Відео

No idea, I don't have a 7104 - sorry.

​@@AussieMikesBees thanks! And in answer to your question... "Sort of". The skirt (note: not installed in this video, see newer video ua-cam.com/video/ELEo6ToVVNA/v-deo.htmlsi=LNttPcyK61bfKNdS) is a ring of foam strip and is compliant enough to pop over low-profile clamps in some orientations with the relatively small force swinging the shoe aside. When it gets hung up, what happens is that as soon as the shoe begins to retract, it pops free of the clamp and swings the rest of the way. It's a little bit of a race condition, as it needs to swing free before it rises far enough to catch on the tool in the spindle, but hasn't been an issue so far. What I probably should do is add a limit switch to detect whether it's successfully retracted before attempting a tool change.

@@funkenjaegerthanks for your reply. The other video shows the foam and the two section bellows. Clever. With the only way in for air being through the small gap under the foam skirt, it looks like no chips can escape. It looks like it performs well over the edge of the stock too. I use a Masso controller and not familiar with Mach. I see you touch off with each tool change. Does Mach not have a tool table where you only need set the tools once and store the resultant offset? Masso also has the option to wait six seconds for a “Dust hood UP” sensor before proceeding. Same for “Dust hood DOWN “. My pneumatic piston has a magnet built in, so there’s a sensor at each end. I hope I can decipher your drawings and adjust to my machine. This is the best solution I’ve seen. Mike

​@@AussieMikesBees I gave up on Mach years ago and switched to LinuxCNC. Pretty much all the code that performs the tool changes, touch off, and dust shoe control is my own code, so it was quite a bit of work to get it all integrated, but the upshot is that I had full control (which is sort of a theme with LinuxCNC in general). LinuxCNC has a tool table of course, and I could have implemented it that way, but that's just not the approach I chose. The extra seconds it takes to touch off the tool after each change doesn't bother me for my use case; if I was doing jobs with a high number of tool changes, and/or doing repetitive production work I'm sure I'd feel differently. The upshot is that by doing 100% "online" tool length measurement, it's less likely that I can end up with a bad crash due to mis-configuring my tool table relative to my physical tool rack.

I haven't modified it yet, but my most likely plan is to use lathe + TIG to just extend the existing rods a couple inches rather than making entire new rods from scratch. IIRC the left-hand rod has gear teeth hobbed into the side of it for the clamping mechanism, and trying to match that geometry on a new rod feels like something I'd rather avoid. It seems to me that if you want to use it on a 12" beam, the length of the leadscrew is going to fall short of letting you reach the center of the beam, let alone reach past center to make a wider mortise without having to flip the whole machine around halfway through each mortise. It's been a minute since I looked closely at that leadscrew, but I suspect that would be a bit of a lathe project to fab a new one.

There's a tensioner bolt on the back side of the bar. I never found that I needed to mess with it on mine, but maybe yours was way off when you received it? imgur.com/a/h057tMH

Thank you!

Yeah, I've definitely got plenty to learn still. My trouble is, all the trees that are giving me problems on my property are either sprawling multi-stemmed / lots of forks, or gnarled pines like this one (or worse) where I have to at least get up 10 feet or so to get to where it's more or less just a single spar, so generally not super straightforward to learn on. This removal was pretty far from efficient, but with dual ropes, flipline and spurs at least I never felt at risk of falling. I do definitely look forward to getting better with spurs going forward though.

Yeah, in retrospect, on a few of those cuts where I just cut through without notching first, I probably should have gone with a snap cut or something.

@@funkenjaeger bro you knew i was tryin pick somethink. Cheers

@@funkenjaeger bro put it out there. Sorry i was pickin

Nope, I've run it on 120v with zero issues.

GRBL has no spare I/O and no capacity for I/O expansion, so I don't see how it could support an ATC. There are GRBL alternatives that are based on more capable microcontrollers like ESP32 that others have integrated with an ATC, but I don't have any experience with those. Regardless, I don't believe any of those work with LinuxCNC so you'd be looking at a completely different hardware, firmware and software baseline than what I used here.

The sacrificial waste board is a scrap piece of Corian

That's extremely useful info, thank you!

I know this is an older comment but if you don't mind could you lead me to where you got the tool steel rods. I am only finding 100 and 150mm long rods. Thanks!

