kw spindle so i can have an ER20 collet set allowing for the use if 1/4 and 1/2 bits
Initial mounting in one end of a new outfeed mobile table. This may change
use support linear rails either SBR20 or SBR25 for the Z axis Spindle movement, likely use double bearing blocks on the top if i can source with a rail set.
Tempted to use MGN rails for the fence axis but likely just go back to SBR due to cost
I have spare Nema23 motors for the fence that will be strong enough but may well source a long 23 or 34 for the Z due to the overall weight of spindle and mount
Basic idea at the moment is to build the Z axis spindle setup independent of the cabinet/table allowing for fraction/fine 90 deg adjustment to a table top.... plenty of ideas out of the 3D build world to make this work.
May well do a bit of this in 4040 extrusion as i'll be using it for the big CoreXY 3D printer project ... Hey Alu is close to wood in that you can use similar tools
Now the help bit from people who already have spindles in use as i need to decide on water v air cooled.
Immediate decision point:
I'm a part time woodworker and can even go years without picking up tools, the existing router table setup has not been used since we moved into this house 13 years ago.
To me this rules out water cooled so I don't have a maintenance overhead.
Air Cooled is noisy but it seems the noise is swamped by cutting noise and why else would i have it on.
I live in Canberra and it gets hot in summer plus for water cooled and my garage can freeze in water, a negative if i don't drain the water
So water or air cooled spindle? ... I'm leaning solidly towards air at the moment
I'm open to all thoughts and ideas?
I'll be doing key dimensional design work in Fusion 360 but may not draw in things like the stepper motors as i have no idea on how to apart from a basic block for them.
-
19th August 2020 06:07 PM # ADSGoogle Adsense Advertisement
- Join Date
- Always
- Location
- Advertising world
- Age
- 2010
- Posts
- Many
-
21st August 2020, 11:08 AM #2SENIOR MEMBER
- Join Date
- Aug 2006
- Location
- Canberra - West Belco
- Age
- 63
- Posts
- 646
I'm going to do some rambling that hopefully helps people decide just how complex they want their router table
Why even go down the spindle motor path and it's complexities when a router bolted to even a simple plate gives a nice square accurate surface.
It's unlikely to be cheaper for a basic setup but the spindle path should be significant cheaper for a high accuracy adjustable system, i'll crunch some numbers after i start ordering parts.
I already have a simple Alu plate based setup mounted in a 60mm odd bench similar to the drawing below. pardon my crude Fusion skills.
This provides a simple aligned cutter to the table assuming your plate itself is flat.
As the router bolts to the plate, it effectively self aligns on the top plane (xy) assuming the plate and router base are flat and the router plate is a consistent thickness
This puts z at right angles to the xy plane, thinking in terms of the planes becomes important when we go to do an independently mounted system such as a spindle motor
All of this applies the same with the commercial router lifts that relate the lift mechanism to the plate bottom surface
it does NOT apply to diy lift solutions that decouple the router base from the table or plate as we then have different reference surfaces to consider and the potential for the vertical plane not to be at right angles to the horizontal planes.
For simplicity i haven't drawn or considered the fence as it's alignment considerations are the same for both solutions with it's reference surface being the table top (xy plane)
RT basic.PNG
This is the simple accurate solution but has the major drawback in it's simple form of requiring undertable access for both lift/height adjustment and bit changes. An above table lift fixes that but adds expense.
It all becomes a dollar/complexity tradeoff
Next post, consideration of and another crappy fusion drawing to look at the spindle body setup
-
21st August 2020, 11:43 AM #3Woodworking mechanic
- Join Date
- Jan 2014
- Location
- Sydney Upper North Shore
- Posts
- 4,464
My router, mounted to a moveable base plate and driven by a stepper can give highly accurate settings. For what I need, I can’t justify the cost of a spindle, VFD, etc when the router will give me the same accuracy although I can see benefits of a water cooled spindle setup.
Mine is still in prototype phase as I haven’t bothered to proceed to a full table build, but I probably will early next year.
