November 2021 – Oddwick Industries

So after getting tired of trying to fasten parts down to the wasteboard and my wasteboard bowing, I finally decided to actually create a rail system for my bigger router. I had toyed with the idea for a while, but truthfully I have been lazy and not very motivated. Winter is a bad time of year for me and in the fall I begin to start feeling down and this year has been a rough one with the city waging war against me, having to move next year, and of course losing my best buddy Damage…

My buddy Damage and his favorite purple towel after bath time

So in order to get me up and moving, I have been giving myself projects that are not too difficult that I can complete with few if any problems to keep my frustration quotient low and keep me moving otherwise I might end up melting into whatever I am sitting on. As part of my continuing upgrade of my larger router, I decided to redo the deck and fastening system.

Now keep in mind that this is just a temporary upgrade because I have never been satisfied with this design and on the next iteration, not only will it be larger, I will be doing away with the v-wheels and the belt drive. Those are both good for low cost and simple machines, but some of the things that I have been doing are beginning to test the limits of both systems. I am not sure if I will be going to fully suported bearings or linear rail, but right now both are being considered. As for the belt drive, I am just beating a dead horse with that one. On the one hand, it is fast, but it is anything but accurate. I have to recalibrate it at least once a month.

So in the future, I am pretty decided that I will be going with a rack and pinion. For my purposes, it should run just as fast which considering that I am wanting to go at least 4×8 and will only be using it for wood or acrylic, speed is definitely huge factor. At that size I am fine with a small loss of accuracy for a huge gain in speed. The other big change will be going from steppers to servos. I think I have gone as far as steppers can take me both in terms of speed, accuracy, and torque and now it is time to go to the next level.

The original deck with the t-track laid out

When the original torsion box was built there were 6mm threaded inserts placed on a grid about 100mm or so. That worked fine for a few things, but I really had alot of problems trying to line things up and it was not very flexible. In the end, I ended up using a piece of 3/4 particle board as waste board and just screwing my workpieces into it. Not efficient, but effective.

After I redesigned Chibi, I went with a slotted aluminum deck. This was one of those moments where you end up asking yourself why you never did it sooner. It was a total game changer. One of the main reasons that I never went with a t-track bed was that because of a design flaw with Kuro, event tho the bottom of the gantry was nearly 5″ above the deck, with a bit in it only had about 40mm of clearance. That is pitiful for trying to route anything other than plywood. But now that Kuro has a new Z axis, it has the full clearance and can make it over t-track hold downs.

Originally I had thought of using the original inserts, but I very quickly realized that there was no way on this planet that I was ever going to get them to line up and make them and the tracks fit together tightly. So, as you can see in the picture above, i laid out the tracks where I wanted them and went to the table saw with some scrap MDF and alot of math that didn’t work so well but got me close and proceded to cut a series of blocks to test the fit. I kept shaving them down until everything sat tight. Then I went ahead and cut 6 full length slats and just like magic, they fit perfectally!

From there it was pretty easy going except for one slight hiccup, but I was saved by what turned out to be a brilliant idea (Completely unintentional of course). When I cut the slats, I cut them about a foot or so longer than I needed. The thought behind this was that I was going to use Kuro to drill all of the mounting holes. At first, I was going to do them one slat at a time, but then figured that as long as I got everything lined up, I could do it in 2 operations the first 5 sets of mounting holes, and then flip the slats around and do the last row. It sounded great in theory and in practice it worked great all except for one small thing: make sure that when creating a job, you DOUBLE CHECK to make sure that inside MOPs are actually selected to cut the inside!

The plan was to do all of the countersinks and then do the pilot holes so I wasnt trying to plunge straight through 5/8 mdf with a 1/8 bit with a 1/2 depth of cut. Since I can’t reach the extreme ends of the deck, I lined all of the slats up to the front end of the machine. Once I cut the first 5 rows of holes, I would flip them around, mark where I needed to cut the excess, take them down on the miter saw, fasten them back down with the all of the holes I just did lined up with the back of the machine and then run one quick job to cut the first row. So the first row cut perfectally and when it got to the 2nd row all of the sudden the countersinks were huge! Well crap, it was running on the outside of sinks rather than on the inside as it was supposed to. I though for sure that I was going to have to rip another piece of MDF and what a waste. But then I realized that since I had only done the countersinks, I could just flip them over and make sure that I offset the holes to where the new ones would clear them. I took each one to the miter and cut about 100mm of each of the ends and realigned them and ran the job again this time making sure that I was actually cutting the insides of each countersink.

