Close



Page 1 of 2 12 LastLast
Results 1 to 10 of 20
  1. #1
    Engineer
    Join Date
    Dec 2014
    Location
    Canada
    Posts
    498

    Why does my printer suck.

    Ok well I have been tweaking and upgrading my Makerfarm i3v 12" for almost 2 years..
    ive become very familiar with calibrating and testing.

    done lots of little things to improve the performance and quality of my printer.

    But when I look at samples from other printers that dont look like really high quality machines.
    it really blows the prints im getting out of the water.

    Yeterday I saw a small delta machine with cheap looking parts.
    print incredible detail.

    and the layers were so smooth
    I realize a thin layer height will get smooth edges but the incricate detail quality was soemthing ive never been able to achieve

    This machine was likely running on a 32bit processor.. im running 8bit Rambo
    it was likely 1.75mm filament, im 3.0mm
    it was a bowden setup with some cheap extruder.. looked like a bulldog... im running direct using Bondtech extruder and e3dV6

    it was running Reprap Firmware. im using Latest Marlin firmware
    Im also using Simplify3d

    what am i missing exactly?

  2. #2
    I'm new to this forum and new to Makerfarm products (I have a Pegasus 10 that working beautifully) but I'm not new to 3D printing and certainly not new to CNC type machines. Two things I'd suggest for most of these units I've looked at is first and foremost is upgrade the Z axis to proper screws. Second get proper idler pulleys on the X and Y axises. (More important than most designers seem to realize IMO) Any structures made of wood (OK fake wood really) glue the pieces together if practical (Use good judgement because if you can't keep it squared up you are screwed once it's glued, hence "if practical") and go OCD on making sure everything is squared up and parallel. Attention to details is what separates the men from the boys when working with devices with fractional millimeter resolution. Frankly you are always going to be constantly tweaking any unit with a wood frame, even more true if it's not glued. I got the full kit when I bought a Pegasus 10 last month and the only piece of wood left on it is the spacer between the extruder and the X carriage. 8 bits is plenty to run a Cartesian format device, the math is pretty simple, the Delta format is an entirely different beast requiring much more complex math and then you can take advantage of the 32 bits. Now if you were going to include a touch screen controller and Octoprint all in a single controller package then 32 bits or even 64 bits starts making a lot more sense. A RaspPi 3 would be a good candidate for such a system although I'd modify it for an external WiFi antenna. The real Achilles Heel for the RAMPS is the lack of a proper RS-232 serial clock in the Arduino architecture 16 Mhz doesn't divide down to 115200 very well hence a jitter problem and the need to run at a out of spec baud rate of 250,000 (Which easily divides down from 16 Mhz and why it's more stable) The problem with 250000 is any driver or program that strictly enforces the RS-232 protocol isn't going to like it

    Other than generalities without seeing the prints to analyze the actual problems it's hard to be specific. I printed a full sized one of these T-800 heads 0.2 resolution in about 15 hours and it looks as good as any of the examples made on an Ultimaker 2 or a Lulzbot Taz and better than most of the others. I even did a manual color change, it's an eSun PLA+ black base and the skull is eSun PLA white. Granted I've been a Maker about 3 decades before they came up with a term for us so your mileage may vary but you can't beat all metal construction. Expansion/contraction due to temperature is minimal and from humidity pretty much non-existent. Don't feel too bad my first printer was a QUBD Two-up .......

