Stiffer frame

This topic contains 77 replies, has 11 voices, and was last updated by  Johnny 1 year, 4 months ago.

Viewing 30 posts - 31 through 60 (of 78 total)
  • Author
    Posts
  • #62406

    Jason
    Participant

    Ok. I think if we added a bit of plastic where my finger is sideways, and added a 6/32 in the two places I’m pointing vertically it will make the gantry a bit more stiff. Not sure how much but I could mock it up with a block of wood for the gap and simply drill the holes, and use some washers.

     

    #62410

    Ryan
    Keymaster

    You have to try it first.

    Use the machine.

    You have many many hours worth of CAM to learn before you worry about changes to any hardware. So many other things that would be a much better use of your time than guessing about the structure. You have CAD, CAM, bit selection, material hold down, prices, time involved, material costs, dust, noise, tolerances, proper part design, etc.

    Please, please try it before you bother with any of this.

    How many ounces of force caused that part to move?

    #62411

    Jason
    Participant

    I’ve run a mill and lathe before and have used solidcam and aspire. I have some computer programming under my belt and helped port linuxcnc from Debian to Linux mint and Ubuntu.

    If it wasn’t for the fact I have a few days off printing left on three printers i would finish this.

    I need a case for the controller, reprint a bottom corner, dual endstop parts, and z max. I’m waiting for the LCD.

    You can ignore my posts if you like.

    However I mocked this up with nylocks inside and it helps. It wouldn’t take much to add some screw ways and a spacer.

    I should have added one more facing down but forgot. Still helps a bit though.

    #62414

    Ryan
    Keymaster

    There is a 5″, 5/16″ bolt right there, and 4 flanking that. That really should see no movement, if built correctly. For my builds that is not the source of flex. Matter of fact I can not make that part move in any way even picking the machine up by the gantry. I have one directly next to my desk.

    If that really does move for you try a drop of CA glue top and bottom, or try a rebuild of the gantry it should not move.

    Again, how much force for the tip of an endmill does it require for that to move. I have a deep down gut feeling you are really seriously over estimating the cutting forces the machine experiences. You seem to be looking for a smoking gun, it is just a balanced build. Better rails won’t help without more rigid parts, more rigid parts are larger, larger parts require more power to move, more power is more expensive. A few upgrades and you have upgraded right to a tormach, the small ones are cheap.

    I urge you to enjoy having the ability to have a CNC machine at your house that can cut aluminum without taking out a loan and running 220V or 3 phase power. Everyone who has commented in this thread here has pushed the boundaries and made me second guess the machines capabilities on a weekly basis and I love it. Front page, Omri has used it with a fashion designer! Barry 5’x5′ build, Dui 18″ Z and all sorts of other stuff. Proving something works, with actual proof is amazing, I love it and hope to incorporate improvements along the way that make sense to the project as a whole point just a niche use. That is the whole point of being as open as I am with my designs. I encourage people to call there changes “mods” and make them available so others can decide to use them or not. If I think it is a poor choice for some reason I will contact you privately to voice my opinion so you can chose to make changes or not. If I love it I will ask to include your change and usually name it after you or whatever you prefer, the “heffe plate” is one such change. Making a thread full of guesses of all my design choices honestly casts doubt on the design as a whole for the less experienced and will most likely will push them to try something else that may or may not be better.

    #62417

    Jason
    Participant
    • Casting doubt on your designs, is not my intention.
    • I am looking for such mods, and it’s not like they are all in the same place. I have seen many on thingiverse, but it is imposible to know which have actually pushed the boundaries and which are just designs.
    • Again I originally wanted to have a cutting area of 2′ by 3′, but it seemed as you had a protest to this, citing the availability of the lowrider.
    • I’d love to have a larger z axis of about 5 1/2 inches.
    • Right now I’m hearing a lot of what not to do, which is fine. It’s helpful.
    #62419

    Ryan
    Keymaster

    The absolute best thing to do is try it as I have laid it out. This is the best collection of parts in my oppionion, I stakw my income on this fact. I address how I handle mods in the previous reply.

    If you want to change something you have to have a baseline to compare it to anyway.  Have a little trust in my oppionion. When the time comes for your mods I will return the favor.

    #62420

    Jason
    Participant

    Where are the mods at again? I must have missed that part. Sorry.

