I started this project back in 2019 and I am finishing up on the 6th prototype here shortly. This will be “RC1” or Release Candidate One.
In this version I am reducing part count, increasing part reuse with the Tool Length Sensor (which is also nearing RC1) and will reduce production complexity and cost.
This version should not suffer the oscillation the thinner metal, longer carriage base experienced. The carriage will allow for a free floating rail to reduce bearing wear as well as reduce the bearing misalignment induced oscillation.
Motion studies all say it will fit with the automated dust hood so I started production on the prototype. The prototype has been cut, tapped, bent and is now in the process of being powder coated. The powder coaters will have it finished around Nov 9, 2022 and I should have it in my hands by the 15th at the latest.
Here is a sneak-peek of the carousel and the tool length sensor in their retracted state;
I have a real problem with wood dust and my allergies. It really doesn’t take much to ruin my day. My shop has to have very, very low levels of wood dust or I just cannot function in it, period. So this aspect has been a huge consumer of my time from day one.
It is a myth that sharp turns affect flow significantly. What does effect flow are changes in pressure and things like the spiral ridges in the tube.
Being a pilot and my hobby as an aircraft designer, Bernoulli’s Principle is something I have a lot of experience with. This dust hood is not 100% optimal but the dust reduction even when machining MDF is significant.
Where the dust hood fails is large chunks of plywood or plastics that shatter off the work piece. They tend to not be ingested readily because the orifice is so tightly coupled with the ISO30 tool holder.
I have been thinking about increasing the height of the dust hood by about 10 mm because there is a small pressure drop at the 4" tube attach. It can lead to some particulate falling out of the stream and catching in the first bend but as the Z axis moves up and down it works itself loose and so the 10 mm bump gets bumped in priority.
But great question!
When I walk around my shop after cutting MDF all day there should be ZERO MDF dust on my other equipment. I am not a clean freak up I am terrible vexed by wood dust!
This dust hood has been absolutely phenomenal over the past 6 month. Definitely my best design in 4 years.
I actually have a project coming up where I will need to use a fly cutter on MDF and I will take some video of how well it handles that acid test!
Not yet. It will be kit and I will probably release the DXFs, STLs for the more adventurist types.
One thing to consider though is a “nice” ISO30 tool holder is about $150. This is a 10 tool carousel, so to put 10 tools on it will set you back $1,500, not including the actual tool in the holder. Figure between $35 and $80 for that.
One mistake people make is thinking they only have 10 holders and swap their tools between them. But that opens the door to risk of sticking a tool in the machine you think is set up and its not and you get a crash.
Just a heads up though, your ESS based controller is probably out of axis channels to make this tool carousel work. You will have to invest in the ModBus compatible external controller and sensors.
As far as the design, I have hundreds of pounds of plastic prototypes. Most are manually removeable brushes. This design is a combination of several designs and came about after a version got warm and sagged during a long project. When it came to dock the brush it was out of alignment and cause the X axis to lose steps and it wrecked my tool carousel! Ouch!
What came out of that event is what you see. A mostly metal design with just the last little part that doesn’t touch the spindle being plastic. Only the back part of the brush moves and there is a sensor there to prove the brush is out of the way before moving the carousel to grab the tool.
The formula for the orifice size to the tube attach point does induce a pressure drop in the dust hood assembly (aka it is slightly narrower than the 4 inch tubing). The thinking was to increase the velocity of the airflow around the tool. Now, does it do what it is was intended to to? Dunno. But it was significantly better than my other designs and just kind of stuck.
I doubt that its perfect and I doubt it will continue unchanged going forwards. Right now I can walk around my workspace and not choke and gag, so that is important.
The powder coaters are dragging their feet as usual. I doubt I will get the release candidate this week
Neat news is Ultimaker now has a Blue Tough PLA!
So in addition to Blue ABS, CPE, CPE+, Nylon (CF/GB) and Stainless Steel Type 17-4 PH we can print in a pretty Blue Tough PLA! The blue pairs well with the yellow powder coated components. Makes for a visually striking addition to your anodized aluminum and red powder coated AvidCNC.
