[RedskinsJBS]

The design is coming along. I'll try to help with whatever I can.

1. One thing to remember that can be difficult to see is that when using the R&P fixture on a carriage, you cannot use the two outer holes to attach the carriage to the frame. The pivot bolt and spring bolt from the R&P intersect and obstruct where the through bolts would normally go. I guess that you could tap the carriage or shorten the length of the R&P bolts, but I haven't done that yet. Of course, I haven't cut anything yet, so all is left to be discovered.

2. This is why I went with 3030 for my long axis. And it is why I have a low profile bed. I made my gantry create more of the height clearance that I desired. If I were you, I would simply plug in whatever the freestanding height of your long axis rails into the deflection calculator and call that your worst case scenario. I used 10.5" with a 75 lb end of beam load on the 15 dimension of 1530 and ended up with 0.006 in of deflection. Now, you will have two beams taking this load at a fixed distance from each other. Plus, you have designed some in some gussets that effectively shorten the beam length. If I use 7.5", the deflection drops to 0.002 in.

In this example, I used a 75 lb load. This load would be entirely generated by cutting forces on the router (plus momentum from the shorter axis). I estimated 75 lbs because of my experience using routers by hand. If I were to experience a 75 lbs force by hand, I would have to move the router slower through the material.

I suggest that you come up with a way to estimate the amount of force that you are estimating you will see and make your calculations from there.

3. I have a NEMA 23 setup, so I can't offer any advice here.

4. I am going to mount my router so that the bottom of the collet nut is flush with the bottom of the 1530. I see this as the maximum depth the router can plunge into the material simply because my G-code software can only take in to account the size of the end mill and nothing else. If I want to mill deeper material, I just have to get a longer end mill. My 8" of use-able cutting volume pre-spoil board was selected so that I could mill half of that without any worries. This way, I could easily mill 4" stockall the way through.

[curriergroh]

A closer look at the cutting bed/long axis rails.
The spacing of cross pieces is now at 20" instead of 24".
All connections will use end fasteners. The gussets are used solely to resist the lateral forces of the z carriage and cutting.
The table will have two layers of MDF (not shown).


A closer look at the corner of the table showing the 2 1/2" steel rail, end condition, and one layer of MDF.


Two changes here:
The gussets are only on one side of the gantry beam.
The gantry beam is shifted 'back' on the bearing carriage.
I think it centers the weight a little more, and increases the cutting area slightly.

[curriergroh]

Reading other posts, I think the gantry connection in the previous post is not possible because the R&P drive intersects one of the holes I'm using for the attachment.

And if I move the 3060 back to the center of the bearing carriage then it would be OK, but I would not be able to use gussets on either side.

Anyone concur?

[RedskinsJBS]

You can use all 4 holes on the inner sets of bearings, so gusset(s) there would work. However, the outer bearings that carry the R&P can only be attached with the inner holes.

I have been contemplating building some plates that I could use to 'pin' the inner and outer bearings together without using any screws, but I haven't found it to be necessary. The two bolts on the outer bearings seem to be enough to hold them in place. Having extra attaching points on the inner bearings just adds redundancy.

Just my opinion.

Jay

FWIW, here is a link to my gantry design. Hopefully it can help you in some shape or form.

http://www.cnczone.com/forums/897949-post116.html

[curriergroh]

From your link:
"All of the 8020 that is adjacent to one another will be securely bolted to each other using 8020 hardware."

What hardware from 8020 makes this type of connection?

[RedskinsJBS]

From your link:
"All of the 8020 that is adjacent to one another will be securely bolted to each other using 8020 hardware."

What hardware from 8020 makes this type of connection?

I'm not sure if 8020 themselves actually have a piece of hardware specifically made to do this, but I'll do my best to describe what I did.

First of all, you have to drill access holes through the 8020 where ever you will be bolting the pieces together. I used a 5/16" drill bit with a fence on my drill press to accomplish this.

The hardware that I used was simple. Button head screws and weld nuts (the cheapest t-nut available). I'm sure that you could use more expensive t-nuts, but I don't really see the point.