    #62421

    Ryan
    Keymaster

    Thingiverse, search mpcnc

    #62425

    Jason
    Participant

    We’re just back at guessing as to which will really improve the machine. I see a belt tensioning mod, but as I understand it, tightening the belts too much is a bad thing.

    Maybe you could add the mods you approve of in a collection label mpcnc mods on your thingiverse. That way there is no confusion as to which mods you support. 🙂

    1 user thanked author for this post.
    #62429

    Jason
    Participant

    Also, where we’re we’ looking to find stainless tubing so cheap? What was the wall size again?

    #62431

    Douglas Bell
    Participant

    @jason

    Hi, I have had my MPCNC built for a while now as per Ryans design exactly then with a slight modification, changing from open legs to an enclosed base.

    Any issues I have had and have are all due to poorly printed parts on my part, including some of the flexing issues you demonstrate above.

    I lived with these issues for a while cutting MDF parts up to 25mm and its had no bearing on the actual cutting and accuracy of this machine in the Z axis, however a poorly printed and assembled Gantry had huge issues on my accuracy in the X and Y which I had to quickly fix by reprinting parts and better assembly.

    My machine is 1200mm x 900mm total with a Z travel of 80mm and so far I have no issues with stability using the 25mm galvanized tubes I have, no supports added along the tube lengths and no tube infills.

    I have cut MDF, Plywood, acrylic and some test cuts in Aluminum.

    I do have issues with Aluminum but that’s more to do with not getting all settings dialed in for Aluminum.

    My accuracy with my spindle is about +/- 0.5mm and my laser is about +/- 0.2mm

    Any issues with stability were always down to trying to cut too fast or too deep, especially depth for me depending on what I was cutting.

    Hope this may help

    D

    2 users thanked author for this post.
    #62448

    Ryan
    Keymaster

    Maybe you could add the mods you approve of in a collection label mpcnc mods on your thingiverse. That way there is no confusion as to which mods you support. ?

    I do, all listed on the parts page. Other things you find on TV are either for a specific use, or I prefer my design for one reason or another. If I like something I ask the creator if I can use it in the official parts.

     

    Also, where we’re we’ looking to find stainless tubing so cheap? What was the wall size again?

    I bought mine at Alcoa, thickness is on the parts page and FAQ’s, but is very easy to use any size really.

     

     

    #62454

    Jeffeb3
    Participant

    I found the cheapest tubes to be from a local place (which is a national chain) called “metal supermarket”. There is almost certainly a metal supplier in any medium to large city.

    #62456

    Jason
    Participant

    Yes, it does help.

    I seem to remember having an issue with the roller plate due to using rectilinear infill. For whatever reason there was more flex on one side then on the other.

    Of this is the case I may have enough pieces to reprint to warrant upgrading to the s.s. tubing if I can find a good price on it.

    In any case adding the 6/32 support system seemed to help.

     

    #62463

    Ryan
    Keymaster

    Yes, it does help.

    Dude. I can only ask nicely so many different ways. This statement right here, do you know what that is? That is a rumor that will grow. You just said your part was better, even after I asked you to prove it many many times.

    You can very easily prove it with a scale or an actual improvement in your cuts. Nope. You very boldly said you improved the design. No one other than me cares if you prove it or if it actually does anything. Now anyone that sees that one little sentence will ask for your better parts.

    I seem to remember having an issue with the roller plate due to using rectilinear infill. For whatever reason there was more flex on one side then on the other.

    The roller plate, it has zero structural functionality, it holds and endstop and prevents the belts from falling off of a poor build. The part isn’t actually needed.

     

    #62476

    Jason
    Participant

    I miss quoted the part. Sorry.

    I’ll update when I get back home. It’s been a few months since I printed out the parts. Obviously that part will have no function in the rigidity of the z axis. You can see which part I’m trying to stabilize in the picture.M

    Mystatement of it helps, is in regards to the fact that one of my parts, whatever it’s name is, may be of poor quality.

    It’s the part that holds the z axis bearings. One side has quite a bit more flex than the other side. So yes. Adding the support helped.

    Yes, it helped.

    Yes, his comments helped. You must be miss interpreting me here. He ended by stating I hope this helps. His comments made me remember that I was curious as to why one side of that part flexed more than the other. My conclusion is it’s an issue with my infill.

