Hello, In the "Mechanics" section of the forum, under the thread "Stepper motor," you talk about your "MKII" version of your DiyouPCB printer. You mention to one of the users that they should wait to build the printer until the "MKII" is completed. I was wondering if there is a timeline on the MKII or if you know when it will be finished? I don't mean to pry but I am very interested in both versions of the printer, and I am rather eager to see what kind of results you guys are able to achieve with the next version of the printer! Also, I am currently sourcing the parts to the prototype, and can't wait to put this printer together!Also, if I may make three sugestions. Wouldn't a more rigid frame construction help with some of the resonace? Maybe by doubling up on the shafts of the frame around the XY plane, each of the two shafts per side separated maybe by 20 or 30 mm in the Z direction. Secondly, I hear the GT2 belts and pulleys perform much better for backlash problems, as their construction is better suited for linear motion opposed to the T2.5 belts which are designed to be timing belts. Thirdly, I have read a little about using cork around the motors to dampen some of their vibrations, however, I am not sure how much this would help. Hope some of this can be helpful! In any case, thank you for all the work you are doing! I look forward to seeing the DiyouPCB MKII!!Best Regards,Nick
Hi Nick,
Thanks for your interest and four your improvements. We appreciate them very much.
About the mechanics:
In MKII we are using leadscrews and Nema17 steppers motors. We think than most of the vibrations and resonances come from the steppers. They vibrate at certain speeds by their own nature. We saw that as slower is the speed more vibrations occur. So the first solution could be to increase the speed but the photoresist is few tolerant to it, I mean, you have a max laser speed to sensitize it well. Leadscrew has less pitch so we need more motor revolutions to move the axis the same length. It means the motors will run faster and with fewer vibrations although we will have other issues: backlash and swinging. Expensive leadscrews and nuts solve this but it broke our main design rule: has to be cheap. So it’s not easy…
A more rigid frame will help in some resonances frequencies, using better bearings, etc. GT2 belts will reduce backlash. Consider also to modify the idle pulleys because use bronze bearing was cheap but not a good idea because they don’t fit with precision in the 6mm shafts and can generates backlash and vibrations.
About using cork: we did it to isolate the motor from their holders but what we saw is that vibrations generates into the stepper motor shaft not the outer frame so it didn’t works.
Use a dampener: we also test it. We 3D printed one and we fill it with small fishing ball weights trying to emulate the commercial ones (the theory says that if you increase moving mass you will reduce vibrations). It didn’t work because dampers are precisely tuned for specific frequencies and we don’t had measure equipment to tune and make them. HP used oil dampers in old plotter models and printers. Now they use DC motors (servos) because are cheap and they don’t vibrate.. This is a better solution but add complexity: you need encoders, reduction gears to increase torque (more backash) and more complex electronics.
So finally we saw it’s almost impossible to eliminate vibrations using that configuration and you have to live with them: basically change his frequency, make the printer more rigid, change the speed, lost precision, etc.
Regards MKII, we have been working in this project aprox. two months with the following design criteria:
The idea is to design a versatile robot that serves for all possible needs that an electronics/mechatronics aficionado could have.
Ambitious? Yes too much.
In two months we “3D printed” a prototype. It’s working now… well, moving… It promises and we solved the first issues as we hit them on the way, but we want to be sure it is feasible or not before show it. (If it’s not feasible we will show it as tech work of art J) .
We migrated from Arduino UNO to Mega and from HPGL to G-Code.
In order to generate gcode from Eagle’s board designing we have been testing PCBCode written by John T. Johnson http://www.pcbgcode.org/ . It was originally though for milling and we already saw that we have to modify the code in some aspects. Fortunally is an open source project.
Also we changed from DiyouPCBShield to RAMPS 1.4 because DiyouPCBShield was though as a cheap solution for Arduino UNO and once migrated to Mega, RAMPS seems a better solution.
The Pickup Driver Circuit: we will use it as is. Also the Pickup.
We also wrote a new version of the firmware and the server software to accommodate new needs.
We will keep you informed about the progress but I think we will need at least one or two more months to have the first conclusions.
Regards,
Victor.