-
21st August 2020, 12:45 PM #4SENIOR MEMBER
- Join Date
- Aug 2006
- Location
- Canberra - West Belco
- Age
- 63
- Posts
- 646
I think that for people who already have a router then the decision becomes a lot harder, i personally have 3 that are 1500w or above 1/2" machines
The issue i was pointing out above and setting up for the next post is that even stepper driven z lift whilst highly accurate on the height adjustment could be aligned badly if not bolted/referenced from the top table surface, the simple solution and the commercial router lifts all provide intrinsic alignment in three planes
-
21st August 2020, 02:56 PM #5GOLD MEMBER
- Join Date
- Jun 2005
- Location
- Helensburgh
- Posts
- 7,696
As Mike and I got further into the build the obvious advantage of the spindle not attached to the top became obvious. The biggest issue with a conventional router and lift is weight and making the top strong enough so it does not bend due to the weight, the second advantage is that if the top is free to be lifted off then it can be light and custom tops made to suit certain jobs if needed. If the fence is parked clear of the top it would nearly be quicker to lift the top off to do a bit change....maybe.
CHRIS
-
21st August 2020, 03:08 PM #6Woodworking mechanic
- Join Date
- Jan 2014
- Location
- Sydney Upper North Shore
- Posts
- 4,464
As I said above, there are benefits in using a spindle.
One of the advantages of using a router could be that you have it bolted to the table top, as it usually is, and use the stepper to move it on it’s normal posts. That way it is exactly square to the table top. I am looking at this aspect at the moment.
-
21st August 2020, 03:55 PM #7SENIOR MEMBER
- Join Date
- Aug 2006
- Location
- Canberra - West Belco
- Age
- 63
- Posts
- 646
That's a couple good points for the frame attached device .... we are actually in the same analysis space as the contractor v cabinet saw mounting of saw components. Mount it to the table top and deal with the weight and potential dishing or mount to the frame and keep a flatter table due to lack of hanging weight.
I'll explore this a little more when i have a diagram to talk to as this also adds the adjustment requirements to get all three planes aligned.
Sorry did actually miss that in the bolted to the table approach above and is a sort of hybrid option. Not sure about the more expensive routers but having my old GMC Mag2050R apart on the bench at the moment the posts are hardly precision items and whilst i never noticed it in normal use there is enough difference even in the internal dimensions to question just how much accuracy can be achieved .. mind you it is wood and likely we will never notice.
I'll be doing a manual lift on the plate mounted GMC this weekend if the mailman delivers some nutserts today, the bottom plate is set up in the mill now to remove pin from the shaft to allow some all thread to go in ... i really should take some photo's
-
21st August 2020, 06:21 PM #8Woodworking mechanic
- Join Date
- Jan 2014
- Location
- Sydney Upper North Shore
- Posts
- 4,464
That is really the crux of the matter- what sort of accuracy are we trying to achieve when routing timber? Do we need to achieve an accuracy of 1mm, 0.5mm 0.2mm, 0.02mm..etc?
Its not that I’m saying spindles are not one way to go but I wonder why we seek such accuracy - is it because we need it or is it because we can say we achieved it?
I’ll probably be adapting an old infrequently used Makita 3620 router that I’ve had for over 30 years, I’ll so some measurements re side play etc unlocked and locked before starting to get some idea of what accuracy I may be able to achieve.
-
22nd August 2020, 12:10 PM #9Member
- Join Date
- Apr 2013
- Location
- Mornington Peninsula
- Posts
- 70
To quote youtuber Jeremy Schmidt - “if you can’t make it accurate, make it adjustable.”
I spent a few years thinking about how to do this and this is my thought process when I finally pulled the trigger...
Made a metal frame. After welding, it was nominally rectangular, but out of square and bowed (which is unavoidable). The advantage of steel is that it will stay that way.
With the frame more or less level, mounted the spindle using a precision level - the important thing here is that the frame is not the reference, but the level (the earth). The spindle can be adjusted with slot screws and shims if needed be for final level or later adjustments.
Mounted top on frame (which is bowed and warped by welding). Then levelled the top (to the ‘earth’ reference, not the spindle) and straightened on the sides and diagonals with a straight edge using shims.
This way, I can adjust the spindle or the top independently, and because the top is level, it is coplanar with the benches next to it. If I take the top off, and replace it, I only have to worry about getting the top sorted and don’t have to touch the spindle.
With respect to accuracy, it depends so much on the application. Using a round over bit, the bit height is important but then a slight lip can be taken off with a couple of swipes with sandpaper. The spindle perpendicularity really doesn’t matter unless is way put. A perfect dado in the morning might be too tight the next day after rain. If the workpiece is twisted and bowed (which it always is anyway), then it has to conform to the table top anyway, whether it’s dead flat or not.
So I guess my point is not to overthink the (static) geometrical accuracy of the whole setup.