8020 normally supplies you with 3/4" button head screws (for 15 profile extrusions) to use with their t-nuts. These would bottom out if you were to try to bolt two extrusions directly together. To go around this, you need to use 1/2" button head screws (5/16-18, of course).

After drilling the access holes, I used a screwdriver through the access hole to act as a stop. I then thread the 1/2" screw on to the nut a single thread so that it will still slide along the 'H' channel that is created. Using a push stick, I move the screw assembly down the extrusion until it contacts the screwdriver. This is the difficult part. I then had to slide the screw assembly just enough so that my hex key could grab the socket in the screw (Ball end works great here). Too far, and you may have to disassemble the entire assembly and start over.

I bolted two pieces of 3030 together for my gantry using this method and it worked great!

[harryn]

Hi, if I tried to build your current design, I would fail. The reason, is that I cannot reliably make high precision, cut parts, that match to the level needed.

Examples:
- The cross braces under the bed
- The support pieces that hold up the X rail
- Cutting the mdf for that bed will be a pain

I am not saying it can't be done, just that it would be very difficult for me to do it with my limited tools. Example changes I would make to take that on:
- Let the cross braces under the bed go all the way PAST the long X rail support. This means that the long X rails are now mounted on top of the bed cross braces.

For me personally, this makes their exact length less of an issue, and the angle brackets can be mounted outside of the bed, instead of inside, simplying the bed mdf cuts.

This will also naturally raise the side rails of the X rail support, potentially reducing the need for making all of those precision matched support parts to make the X rail the right height.

If the cost of adding in enough of a support height is the challenge, go the metal yard and buy a 2 x 4 x 1/4 in steel tube to fill in the gap. Alternatively, just fill it with a kiln dried 2 x 4 that has been run through a planer a few times to parallel the edges.

For people like me with limited tools and skills, it just isn't that easy to make a lot of parts exactly match. Your design assumes that you can do that.

As far as pricing, when I looked at 8020 type materials, the main price issue was connectors. You can save a bundle by using something like steel tube cut offs instead of those fancy Al gussets. It does take some drilling, but it isn't that bad with a drill press.

[Zuni]

When I started looking at the first post in this thread, I screamed "Thats EXACTLY what I have in mind for my upcoming build, except I'll use a spindle instead of a router!" I like how this design is evolving, but I feel kind of strange sitting back and watching while curriergroh does all the leg work.

- Let the cross braces under the bed go all the way PAST the long X rail support. This means that the long X rails are now mounted on top of the bed cross braces.

For me personally, this makes their exact length less of an issue, and the angle brackets can be mounted outside of the bed, instead of inside, simplying the bed mdf cuts.

I like that. If you don't need the bed depth, that REALLY simplifies things (back to the original design, for the most part). I think you'd have to change the long x-axis rails to 1545's to eliminate interference between the corner braces and the rack. If I understand it right, this (crudely) illustrates what you mean?:

[harryn]

Yes, exactly. To make those sides taller, you can use either ordinary rectangular tube (steel or Al) and put the 8020 on as a top plate if needed I guess.

The costs can be knocked down by using square steel tube (or possibly just an angle) to replace those gussets. It is not as aesthetic, but my experience with 8020 assy components is that they are on the spendy side.

Remarkably - very similar to your early bed design.

[pbarnes8020]

To try and save on the weight of the frame and the cost (while keeping a square profile with two T-slots per side) you can look at using the newer 3030-Lite profiles or 3030-LS (Light 3" x 3" with a smooth finish).

3030-Lite T-Slotted Aluminum Extrusion

or

3030-Lite Smooth T-Slotted Aluminum Extrusion

[Mountaincraft]

This is how I plan to solve the gantry end support issue... It can be cut from a single piece of 1/4" steel, and then bent up with a single weld on the outside of the corner...









Alternatively, I was thinking of making a block out of 3/4" aluminum to bolt across the bearing cars, and another bolted to the back of the first piece to form the back of the box, then a piece of 1/4" steel on the outside, bolted to the first two pieces...

My CNC Router/Plasma Build Thread

[CarveOne]

I think you are going to have problems finding a machine that can bend 1/4" steel with square corners accurately like your rendering depicts. Maybe one bend and a second plate welded in place to complete it is more do-able.