Hello Victor,Wow, I did not expect to hear back so soon or with so much detail haha! This is great! Your tests all seem quite thoughly done and well thought out, good job! With regards to the GT2 belts, pulleys, and other linear bearings, just thought I should mention I have sourced almost all the parts at this point. While I was sourcing things I actually could not find bronze bushings (at least not the sizes you guys had listed), and ended up making my own substitions for these among many other parts. I thought for your own reserach and mental notes it may be helpful, so I'll note what I am attempting to substitute. First, for the all the bronze bushings I am substituting LM6UU linear bearings (obviously I will have to modify your original STL files). Second, for the belts and pulleys I am getting GT2 belts although I have not purchased the pulleys because I don't really know which tooth count would be most helpful for this application. Third, for the Idle Pulleys of the X and Y axis I am substituting 625ZZ linear bearings. I plan on using two 625ZZ linear bearings for each Idle Pulley and placing them right next to each other. Also, I plan to place extra washers on either side of both of them to form a semi-loose but still tight fit. Also, instead of using the two M6x20mm steel shafts for the Idle Pulley shafts, I am using old M4 threaded rods I having lying around. Fourth, I made minor modifications to your frame namely for convience in buying parts. The steel shaft lengths I am using are two M6x600mm, and I am cutting each of them to lengths of 330mm and 270mm. This allows for no waste of the materials on my part while maintaing a L/W ratio of 11/9 which seemed to be what you guys were doing. It also allows me slightly more space to make the bed 10mm larger in length and width. I have not yet calculated how much longer the aluminum plate, threaded rods, card board, and glass will be as a result of these changes, but shouldn't be too much different. Fifth, I am planning to mount a laser on either side of the print bed in order to print both sides of a PCB in parallel to speed up overall print time of a PCB. I realize this may not be realistic due to the reasonace issues, but I am still going to try once I get the MKI up and running. A guys gota keep on dreaming right? Haha.So crazy thought, have you guys considered modeling this after the Prusa i3 Mini (since you guys want a smaller machine + 3D printing + milling + still your DiyPCB machine)? You could rasie the Z Axis slightly (maybe 10mm or so) in the Z+ direction to allow for a drill chuck etc to be mounted without affecting the build volume. You also could extend the Z Axis more agressively (maybe 100mm or so) in the Z- direction. This could allow for a single build platform with the 3D printing/milling portion of the printer in the Z+ direction (with respect to this hypothetical build bed), and in the Z- direction you could have the Laser Pickup and Pickup Driver Circuit do their thing. The XY plane (the print bed) frame of the printer would obviously still need to be glass for the laser, but these modifications would also allow for PLA 3D printing and milling. Basically, it would be like minaturizing your printer, and throwing a 3D printer on top of it haha. Is this pretty much what you guys were thinking for the MKII?However, I do have one concern, if you plan on having the MKII printer do all these things I fail to see how RAMPS v1.4 has enough to bring to the table. Specifically, it does not seem there is enough Stepper Driver support (at least not how I am invisioning the MKII). In my head, you need three Stepper Drivers for the XYZ coordinate system for the 3D printing/milling + one Stepper Driver for the 3D printer extruder + one Stepper Driver for the Milling head + two Stepper Drivers for the XY coordinate system for the DiyPCB machine, which totals to seven Stepper Drivers and eight Stepper Motors. I assume I am missing something?In any case, your design goals sound wonderful! I wish you guys the best of luck, and look forward to the hopefully inevitable success of the project! :DRegards,Nick
Hello Nick,
I think you can use bronze bushing or linear ball bearing. Personally I’m comfortable with either. Bronze bushing are cheap and more precise but also more tricky to align and adjust. Ball bearing are more tolerant but less precise. Good one’s are expensive, and cheaper ones scratch the shafts. If you will use linear ball bearings consider tying them loosely to the carriage because they work well when they are not very squeeze into the plastic. By the way, we are using them in the MKII, mainly because we are too lazy to align the bronze one’s well. I saw at internet some bronze bushing with auto-aligment rubber gaps, but I couldn’t find it locally, so we didn’t test it.
We used the RepRap linear motion configuration mainly because it’s cheap and works well in those printers. We don’t see any advantage to increase tooth number, because some of this stuff is indecently expensive. A good collateral effect of the diffusion of RepRap is that some traditionally expensive stuff like: pulleys, belts, stepper motors, even electronics, are now cheaper than in the past so our recommendation is try to use it.
Good idea about use 625ZZ as a pulley. As I told you in a previous post this is an area of improvement.
No problem about change the printing size. The printer is scalable. Perhaps we did it too big.
Good idea also to print doubled-side PCB’s at the same time. The only issue I found is the cost because you has to duplicate the pickup and some electronics. From our experience, designing always has a trade-off. It’s the main problem you will find. Sometimes it’s difficult to choose which design aspects you want to prioritize and what to sacrifice. In this case you give priority to speed vs cost. It’s ok.
And yes we are complete crazy trying to go beyond the limits but mainly is because normal thinking is already made!! :-) I think you are also very interested in thinking out of the box with your designing. This is good. Personally I think is funnier although this kind of alternative thinking gives us useful problem resolution skills too. Also consider what Einstein said (in spite of the obvious gap between this genius and us):
“If at first the idea is not absurd, then there is no hope for it”.
I’m totally agreed with him. Only I would add:
“Reality will always put your ideas in the right place: at bottom (because the gravity force) :-)”
Just kidding. What I want to say is that thinking absurd ideas is not enough because we also need robust implementations; inside the limits of physics, mechanics, electronics and of course cost. It has to work and at a reasonably price. It seems obvious but difficult to achieve.
Good luck Nick. I you need something more don’t hesitate to contact us.
Regards,
Victor.