Oof, did your partner have to weld you? A genuine TYFYS. I am still fairly new, second year as an ACP, but my wife has 8 years ACP. And we still speak!
The job is challenging and fun, until it isn’t That’s where craftwork, teaching, and music step in.
so funny story. My brother in law used to be a service engineer for GE medical. He got called into a hospital to fix their CT scanner. Worked on it for three days, 14 hour days each. Too many hours hunched over and on his knees. Finally got it up and running. As he was packing up his equipment he felt dizzy. Stood up as an ER doctor was walking by. Doc saw it in his eyes. Bro in law Promptly collapsed as the doc was grabbing a gurney. BIL had a stroke. He was the first patient to be scanned in the machine he just fixed. Which was good because the next available CT was 100 miles away and he probably wouldn’t have lived long enough to use it.
To this day he still hasn’t decided if fixing the CT saved his life or almost cost him it.
He’s now in IT, working as a database administrator in the same hospital he nearly died at.
None of this has anything to do with the new Maslow. Except maybe the lesson that no job is worth working yourself to death over. So @bar, let us know if you are feeling dizzy as you try to get this rolled out ok?
it’s the third kickstarter that you’ve done, and you learn each time
looking forward to this, and to adapting it for metal working (both with a spindle to get lower speeds, and building a metal version that can be used with a plasma cutter)
I’m assuming the code will all be opensource again, correct?
Of course! Code, electronics, and hardware…but it might take me a little bit after the kickstarter launches to get it all organized and documented properly
I’m absolutely going to take you up on that. I’m too deep in getting all the business paperwork and stuff in order before the launch to even show you where to start (and it’s so hard to work on hardware without having the hardware), but I would absolutely love help at some point in the not too distant future.
It might be a little bit before I can get you a full machine to test on, but if I sent you a control board you would be able to at least run the firmware. Is that something that you would be interested in?
Certainly interested in testing the control boards - just be aware that physical things take some time to get here (Timor Leste), and then I have to go through the dance with customs.
Thankfully though, I do have a business registered here, so I can get things cleared through customs in the business’ name. In other words, hard, but not impossible.
If there’s enough time to organise things I may be able to get it via a business in Darwin Australia. Which would cut out the hassle with customs as they would just clear everything together.
I am hopeful that on this version better code commenting and maybe even a launch version functional description could be put together in order to help future contributors understand the functional intent of the code better. One of my greatest hurdles with trying to modify the code from the previous versions was that I had a very difficult time trying to follow the intent.
Maybe on this go around we can emphasize the necessity of descriptive commenting being included when submitting pull requests?
Just let me know what I can do to make it happen. I am happy to pay for any customs fees that might come up.
Absolutely. An ESP32 S3 is the correct board, but it might not be that useful since the project is built on FluidNC which is already very well tested. As far as I know the only real issue right now is that occasionally it segfaults while running a file. The issue only happens when reading from the SD card so I think it’s something todo with trying to read from the SD card and the encoders at the same time. I haven’t had a chance to look into it yet. A stand alone ESP32 wouldn’t have the SD card or encoders connected so it probably wouldn’t help with that sort of testing.
I fully support that. Unfortunately FluidNC doesn’t have the most easy to read code. They are quite responsive when I’ve asked them questions. We can at least make our parts of the code easy to understand.
But it’s worth “poking it with a stick” to find out.
“But why?” you may ask. If we’re dealing with something like segfaults then that’s already reason enough to explore. Getting non-professional programmers (who are probably more used to Python, Javascript, Java or other “Garbage Collecting” memory management languages) to pay careful attention to “programmer writes the code” memory management (used by C, C++, Rust) is best avoided where possible, although Rust does a great job of helping you avoid dumb stuff.
Aligning.
As seen on picture the unit seems very closed. Is it at all possible to view the cut while working ?
How would you realign if you get interruped ?
Or need to do several exsact same cut on serveral sheets. and using the sheet edges as part of the cut.
Eg. you need a precise point on the sheet to start.
generally, in CNC work, if you get interrupted the part is toast and you have to
start again. sometimes you get lucky and can re-start with the same workpiece
you need to have a known 0,0 point (and I don’t know how you can do this if it
does a recalibration each time, but it could be that it positions itself to
equal belt lengths in all directions as it’s 0,0)
Bar??
Or need to do several exsact same cut on serveral sheets. and using the sheet edges as part of the cut.
Eg. you need a precise point on the sheet to start.
if you have a known 0,0 point, then you put stops on the machine and align the
edges of the workpiece against that.
you can also position the router next to the edge with the motor turned off and
extend the bit to ‘feel’ for the position.
It has current sensing on the driver motors so I would think the same blocks which would form a corner alignment for your plywood sheet could also be used for setting 0,0 or some other fixed/known locator.
It is quite enclosed, but you can see in enough to align the bit with a feature on the wood if you need to. The top of the dust collection system is clear acrylic and you can see down to the tip of the bit
It’s a fully custom implementation. I found that with the controller mounted on the router the vibrations caused any “headers in sockets” type connections to be unreliable so everything is integrated into a single board with no moving parts. Even the “SD card” is actually integrated flash memory that imitates an SD card because regular SD cards couldn’t handle the vibrations.