I believe the dynamic rigidity is a seperate and more important issue. If the spindle deflects under load, you’ll get unpredictable results depending on loading with chatter, poor surface finish, greater likely hood of tear out etc. that’s why I chose to go with the ‘overkill’ z axis linear solution. The geometry can be adjusted, but the rigidity can’t.
With regards to water vs air cooled. I vote for water. Lots of advantages. I don’t see the maintenance issues. My cooling pump sits in small Bunnings plastic storage bin with a couple of holes punched in the top. If I don’t want the water sitting there (maybe I’m not using it for a while), I can chuck it out and then put a couple of litres back in when I want to use it again. You don’t have to use fancy coolant.
Leigh
Sent from my iPad using Tapatalk
-
26th August 2020, 10:57 AM #10SENIOR MEMBER
- Join Date
- Aug 2006
- Location
- Canberra - West Belco
- Age
- 63
- Posts
- 646
Small update, the ST Nucleo-144 turned up yesterday, this board is the heart of the MandJ software solution and i need to do some learning on it and source the appropriate touch display and do the interface pcb or as referred to on the CNC forums the BoB (Break out Board). Small slow steps as i juggle projects.
Working on an above table adjustment (manual) solution for the old GMC at the moment so i get a table mounted router back in operation.
Some comments on Leigh's post above ... i had carefully quoted the sections i wanted but can't work out how to add the partial quotes in the forum software....
if you can’t make it accurate, make it adjustable.
but out of square and bowed (which is unavoidable)
the important thing here is that the frame is not the reference, but the level (the earth). The spindle can be adjusted with slot screws and shims if needed be for final level or later adjustments.
For ease i would not use the earth level as the reference as it adds a complete step to effective level a surface to earth that is not actually required in 3D space, just pick a reference plane and work from both a single corner point for dimensions and from the plane for all 90 degree angles.
Once the top is ultimately aligned to your reference frame then it becomes a simple matter of shimming legs to make it level to surrounding bench tops.
When you inserting into an existing benchtop then the benchtop becomes the reference plane by default.
The spindle perpendicularity really doesn’t matter unless is way put.
My 3D printer with it's extruded plastic still achieves very high repeatability with better than 0.1mm dimensional accuracy when i take care with the filament calibration.
I believe the dynamic rigidity is a seperate and more important issue. If the spindle deflects under load, you’ll get unpredictable results
At least in all this my garage reorganisation is still progressing and i might end up with some more working space soon.
-
26th August 2020, 01:16 PM #11GOLD MEMBER
- Join Date
- Jun 2005
- Location
- Helensburgh
- Posts
- 7,696
My reference to set the spindle will be the top mounted on the frame. Make a T that can be inserted into the chuck and the cross leg the same length as the table and measure from the table top at the ends of the T. The spndle will be mounted to allow adjustment in both axis. Not getting it close to right will affect the profile the cutter makes on the timber.
The manal version we built has absolutely no deflection that can be measured or seen and it gets used on way bigger stuff that the average hobbyist would do, if you are worried about deflection simply add another bracket to the body of the spindle. I think we used 18mm rails on that one but I would have to check. There is no way on earth that enough load could be applied on a router cutter that would result in deflection using 18mm rails. I am only putting this out there as my experience to assist any decisions that need to be made.CHRIS
-
26th August 2020, 04:40 PM #12SENIOR MEMBER
- Join Date
- Aug 2006
- Location
- Canberra - West Belco
- Age
- 63
- Posts
- 646
Effectively the same as tramming a mill bed and column only upside down . i am struggling to picture simple adjustment mechanisms at the moment that are simple and rigid when locked.... open to all suggestions. My view is no sense building a battleship if people can't replicate it.
Yeah i really didn't articulate myself well on deflection v rigid. Deflection of the round body spindles at 80mm in the bulky looking bracket is considered a non issue in the CNC world that will have higher loads than we are likely to achieve on a router table and any deflection on a round body bracket compared to a full length square body mount is non existent in the real world... but ... that square body mount may be more rigid from a harmonics vibration leading to chatter ... again very much in a maybe world but even the short time with the mini vertical mill has taught me that every little bit of rigidity that you add is a good thing.
TLDR: for most considering a spindle the average round water cooled option is likely the best support choice these days
-
27th August 2020, 12:01 PM #13SENIOR MEMBER
- Join Date
- Aug 2006
- Location
- Canberra - West Belco
- Age
- 63
- Posts
- 646
Do we need DRO scales on a Stepper Driven RT?