Ulimaker Tough PLA a great and cost effective plastic that is similar strength to ABS. Its low shrinkage makes it ideal for higher tolerance results, especially when compared to ABS, CPE and Nylon.
So far, so good. Waiting on some zinc plated M5 for the worm drive adaptor plate. The black oxide are there to check fit.
The carriage is wicked evil with all those M8 and M5 threaded holes and the powder coating.
I need to slice off another 1 mm on the risers under the M8 threaded holes so that you can torq the risers flat to the carriage mounting plate. Small visual defect for me but that will be a fix before production.
I plan to toss the spoilboard tomorrow, check the cross members for straightness. Install the new carousel. Use the probe to map the cross members so I can cut accurate holes for the M8 screws to attach the new spoil board to the cross members. Record video of the operation to cut the spoil board to show off the carousel.
Then it’s calling all my suppliers to start pricing the kits. Then calling the bank to make sure I have the capital available.
Here is a quick motion study of the system. I had never clicked the button in Fusion 360 to publish one so you know, first time for everything. Its too short and it was meant to play as ping-pong loop.
I ran into a bit of an issue with the ball screws. I had to go with the 350mm length and when I did that I had to come up with a more suitable mount with bearings to hold it stationary.
If you have priced ball screws from Thomson right now you know my wallet has a serious hole in it.
The motor and bearing mount has proven troublesome. I am using the Teknic SDSK motors which have the built in controller and they have an L-shape to them. The part that hangs over the two mounting screws makes the maximum sized screw you can use to secure the short body NEMA23 models 40mm.
A regular stepper motor this would be no problem and I could use the 50mm screws. But, at 40mm, that is way to short!
That means I had to break out the PVA and print large overhangs with PETG… and then I ran out of that! Argh! But I really should have been printing in CPE because of the exposure to oils and other lubrication. CPE is more tolerant but only just a little weaker than PETG. Lot’s more warpie though
Now, if you are a 3D printer buff you know PVA doesn’t stick to PETG or CPE. That takes a certain kind of magic. Send me $500 and I might reveal my secrets
Anyways! I will have the CNC back together in the next two weeks and everything dialed back in. I have added a new sensor to check if the slot is empty or not before a operation. Keep you from trying to stick a tool into a slot that already has a tool in it or pull a tool from a slot that is empty.
Turns out 61" x 121" 3/4" MDF is really hard to find down in Georgia right now
Good News! : I have done a bit of research and the carousel will fit all models of AvidCNC PRO v21.1 machines.
Turns out the PRO60120 (my machine) actually has the LEAST room at the end to place a retracting tool carousel. But the overhanging design will remain the same. The rack and pinion drives and the proximity switches mounted on them are the actual factor in how far back the gantry will go on all of the PRO level machines. That distance was the design consideration for how far back the carousel carriage had to move to completely remove the tools from the work envelop.
Now, the bad news
Yep, the smaller the AvidCNC machine you have, the more expensive the ball screws and linear rails will be because they will need to be longer. It is noted that you probably purchased the smaller machine to save money in the first place. Can’t win them all
Also, you still cannot put longer tools on the table because the back of the carousel still crosses over the end piece 40-4080. The maximum length for a tool is measured from the bottom of the gantry to the top of the 40-4080 end piece, minus the HSD stud, ISO30 holder and nut. Maximum tool stick out will be about 80mm or less for the 50mm ISO30 holder and shorter for the long reach. So stick with stubs to about 3 inches.
Good news on that front, give the too long tool a slot number of 99 and then the M6 will drive to the tool changer and let your do a manual change! This means you can leave longer tools on the side or leave the probes and markers on the side and keep the table full of cutters for when the fun starts
Also, the carriage will need to be located about 20mm to 50mm from the leg braces. Further if you have the welded steel braces.
I have everything together and just need to retune the machine and I will start working on some videos for folks.