Word of advice, be skeptical of whether you'll receive the one in the photo unless the auction listing is absolutely explicit and unambiguous about that. Also be prepared for the risk of shipping damage, these units are large and awkwardly shaped so are pretty tough to pack. I've now got two, both of which arrived damaged in different ways - fortunately I was able to repair both.

Agreed. The first one I bought came pretty mangled. Handles / trigger were completely busted off the cast surface. It was basically wrapped in one layer of newspaper and some masking tape. The shipper just refunded me and I bought another one and it came fine. Be prepared for potential shipping damage. The good thing though is they didn't have me send back the damaged one, so eventually I'll fix that one up. I posted a comment up above on how I extended the depth for clamping. It was actually surprisingly easy. 20mm diameter for one of the rods and 13mm for the other.

​@@RyanGuisewite my first one was a flop, seller's assistant was "involved in a traffic collision on the way to the post office, and it was damaged beyond repair" ... more like they never had it and were just covering themselves. Second one from a different seller arrived intact and for $100 less! Packed in miles of Japanese newspaper! Will definitely order the tool steel rod you mentioned to extend the clamping ability 🍻

My first one (the one in the video) came with the base casting sheared off where it held one of the two horizontal rails. That one I fixed with machining (you might spot the two black button-head cap screws on the left side). My second one came with the right-side handle sheared off near where it was screwed to the motor casting. That one I fixed with TIG welding - I was pleasantly surprised at how easily that cast aluminum welded.

@@mr2ferrari Hah! Yeah my second one came with miles of newspaper as well. Still using it to start the occasional fire in my workshop 😆 I wish I could post a picture of the first one that came. The packaging was... comically bad. @funkenjaeger I should try welding my handles back on. That would make the repair job considerably easier. I think a neighbor friend has a tig welder. I've only got stick and mig and have never done aluminum, though I'd love to get into it. Such a handy skill to have.

At the time of this video I was using an ESS board with Mach4. I've since abandoned both and gone to LinuxCNC with Mesa boards, and never looked back.

That part is manual - the two knobs on the front clamp the aluminum vacuum tube, so I just loosen those and slide it to the desired height. Automating the shoe height is interesting and might be on my long-term roadmap, but I didn't go for it in this iteration for 2 reasons: 1) a linear profile rail and pneumatic cylinder were more compact and simpler to integrate than a motorized linear positioning mechanism 2) establishing the Z offset for the shoe, both on-machine (manual touch-off? some sort of probing?) as well as the means of programming it in CAM, would add a fair bit of complexity that I didn't want to tackle right out of the gate. The risk of a crash with a dust shoe in the equation is much higher, as are the consequences (something would almost certainly break), so I'm taking it a step at a time.

I would guess that the answer is no, but you'd have to ask Teknic.

Sure, it's just a simple CNC. The problem is that LinuxCNC setups typically use hardware like Mesa Anything IO boards which are much more expensive than most pen plotter budgets. Consider grblHAL which is similar in functionality but runs on cheaper hardware.

@@MartinJacobsonInstruments much appreciated !!!

with a clearpath servo driver of course. gcode isnt a driver its instructions. software just creates the instructions. the CNC controller translates those instructions for the servo driver which by name drives the servo.

TPU (Sainsmart). They're a little stiffer than I would prefer so I'm going to try printing a set in TPE (83A durometer vs 95A)

@@funkenjaeger Good stuff, can't wait to see follow-up videos! Thank you!

Very clever

I just published it on Thingiverse and added a link in the description.

The model is kind of a mess because I did a lot of iteration on different portions of the design, but I posted a snapshot: grabcad.com/library/cnc-router-actuated-dust-shoe-wip-1

@@funkenjaeger Thank you

They're readily available on eBay, AliExpress, etc. Search for "vibration motor".

You can set tool numbers independent of pocket number in the tool table.