CarveOne

[Mountaincraft]

I think you are going to have problems finding a machine that can bend 1/4" steel with square corners accurately like your rendering depicts. Maybe one bend and a second plate welded in place to complete it is more do-able.

CarveOne

Yeah, that's why I included the 'alternatively' part....

I'm gonna pre-fab mine out of 3/4" plywood during mock-up.. I may end up just welding all three pieces out of 1/4" steel.. 3/4" aluminum is expensive!

In reality, my gantry will probably extend about 1.75" past the outer bearing car, so if needed I could bend 'that' angle... and then weld a plate on the back side...

[CarveOne]

It's going to take a serious commercial hydraulic fabrication machine to make a square inside corner with 1/4" steel and not have fatigue cracking in the bend. That's stretching the steel quite a bit in the outer part of the bend.

I would recommend welding three precut pieces using a welding jig plate fixture to maintain 90 degree angles. The 3/4" aluminum plate may be cheaper in the long run if you have to pay a welding shop to do it for you.

CarveOne

[Mountaincraft]

I have a little time to work all that out.. But I have about a week to decide on the motor/driver/PS package.... And I'm clueless...

I could really use some help in that area.. My new machine build thread is a couple threads down in this forum...

[curriergroh]

I've actually come full circle, as some others pointed out. I will be using the gantry interface plate shown in the first post. Advantages are a smaller machine footprint relative to cutting area. It also moves the complexity to the gantry assembly instead of the table assembly, which will save about 200 inches of aluminum and $200.

I was gone for a few days looking at some shop spaces in Cincinnati. I think I found one that will work. So now it is time to start purchasing. I will start with the motors and electronics first and continue to revise the design of the aluminum and steel frame.

I made the first purchase last week. Aluminum heatsinks for the drivers.
They were only $1 and the perfect size. I couldn't resist.

[curriergroh]

Power Supply x 1
Keling KL- 7220 Unregulated Power Supply 1440W, 72VDC/20A ,120VAC or 230VAC - file attached

Breakout Board x 1
PMDX-122 or PMDX-126

Driver x 4
Gecko G201X

Motors x 4
Nema34 640 oz-in Current 7A, Inductance 2mh - file attached

I don't know how I am supposed to evaluate the breakout boards.
Here is what I know I want: 4-axis, e-stop, limit switches.
I would also like the router and shop-vac to be turned on and off. What else do I need that I am unaware of?

I have a friend using the PMDX-122. So that one has the benefit of being able to share experience.

I think this works. If anyone sees a problem with this setup in general or in relation to the design please let me know because I'm itching to pull the trigger on these.

[Zygoat]

Can't beat that for $1

[RossMosh]

I think you should look into using 3030 as your primary material. I was quoted $0.67 per inch on 3030 where 1530 is nearly double at $1.08. Personally, I thought the design in page 2 makes a lot more sense than page 1. Can't say I think very highly of the way the way the gantry and carriages connect on page 1.

I'd also look into minimizing the gussets from 8020. They are ridiculously expensive considering what they are. Just switching to an inside corner bracket would save you about $3-4 per piece. I assume you plan on drilling and taping the end of the extrusion and using carriage bolts to assemble everything so I'm fairly certain that a 4 hole corner bracket would be more than sufficient at making everything more rigid.

Hi, if I tried to build your current design, I would fail. The reason, is that I cannot reliably make high precision, cut parts, that match to the level needed.

A lot of people spend the $2-3 per cut and have the 8020 supplier make the cuts. They have ultra precise cutting methods that leave you with consistent and accurate prices. Automations4Less doesn't even charge you for the cut.

[Mountaincraft]

Power Supply x 1
Keling KL- 7220 Unregulated Power Supply 1440W, 72VDC/20A ,120VAC or 230VAC - file attached

Just curious why you chose that power supply? Is it because it's the highest wattage?

Does this hook directly to the stepper drivers, or is there regulation/filtering, etc. that must be added in between? Are the drivers protected in any way from spikes or whatever that might be put through this PS? Or is that not an issue?

In general, is bigger, better? Or must the PS be sized so that the intended motors/drivers use up a certain percentage of available power?