1st one completed, nuts, nae nuts Trip to Toolstation later ......... Top edge of the bearing is flush with the bracket base - YES !!!! Started printing the next one. Paul
As that was successful for the center I have grafted two together to make a double bearing version Now printing, as the single took 41 min to print, the double should take approx twice the time, but I'm also printing two copies at the same time Ah well I may just have to watch the football instead Paul
Very interesting thread, very much up my alley (not that I want a traverser), but I am getting into Arduino's myself (with lights and servo's) and I am sure one of my next steps in the future will involve stepper motors. I design industrial servo systems for a living (amongst other PLC based hardware), so might I suggest you run the Homing move at full tilt, by adding a second micro switch as a pre-home (say 5-10mm before home) and only run slow between pre-home and home (the gap needs setting to the overrun of deceleration) This simple addition (of 1 pin and a micro) would make the home move sweet! You might also think about tensioner pulleys for those long belts?
Hi TimberSurf Thanks for the sugestions, I have actually bought some proximity detectors to do the same as you suggest, but I will need to see if they are repeatable enough to use, and if so may get retro-fitted to this traverser (the mk 2), if not the extra micro switch could get used instead.. The new version (mk 4 - I think) will be using the fully supported linear rail - they have arrived see below for photos. This is a change from the mk3, where I had 3D printed brackets for twin shafts and the lead screw all in the one assembly, there was still some flexing, and by the time I added up the cost of rod, bearings and the time printing the parts another solution was looked for, so the mk3 is consigned to the bin - well actually Toto may get bits for a mini traverser for his shed. Stainless steel rail, with an aluminum support track The rail is held in place with cap screws, accessed from the underside. Each rail is supplied with two SBR12UU Block linear bearings I've also ordered Additional Lead screws with Anti-Backlash nuts plus Anti-Backlash nuts for the existing lead screws. Hopefully all the parts will be here within the next week so I can start and assemble them. Paul
It's hard to tell from the pics, but is the aluminium backing on the base of your system? If it is, then you only get the strength of the vertical aluminium, the whole point of supported rails is not that they are more rigid, but they can be attached to a rigid frame that will not bend. You would need to mount them on their side to steel! thus you take on the rigidity of the base steel and not rely on the rigidity of the rail and aluminium. The other option is to use twin rails (again on their side) for integral rigidity (as you only have a wood base), but again the price may put you off, I am not sure how it would compare with two round rails costing. In industry, we have gone on to shaped rails with fixings every few inches to a rigid base, but I'll bet they are out of your price point.
Point of information, proximity switches work as follows, the smaller the diameter the more accurate the switch point, the smaller the the sensing distance the more accurate the switch point. i.e. M12 with 10mm sense distance will only be repeatable to 2-3mm M6 with 2mm sensing distance will be repeatable to say 0.25mm It's actually hard to beat a quality micro switch for repeatability (without being in danger of getting the prox so close, its in danger of being clacked if misadjusted!)
Currently the rails were resting on a piece of ply for photo's, but will be mounted on a plywood frame, which if needed could be reinforced with steel tube or angle. I did look at Hiwin rails, but as you say the price starts to get prohibitive, especially as I have two to build by October. The mk1 traverser - OO gauge used on my Victoria Road layout, used PC keyboard runners, they worked well enough. I did consider using drawer / server runners for the mk2, however the spec was for a 20 inch wide traverser, with four tracks, all being able to access the middle entry exit point. This resulted in the traverser table going beyond the traverser frame and thus the posibility of the bearing races of the drawer runners falling out. So the mk2, just uses 8mm rod with linear bearing blocks, which for my O gauge stock is fine, but with Toto's Heljan loco's at several kg each they sag. However it runs very smoothly. The mk3 was then designed with twin 12mm bars at each end, and also included the lead screw assemblies as well, looked good and did work, but sounded rough, but that was probably down to using steel tube rather than ground bar. Which brings me to the mk4. From your comments above, you are saying I would be better putting a verticle cross brace across the traverser and mount the supported linear rail to that, rather than mounting the rail on a horizontal crossbrace. Paul
Yes! Like the side of a baseboard, the larger the span, the deeper the side frame needs to be. Rails on their own only have their own rigidity (and surprisingly flex more than you think when loaded in the middle of a long span). Supported rails are not made to give extra metal to stop flex, they are designed to transfer rigidity from the substrate to the rail. Max rigidity would be gained from the twin rail (it has inherent rigidity if mounted vertically and gains further rigidity if mounted to a steel substrate) You could add a flat steel bar (1/4 x 2in) to the {on it's side mounted} supported rail base then attach that to the vertical stringer.
Toto, if you are the Arduino wizard, you don't get off lightly either! Have a look into "S curve" acceleration and deceleration, TimberSurf Lumsdonia <--- Hit the link to go to my website for full story and wiring advice!
We have accel and de-cel covered, by using the AccelStepper library, you only provide the destination (in steps), the library knows its current position, so will calculate the number of steps, direction, and based on the defined values at startup the accel and de-cel for the move. Keeps the programming nice and simple. Paul
Works really well, this link may be of use Sparkfun Big Easy Driver this is the stepper controller I board I have used. There are the board / controller specs on the third tab. The library will work with any 4988 based controller - available cheaply as stepper controllers for 3D printers Motor shield and controller A4988 Controller One feature of the Big Easy Driver I found very useful was is its regulated 5V output, which I've used to power the Arduino and ancillary components, with the board being supplied with 16V DC / 7A (ex Laptop PC power brick). Paul
Fun and games today ........ what's new there then I hear After spending the last few weeks trying to find my PowerCab controller - it got to the point that we weren't sure if had ahd it or it was at Toto's, but I found it behind some furniture moved for my Mum's visit at christmas. So this morning was spent locating power supplies and making cables up to connect the PC via usb, power supplies and DCC track bus. Powered everything up and it worked first time - as if it was supposed to I swapped out the nano for a spare, and loaded up the sketch (program) and it worked. Now to adjust the positions as the track had been moved for Luib Bridges rebuild. Entered the new setting and not a lot of effect, even halved one of the settings to no effect Now I remembered that I had upgraded the Arduino software to version 1.8.0, for the Tutorials, and the traverser was written before then. When you upgrade the software it deletes the previous versions, so I didn't know which version I'd used Decided to download and install version 1.6.0, which was the last major release before 1.8.0 (but probably newer than the one I was using). IT WORKED Adjusted the settings and seems good. Next up the other end. I will try and post a video later - read that as tomorrow Paul
That's moving along swiftly. it certainly looked nice and smooth earlier when I had the chance to see it in action on Skype. Great progress Paul. Cheers Toto
Oh heck I'm gonna have to look into this 'hard we know' thingy whatsit as I nearly passed out trying to fathom this thread maybe later
Not quite Toto I was thinking more along the lines of an automated corkscrew.. just as a simple 'exorcise' Ian vt