So this is a genuine question after reading a lot. I come from a 3D printer design angle where you home the axis and work from a known 0,0,0 point in 3D space just using step count. There has been considerable testing for lost steps, position accuracy etc. 3D printers have a much higher demand on positional repeatability than any router table will and they seem to get it right by just homing at the beginning of the print that might last for days at a time. CNC basically does the same thing, you set your 0,0,0 point and all movement is relative based on step count/pulses.
I can see how Mike originally used the DRO scales to get the positions in 3D space for the manual table and how that got carried forward into the software driven stepper version of the table but a clean slate would allow us to drop the DRO scale .... not even vaguely sure yet if the software can be adapted to this or it's a clean software slate as well.
On a RT we need to zero bit height every time we change a bit anyway and the fence could run like the arms of a delta printer, the home position is at the rear of the table and the collet centre is a calibrated offset from the home position, this brings in the benefit that on the dual motor drive of the fence once you set the home position each side at setup you just drive the fence back and home each side, this handles any missed steps or fence being knocked out of square when the steppers are turned off.
Another day - another thought
Hey i know that if your reading this thread that you likely want to see something built .... hang in there please
-
27th August 2020, 01:00 PM #14GOLD MEMBER
- Join Date
- Jun 2005
- Location
- Helensburgh
- Posts
- 7,696
Peter, the thinking behind getting rid of the router was several fold and I think that has been lost a bit over the period of time Mike and I have been playing with the idea. The main one was cost, an all singing and dancing RT as mike intended can be built for less than a new router and lift can be built for a way better and more versatile result. The second reason was that hanging a lot of weight from under the top is asking for trouble and removing that makes it possible to make far lighter tops that can be removed without tools and can be changed for custom tops if needed. The next reason was speed control, the VFD approach gives way better speed control hands down. As an aside we made the lift for the manual version for the cost of a DRO, two rails and a lead screw and the spindle was cheaper than any comparable router available and way tougher. All up the whole thing cost us (then) under $700 built and was at least as precise as any router lift.
Things we found on the way, a water cooled spindle is far quiter and it generally comes with a range of collets which means anything in reason can be chucked into it such as sanding spindles.
I can understand people already having a router and wanting to use it but it is nearly comparable to buy a spindle and build a lift and you would then have a spare router. I haven't priced lifts recently so take my comment on that with a grain of salt and the prices of spindles have risen since I bought mine. I have just had a look and both CT & Timbecon have a lift for about $550 which is way below what lifts were costing back when we started this so that has changed a bit. To me the absolute overriding reason to build a RT this way is to get the weight off the top because it is a fundamentally flawed idea and restricts the versatitlity of the RT. Even if it cost more to use a spindle I would not use a router and lift for that single reason and I have three of them that could be used in a table.CHRIS
-
16th October 2020, 11:30 PM #15SENIOR MEMBER
- Join Date
- Aug 2006
- Location
- Canberra - West Belco
- Age
- 63
- Posts
- 646
Just adding a reference back to Mike's thread so i don't lose the links to the touchscreen
Automated Router table
This little project is sort of back on the juggling act, building a basic cabinet to house the current router set that will transform later whilst i get my head around coding for the ARM.
I want to be able to run Mike's code if i can't wrap my head around what i want to do.... at the moment based on my 3D printer experience i want to run without the DRO hardware and use step count/limit switch/homing. 3D printer run for hours upon hours and never seem to lose positional accuracy, homing seems a small price at power up or as needed.
now to order a display and start learning
Similar Threads
-
Challenge; Create a project with the router OUT OF THE TABLE
By Template Tom in forum ROUTING FORUMReplies: 2Last Post: 27th October 2013, 08:16 PM -
help finding timber for chest, wood-tech project western Sydney
By LucasT in forum TIMBERReplies: 11Last Post: 5th June 2013, 09:22 AM -
New router table project
By peterst6906 in forum WOODWORK - GENERALReplies: 9Last Post: 5th November 2012, 04:33 PM -
The Router Table Project
By DemetersRest in forum ROUTING FORUMReplies: 7Last Post: 24th March 2012, 12:35 PM -
Table saw blade alignment - Lo-tech / Hi-tech
By niki in forum TABLE SAWS & COMBINATIONSReplies: 6Last Post: 25th February 2008, 04:32 AM