The final touches are being put on the laser. One thing that I ran across that turned out to be a huge issue is that when I build things, I tend to over engineer it and my creations end up being built like tanks. This laser is no exception. The lid ended up being pretty heavy and the first set of gas shocks I had got were rated at 10lbs each which when figuring the lifting angle was not enough. So the second set I used were rated at 20lbs apiece and that is when I ran into the problem. When I had originally designed the dibond lid piece, the intention was to just fasten the gas tube bracket straight to the dibond. That worked ok for the 10lb tubes, but when I closed the lid using the 20lb tubes, between the weight of the lid and the force of the shock, it started bowing the dibond up. Luckily the way I positioned the frame side pieces, this left me enough room to fasten a bracket mount using some left over aluminium L. I was able to use the original holes in the dibond and line them up with the new bracket which was attached to the frame rails and took the force off of the dibond. In the end, the shocks sat about 3mm lower than I had planned and slightly below horizontal, which ended up working better than expected.

The wiring ended up being messier than I wanted, but that is the story of my life. The controller is a Cloudray Ruida RDC6432G which after a bit of some initial issues, ended up working fantastically. But more on that in the next post. The motors are Stepper Online nema 17 84oz.in 1.8º steppers. In my first version, I was using 0.9º steppers to try and increase the resolution mechanically rather than with microstepping, but somewhere along the way, they developed a very nasty whine whenever they were ran fast. I thought that it was the bearings in the wheels going out or possibly the y carriage draging somehow, but when I put the motors in this laser, they made the same noise. That was what had actually started this project in the first place. But once I switched to these motors, then all was well. The first ones I replaced them with were something like 60oz.in, but they were badly underpowered and stalled at anything over 200ipm.

The stepper drivers are StepperOnline DM420Y digital drivers. I first started using digital drivers when I did the MK3 version of my Chibi Router which uses DM542Y drivers and was blown away with how smooth and quiet they ran. Of course like every other DIY maker in the world, once you stop using arduinos to run your machines, the first stepper drivers you get are the infamous chinese TB6560 drivers which crap out if you look at them too long. Then you move to the tb6600, which if you do the power override mod on them, then they are more reliable and powerful than the 6560s, but watch out they get HOT and make sure that the resistors in them can handle the wattage, because the ones that come in them by default are woefully underrated. But when you power them on you get the awful thump of the hard power up. With digital drivers, they run MUCH cooler and have a soft power up so it does not sound like your machine is being ripped in half on startup!

To mount all of the electronics, I designed a rail system that could be used with minimal modifications for each component. When designing things, my main goals are always accessability, interoperability, reusability, and ability to transpose the part. This rail system hit all the marks. The only way it could have been better would be if i were able to only cut 10 copies of the sam part, but the screw holes were too varied to pull that off, but otherwise it works fantastically! Everything is adjustable and can slide on the inner rails so I can get my hands in there and fasten everything down.

The relays ended up being the biggest facepalm part of this entire project. When I do a writeup on the software side of things, I will detail the steps I had to go throught to get my controller to work, but in the process I discovered some hidden features, one of which was a fan control timer. Of course I discovered this AFTER I had already got a time delay relay installed. What I wanted was just for the exhaust fan to run for a number of seconds after the job had completed so as to completely evacuate all of the smoke from the chamber before I opened the lid. Who would have ever thought that would be so difficult? The instructions for the relay seems to have been written in chinese and translated into half a dozen other languages before they decided to settle on english. So once you kindly signal the hiccuping giraffe until the lucky lumber is at peace (makes perfect sense right?), then and only then will the fan remain on 20 seconds past the end of the job. So after all of that, once I get it up and running, I discover that there is a hidden setting in the controller that will do exactly that. And by hidden, I am not even kidding. Burried and and completely undocumented.

Adding graphics to the cut plates prior to assembly

And there we have it, a finished laser engraver that works beautifully!

Month: November 2021

Kuro Gets a New Deck

Finishing Up