    Thus, I realize if I’m going to need to reprint many parts, I may as well upgrade my rails, provided I can find some at a reasonable price. Where did you purchase yours from for so cheap, Ryan? Did you quote 42 dollars?

    I haven’t been using a weight scale to measure deflection. You can do it with a known static weight and measure actual deflection, so doing that side ways isn’t as easy. Pulling at the z axis with my fingers a pound or so caused a bit of flex( in the part I’m saying may be a bad print) I could visually see.

    Again, no reflection on your design, just noticing this on my build. Again, I may need to reprint a few parts.

    Can’t seem to post. I hope this doesn’t get posted double 🙂

    #62483

    Ryan
    Keymaster

    Pulling at the z axis with my fingers a pound or so caused a bit of flex I could visually see.

    Lets go with this real quick. At what Z distance? Torque=force * distance, hence the half the length twice as strong statement I keep making. You had all the figures and links earlier if you want to discuss this you need to still with complete information. F=ma,  think 1lb, concentrated down to the surface area of the bit that is in contact with the surface of the material about 1mm squared and a majority of that energy is kinetic rotational. What would one pound balanced on the tip of a pencil feel like? So maybe just stop with the over analyzing until you try it, at this point I feel like you will not see the complete picture and just keep focusing on one tiny aspect and missing the complete picture. This is not a math problem in a book this is a real world system. Every part depends on another part. FEA can not even be accurately modeled because 3D printed parts are to variable.

    This thread is a very negative one. Too many assumptions, no evidence, no numbers, all guesses.

    People are trying to help you improve a build that you have not tried to even use yet. You can see this might be an issue. You say your part flexes unevenly, it shouldn’t it is a symmetrical part as is infill, but how do you know that it flexes too much? You are blindly stating you have too much flex but have not proved that. Use the machine. If you can not mill pine wood at 4mm DOC, at 8mm/s, with Dewalt 600W DC motor, single flute up cut bit you have a major problem. You are using a belt driven brush less motor that will be belt driving a collet system with some ratio. We will have nothing to compare it to.

     

    It’s the part that holds the z axis bearings. One side has quite a bit more flex than the other side. So yes. Adding the support helped.

    You go through explaining what you quoted and that I mistook it, and then reinforced your own assumption. So now your rumor is here twice.

    Who cares if it helped some sort of flex you may or may not have. Does it do anything to a real cut? Please be more considerate.

    Early on in this thread you felt the need to tell me solid rod is stronger than tube. You did not complete the statement by adding the increase in weight that comes with that 2.8x’s increase in rigidity and the reduction in accelerations this causes, the price increase, the difficulty in cutting that material, as well as difficulty in finding it. You keep doing this. Cherry picking information you like and not completing the system. I keep filling in your gaps.

    How different would this thread be if you would have built a machine with your braced tube or solid rod, made some cuts and shared your findings. Side note here, there are several build that incorporate solid rod, and/or braced tubes in the forums.

    #62492

    Jason
    Participant

    I feel your right about this being a very negative thread. Not my intentions but I don’t feel it’s entirely my doing.

    All I wanted to do was strengthen the outer rods. What’s so different from doing that than Douglas Bell’s enclosed build or using 1″ tube? 3/4 emt is pretty flimsy in comparison. It was expressed early on that a 2′ x 3′ cutting area was too big. I gathered that by your comment suggesting I build the low rider.

    In the examples given other who built a machine of comparable size built with thicker pipe.

    You quoted me saying one side of one of my parts had more flex. I openly said I believe it to be a poorly printed part. I’d return your request here to be more considerate.

     

    #62497

    Ryan
    Keymaster

    What’s so different from doing that than Douglas Bell’s enclosed build or using 1″ tube?

    Because they all tried it and told us how it went, there was no second guessing every single aspect of the machines design. Actual real numbers, and usually an odd use case to go with it. Totally awesome!

     

    3/4 emt is pretty flimsy in comparison

    Dude, stop it. Unsubstantiated claims on inadequacy. There are thousands, literally, of 3/4″ builds. There is no reason to build something out of carbon fiber and titanium if $4 EMT is enough.  If a spaghetti noodle worked and was cheaper I would use it.

     

    Not my intentions but I don’t feel it’s entirely my doing.