  3. #3
    Engineer
    Join Date
    Dec 2014
    Location
    Canada
    Posts
    498
    I ended up upgrading the Z screws to fine Acme screws and run geared nema 17s
    the machine sits on a some ceramic plates to keep everything flat and level.
    i upgraded he bed to a aluminum plus boro glass & PEI



    Quote Originally Posted by longjohn119 View Post
    I'm new to this forum and new to Makerfarm products (I have a Pegasus 10 that working beautifully) but I'm not new to 3D printing and certainly not new to CNC type machines. Two things I'd suggest for most of these units I've looked at is first and foremost is upgrade the Z axis to proper screws. Second get proper idler pulleys on the X and Y axises. (More important than most designers seem to realize IMO) Any structures made of wood (OK fake wood really) glue the pieces together if practical (Use good judgement because if you can't keep it squared up you are screwed once it's glued, hence "if practical") and go OCD on making sure everything is squared up and parallel. Attention to details is what separates the men from the boys when working with devices with fractional millimeter resolution. Frankly you are always going to be constantly tweaking any unit with a wood frame, even more true if it's not glued. I got the full kit when I bought a Pegasus 10 last month and the only piece of wood left on it is the spacer between the extruder and the X carriage. 8 bits is plenty to run a Cartesian format device, the math is pretty simple, the Delta format is an entirely different beast requiring much more complex math and then you can take advantage of the 32 bits. Now if you were going to include a touch screen controller and Octoprint all in a single controller package then 32 bits or even 64 bits starts making a lot more sense. A RaspPi 3 would be a good candidate for such a system although I'd modify it for an external WiFi antenna. The real Achilles Heel for the RAMPS is the lack of a proper RS-232 serial clock in the Arduino architecture 16 Mhz doesn't divide down to 115200 very well hence a jitter problem and the need to run at a out of spec baud rate of 250,000 (Which easily divides down from 16 Mhz and why it's more stable) The problem with 250000 is any driver or program that strictly enforces the RS-232 protocol isn't going to like it

    Other than generalities without seeing the prints to analyze the actual problems it's hard to be specific. I printed a full sized one of these T-800 heads 0.2 resolution in about 15 hours and it looks as good as any of the examples made on an Ultimaker 2 or a Lulzbot Taz and better than most of the others. I even did a manual color change, it's an eSun PLA+ black base and the skull is eSun PLA white. Granted I've been a Maker about 3 decades before they came up with a term for us so your mileage may vary but you can't beat all metal construction. Expansion/contraction due to temperature is minimal and from humidity pretty much non-existent. Don't feel too bad my first printer was a QUBD Two-up .......

  4. #4
    I'm following this. I've always been underwhelmed by the quality of the parts I get out of my i3v 10 inch.

    longjohn119, are there pulleys you recommend?

  5. #5
    Quote Originally Posted by mattanonymous View Post
    I'm following this. I've always been underwhelmed by the quality of the parts I get out of my i3v 10 inch.

    longjohn119, are there pulleys you recommend?
    I got a pair of these from Amazon but you can likely find them even cheaper. Some people like the smooth bore pulleys rather than with the teeth, something about backlash with the toothed idlers but I haven't noticed any problem in that regard. I just know my Y belt isn't riding in a different place and flopping back and forth all the time which couldn't have been good for accuracy. Frankly I can't see all that much savings using a pair of bearings instead of proper idler pulleys. I can understand using normal threaded rod where there is a pretty big cost difference and it can work decent **if** the rods are perfectly straight and stay that way but that actually rare. Especially with 5 mm where just using too much pressure to cut them can bend them a little and it doesn't take much of a bend to show up in your prints. Personally I think the Acme lead screw upgrade Makerfarm offers for 60 bucks is a very fair deal for what you get and that includes a better stepper driver

    I like the DRV8825 Stepper Driver better because you can actually heatsink them properly. The type of IC package used is made to pull heat from the bottom of the chip to the back side of the PCB where there should be a fairly large ground plane and you put your heatsink on the back side. This type of IC does not sink heat properly from the top, it'll help a little but not a lot like the heatsink on the back side. The A4988 drivers aren't properly designed for heat sinking with most of the heat drawn through the bottom into an area that is boxed off and impossible to add a heatsink to. Putting a heatsink on top of the chip helps but not really all that much because of the way the IC is designed. The Pololu design will work but you need to derate the current by approximately 25% so instead of a max 2 amps it's more like 1.5 amps, anything more and you'll likely go into thermal runaway and burn them up. I wouldn't run out and replace all your A4988 drivers because even 1.5 amps max is fine for most designs but if one were to go bad I'd definitely consider the upgrade to a DRV8825