@@funkenjaeger Thank you very much for your reply. I'm arguing with some guys on a facebook group regarding my brand new Syil X7 cnc, it seems that this machine with a 16 position arm type ATC, powered by the industrial LNC6800 controller, it can't load in the ATC magazine any tool numbered with anything above 16. For me this was very weird, on my Tormach I can store tool 100 in any of the 12 position of the ATC, and now I'm searching to find any other machine that has this limitation, until now I haven't find any (LinuxCNC, Mach3, PathPilot, DMG Mori .. all can store any tool number , in any pocket of the ATC) The thing on Syil is that when the machine needs to execute the code T99 M06, it checks for Pocket 99 (instead of tool 99), and shows an error that the number is beyond the maximum pockets available in atc. The machine should check in the 200 position tool table, on what pocket the T99 is stored, and access that pocket. I really pushed hard on this, I really argued with everyone on that group, they need to find a way to change this behavior. They told me I need to change my workflow, but I refuse, I told them they are doing a wrong thing, I can't imagine how they accepts to continuously change their tool numbers in CAM software, daily they change their numbers to match numbers 1 - 16. I usually generate about 4 - 5 gcode programs daily, each has around 10 tools ,,it's a pain in the a** to renumber my tools every time (the tools that are not already in the magazine)

@@ovisopa I agree, from what you describe it sounds like shortsightedness and/or laziness on the part of the developer. CAM tools like Fusion 360 (and MasterCAM, from the looks of it) have a tool library manager for managing all of your tools globally, indexing them by tool number. Being forced to set tool numbers specific to each job breaks THAT workflow - for no good reason.

@@ovisopa Less likely shortsightedness of the developer, more likely they are removing a potential user failure mode that could result in dissatisfaction from more users than the ones who would want to keep track of a larger library of tools.

I'm using a Mesa 7i93 with two breakout boards - one 7i37-COM for GPIO like air solenoids and tool pocket sensors, and another of my own design for all the core machine functions including servos, limit switches, tool setter, E-stop, etc.

I've been incrementally and iteratively building this machine over the course of the last few years and since I only set out to do this as a one-off personal project, I have not been diligent in making my models complete, organized or keeping them entirely up to date as the design evolves. I've done a few writeups covering certain parts of the machine build ( funkenjaeger.net/cnc-router/ ) but those are far from comprehensive - it's on my to-do list to make more progress on those, but I'm unlikely to get caught up any time soon.

You can't tune two together. Teknic provides instructions for this scenario. I tuned a Y servo with it installed on the X axis, with extra weight bolted to the X carriage to simulate half the weight of the gantry. Then you copy the tune settings to the other Y servo with the direction bit flipped.

@@funkenjaeger hot garbage

It's a 2-ton hoist, but I had to use it on the 1-ton arm position to be able to lift high enough for this.

@@funkenjaeger Thanks

The GUI is qtdragon (qtvcp) with minor modifications.

My config is on GitHub: github.com/Funkenjaeger/fj-lcnc-cfg

@@funkenjaeger thank you for doing this! 🙏 few more questions if you'd be kind enough to answer, which clearpaths did you end up going with and what's a comfortable ipm you're able to run them at, second is will the code for the ATC work in mach4? I currently have the ESS and mach4. Did you ever use the s30c spindle with any other vfd or just the wj200, I'm running mine on the delta vfd but just got a wj200 after the hitachi NE1 just wouldn't respond to speed changes in mach. I'm hoping the wj200 will provide more torque

I listed the clearpath servo model numbers (different model on each axis) in a comment on one of my other videos: ua-cam.com/video/98X_LPUHFn0/v-deo.html I have several VFDs on different pieces of equipment, but the only VFD I've ever integrated with CNC control is the Hitachi. I would not really expect LinuxCNC scripts to be of any use with Mach4. Incidentally, I was running Mach4 with an ESS on this machine up until a few months ago; that was before I had the ATC spindle, so I didn't have nearly as much automation integrated. I gave up on Mach4 after having used it for a couple of years for a number of reasons, but among them was that I felt there was a lack of resources for figuring out how to integrate custom behavior (such as ATC routines) - the documentation is too sparse IMO, and it doesn't seem to have much of a user community - even relative to Mach3, let alone LinuxCNC.

@@funkenjaeger definitely agree on the lack of community on the mach4 support. I bought an MB3 board with an ESS for the simplicity of using the avid xml file and now I'm thinking I should of just switched to acorns or Linux cnc... oh well, lesson learned. Maybe one day I'll have the time to figure it all out. Thank you for taking the time to respond. I'll be on the look out for future videos!

@@funkenjaeger buena tarde y saludos , aquí se puede ver tu código fuente de la sonda para la herramienta , me encanto lo del sensor para bajar la velocidad y solo hacerlo con una sondeada 😍😍😍