    You are making me defend my design. I will. Always. This is my actual full time job, for years, my income depends on it. I take it very seriously. When you make statements that are false, or untested, I will be here to correct or point it out.

    #62506

    Jeffeb3
    Participant

    I think it’s the perfect time to just forgive everything that has happened so far in this thread. I think you both understand each other much more, and if you still wanted to have this conversation starting tomorrow, ideas would flow much easier, and with less conflict.

    #62508

    Jason
    Participant

    No one said 3/4 emt wouldn’t be adequate in a machine. I’m not attacking your machine. Your jumping to conclusions and being a jerk. You quote me outta context. Thanks

     

    #62518

    Ryan
    Keymaster

    Yikes, I am sorry it came across that way, obviously I am better at CAD than putting the right words together. I assure you no ill feelings were meant. I’ll just stay out of your thread.

    #62535

    Jason
    Participant

    I will share this with you guys. The original parts I printed we’re in standard abs. Everything was a little tight, and besides these parts I’ve had issues with layer bonding and warping occasionally.

    I changed to esun abs+, and after messing with it for awhile, my parts come out fitting like they are supposed to. When I first started using it, I treated it like abs, but it didn’t seem as strong. Slowing the print speed down made a huge difference. This may be the case with standard abs as well, but slowing it down usually resulted in curling.

    I didn’t print a hook and load test it, or anything, but it seems a little stronger overall. Maybe one day I’ll test it to be definitive.

    #62543

    Jason
    Participant

    I’m really not finding 24ft of stainless for less than 100$. 8ft is about 35$ or more.

    Am I wrong? Maybe I’m not googling right?

    #62544

    Jeffeb3
    Participant

    Anyway, it’s been a while since I bought my steel for my low rider, but I do remember it being way cheaper offline (even at metal supermarket). Just calling a local shop and asking them is worth it. Asking for smaller pieces could also help.

    It could be different now.

    #62546

    Ryan
    Keymaster

    Yeah it’s all in the shipping anything over 18 in gets really really expensive you have to find it locally.

    #62549

    Douglas Bell
    Participant

    Hi All

    Sorry if me contributing to this thread has exasperated the situation a little.

    My goal was just to let @jason know that I followed the plans as described using the size chart to make my machine the size I wanted, and that any issues I had were not with the design but my poor construction / printing.

    My cheap 25mm Galvanised tubes cost me about £8 for 3 meter lengths, I didnt think they would be up to the task, but even at 1200mm long I have no issue with them, yes they flex if I push down on them but I am putting a lot more force downwards than my machine will ever do.

    As my comment on an enclosed base this one done a long time after using my machine and was more that I was fed up hoovering my floor as I don’t have a vacuum system on my machine.

    Nos

    #62564

    Ryan
    Keymaster

    Doug, no way, don’t feel that way about anything. No situation here, just a misunderstanding of emotion over text.

    #62569

    Jason
    Participant

    Hi Douglas. I didn’t mean to drag your name in, in the way it came out.

    I realize I must be the umteenth person to try an “fix” the emt. I’d like to say though, I never said there was an issue with the machine as it’s designed. Like you said, the rails do flex under pressure. I saw that and wanted to stiffen it up. Whether or not it will help the machine is another question.

    This morning I pulled off the new xyz part ( this is the correct name for it) from my printer. It does not flex like the old one. I believe infill patern is to blame.

    For the weak part, adding something to stabilize it did exactly that. I didn’t realize how weak it actually was in comparison with this new part, until I held them both in my hand.  I assumed once I bolted it all together it would be fine, but there’s no substitute for a properly printed part. With this new part I don’t see a need to resupport it.

    @heffe

    That’s where I started, but locally they are using internet prices to model their prices. They cater to big industrial companies and can get away with it. Bummer.

    Just thinking out loud, but I bent 3/4 emt and 1″ emt daily for a few years out of high school, and 1″ was pretty tough in comparison. I wonder if the o.d. would work. This is just thinking out loud. Nobody has to go look it up, I’ll check later.

     

    #62570

    Jeffeb3
    Participant

    1″ OD is yhe biggest the parts get, so 1″ EMT won’t fit. Complete rework for that.

    The low rider uses fewer tubes, BTW.

Viewing 30 posts - 31 through 60 (of 78 total)

You must be logged in to reply to this topic.