Reversal Zits: SD Card Variations

Whenever what might possibly be a timing problem crops up, the first suggestion seems to be “Print from the SD Card!” because the communication USB link between the PC and printer is so fragile it can’t keep the printer fed properly. This seems odd to me and some recent findings indicate that the latency problem is largely due to a weapons-grade blunder in the Java stack, but I figured I’d try it to see what happens.

This rather garishly contrast-stretched image shows the zits along the left side of the right-hand critter’s dome, plus an assortment below their eyes. The one on the left was from the SD Card, the one on the right was through the USB link:

Octopodes - SD Card variations
Octopodes - SD Card variations

The zits may be a bit hard to see in the image, but I can’t see any difference in person, either.

For what it’s worth, I have never personally seen a problem that was resolved by printing from the SD Card. I think Linux does a better job of keeping the USB pipeline fed, despite Java’s baked-in RXTX delay, than Windows, but I really don’t have any data to support that. Other than that most of the complaints about USB latency seem to come from Windows users, that is.

12 thoughts on “Reversal Zits: SD Card Variations

  1. It might be the lighting, but I think the zits on the left print are slightly smaller. The zit which stands out the most as being smaller is the one which occurs almost directly below the left eye.

    I’ve also wondered about the latency issue, as a precaution I’ve always printed from SD. A handy side effect of this is when your £15 car boot sourced Dell pc driving your printer craps out, the print doesnt stop :)

    1. It’s the lighting… there’s a certain random difference in size, but there’s no “Wow! They’re gone!” level of improvement. The whole SD card apparat is basically an admission that an Arduino lacks enough juice for the job, although what’s going wrong here has nothing to do with that.

      Dell pc driving your printer craps out, the print doesnt stop :)

      One of the first times I tried using the SD card, I thought that RepG could do something else while not driving the printer. Turns out you can’t slice a second model while printing the first: RepG is really busy doing nothing!

      1. You should be able to disconnect repg after you start a print from sd card and then slice a second model while it’s disconnected.

        1. disconnect repg after you start a print

          Indeed, although that pegs RepG’s usability at a new low.

          What I’ve done is jiggle the upstairs PC with the same revision/config/setup files as the one running the printer, with the benefit that it slices at about twice the speed. At some point I’ll figure out a defensible way to put the config files in a common spot so one change propagates everywhere.

          Anyhow, that lets the printer’s RepG exert what little control/monitoring it does during a print, while I watch it on a webcam from upstairs. Not ideal, but it’s workable… and the printer hasn’t had another thermal runaway.

        2. Skeinforge can be run outside of replicatorg… or you could get an interface controller if you still want to control a print from SD card after disconnecting replicatorg. :)

          1. Skeinforge can be run outside of replicatorg…

            That’s pretty much the plan when I switch this thing over to EMC2. Until then, I may as well use the Official Interface.

            you could get an interface controller

            That box is directly from the Eighties. I’d rather get a touch-screen and run EMC2’s Touchy. Costs about the same and you don’t have to hack the firmware that’s also doing time-critical motion control to customize the UI.

  2. Ed,
    I believe the problem has more to do with the advancement of the filament after a retraction. The suggested speeds for retract are pretty fast and skeinforge uses the same speed parameter to prime the nozzle before the next print. Jamming the filament back into place so quickly creates a lot of inertia in the molten plastic. The only thing slowing it down is the restriction of the nozzle.
    Here are three possible solutions for your consideration:
    1) A slower velocity for priming the nozzle
    2) a graceful acceleration and deceleration of the priming (new reprap Mcode? [gag])
    3) over priming the nozzle by a fixed distance (0.01mm?) and retract it back to zero in an attempt to over come the molten plastic’s inertia coming out of the nozzle.

    Going from a .5mm nozzle to a .35mm nozzle has significantly reduced zit’ing on my prints and I attribute it to the extra back pressure from the smaller opening plus the reduced retraction distance. I am a little apprehensive about the distance of filament needing to be moved but my gut says that means less mass is accelerating and there for less energy is being added to the system.

    1. The zits appear larger at the end of the thread and smaller at the beginning, which suggests the retraction is either too slow or too delayed. More study is needed.

      What you’ll see over the next few days are examples of what didn’t have any effect, which should let me concentrate on what might work.

      I have a 0.4 mm nozzle on the shelf, but I am not disturbing anything before the presentation on 3D printing I’m doing in early September: Must. Print. Tchotchkes. [grin]

      a graceful acceleration and deceleration of the priming

      Notice how that subject keeps popping up over and over and over again? Insert standard EMC2 rant here…

      1. “I have a 0.4 mm nozzle on the shelf, but I am not disturbing anything before the presentation on 3D printing I’m doing in early September: Must. Print. Tchotchkes. [grin]”

        I don’t blame you one bit! I didn’t want to change from the 0.5 I was using but was forced to while rebuilding the heater portion of the extruder. I compared the extrusion diameters with a micrometer (new addition to the arsenal) and caliper which didn’t agree. The previous skeinforge parameters always needed tweaking when using the calipers, so this time the micrometer values were used. The mic also showed the filament diameter to be 0.1mm larger. The prints were nearly perfect out the gate.

        “Notice how that subject keeps popping up over and over and over again? Insert standard EMC2 rant here…”

        I quit talking… er listening about it and finally did it! [grin]It still needs a little tweaking but it has been hands down the best modification I’ve made. I was thinking a custom acceleration profile for when the extruder is primed, other wise the linear acceleration rates are perfect. Honestly, it hasn’t been enough of a bother to warrant the effort.

        Suggesting using something other than the open source reprap hardware rubs many of the core reprappers the wrong way. But if any one asks, I would never go back as it is too volatile. Since March I’ve seen the evolution of two hardware platforms, tried three different reprap approved firmwares with multiple subversions, four host programs, and wanted off the bleeding edge. Hopefully, for the sake of the 3D printer movement, the dedicated motion control system stabilizes soon.

        Oh, have I mention the bed has not needed to be zero’ed in two weeks?

        Good luck with your presentation! Consider suggesting the audience acquire a Xanax prescription before commencing their first build [grin]

        -CodeRage

        1. wanted off the bleeding edge

          My intent in buying the Thing-O-Matic was not to make a series of fundamental engineering contributions to the field…

          At this point the deleted printer pretty much does what I bought it to do: build widgets that aren’t fussy about their appearance. All I know is what I see in pictures, but its overall print quality seems right up there with the best of what you folks are doing. So I’m satisfied with how it works… even if that’s been a long time coming.

          I’ll put the nozzle in when I rip up the Thermal Core for some other reason, without expecting a world-shaking resolution improvement. For that, it needs a better mechanical design and better control software; both of which are do-able, but not right now.

          suggesting the audience acquire a Xanax prescription

          It’ll be the lowest-hype 3D printing presentation on the planet: on that you can depend!

          I’ll hand out tchotchkes, talk about what it can do, and we’ll have a lot of fun.

        1. miss the announcement

          Poughkeepsie being where it is, you’d have a long drive: it’s for the Mid-Hudson Valley LUG.

          I’ll get the slides (remember when presentations used actual slides?) up here shortly before I start talking…

Comments are closed.