  6. #6
    I'm interested in this. After having my i3v for a year and a half, I'm not at all impressed with the design. I finally switched to a solid-state relay after burning through enough of the relays Makerfarm provides, and I'm tired of constantly having to tighten screws on the frame and tighten my belts. I'm trying to take a bottom-up approach to tuning in my machine and increasing the precision. I recently ordered ACME lead screws to replace the threaded rods on the z-axis, aluminum extrusions to make an all-metal frame, and aluminum motor mounts, carriage, idler, and bed. That will be my weekend project if everything comes in by then.

  7. #7
    Technologist
    Join Date
    Aug 2014
    Location
    Oklahoma
    Posts
    191
    Add usarmyaircav on Google+ Add usarmyaircav on Thingiverse
    Is it worth it to go from an I3v 10 to the Pegusus, and the ACME lead screws? I already have an E3dV6 Hotend? I need to switch to Simplify also.

  8. #8
    Engineer-in-Training
    Join Date
    Jul 2014
    Posts
    305
    Quote Originally Posted by usarmyaircav View Post
    Is it worth it to go from an I3v 10 to the Pegusus, and the ACME lead screws? I already have an E3dV6 Hotend? I need to switch to Simplify also.
    I actually have both a 12" I3V and a 12" Pegasus, both with stock machine screws.

    I have found them to print at the same quality except for the Z banding. I am having some banding issues with the Pegasus. But I figure those will be resolved with a little more tinkering.

    After spending a lot of time on my I3V and getting it squared away, it is a solid machine.

  9. #9
    I finally got some decent results out of my machine!

    Here's what I think helped:
    - I reassembled the machine paying close attention to things like the parallel aluminum extrusions being equidistant from one another at the top and bottom and making sure everything important was square and true
    - I printed and installed the 2 nut spring version of these http://www.thingiverse.com/thing:795954
    - I finally ordered and installed real couplers https://www.amazon.com/gp/product/B010MZ8SQU
    - I installed the pulleys longjohn119 recommended https://www.amazon.com/gp/product/B0188IFK4Q
    - I calibrated e-steps, set the filament diameter to an average of several measurements, and tuned the extrusion multiplier with this procedure https://craftunique.com/forums/view-thread/694/#7130

    No need for expensive upgrades like acme lead screws.

  10. #10
    Engineer-in-Training
    Join Date
    Jul 2016
    Location
    Pennsylvania, USA
    Posts
    255
    Hi

    Never underestimate the influence of filament and temperature on the whole process. In some cases very small changes can have significant impact on the results.

    That said, I'm *very* impressed by how the 12" Pegasus prints.

    =====

    That said, why did my earlier printers have issues from time to time:

    1) I didn't get everything level / parallel / orthogonal and keep it there. On some printers that was a lot of work and a lot of repeated work.

    2) I didn't keep everything tight (in terns of belts) and loose (in terms of friction and binding).

    3) I switched to the latest version of XXXXX and didn't get it straight before I tossed the latest version of YYYYY into the mix. Firmware / software bugs (or settings issues) are a pain when they start to pile up on top of each other.

    4) I tried to run things fast, convinced myself that "it's fine" and moved on. I spent a lot of time chasing things that actually were speed related without dropping the speed.

    5) This filament was fine with this speed, that temperature, and when it was new. Now two years later it seems to not work as well after I forgot all those settings and it's soaked up a ton of moisture and beer fumes....

    Simply put, there are blind spots in the process. Everybody has them. If you look at my printer, you will see things a bit differently than I do. You can't give me a list, because it's just the stuff you see. Even if you gave me a list, the list is based on your background and tools. Your list doesn't work for me. Equally, my list isn't really of a lot of use to you.

    Bob
    Last edited by uncle_bob; 09-11-2016 at 09:01 AM.

Page 1 of 2 12 LastLast

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •