User:DaveX

My Wallace Build

My Wallace is currently in a mostly assembled state with these pieces:

• Wallace, NEMA14/6mm Printed by http://www.fabbr.com/
• Pololu steppers per http://www.pololu.com/catalog/product/1209 1A, 1400gf-cm (20 oz-in)
  • Some (too-small, but shimmed with paper,) PLA LM6UUs per http://www.thingiverse.com/thing:16464
  • Replaced the LM6UUs with steel ones, but they still seem pretty stiff. In trying to get the Y axis to move, I think I burned out a stepper.
  • 350x300x430mm smooth rods makes a build volume of 172x218x22 and an overall envelope of 405x531x442mm (per notes in Wallace_Build_Manual#Dimensions, and measurements shown in the pictures at User:DaveX/WallacePics
  • Pegboard bed: 274x313mm
• Greg's hinged accessible per http://www.thingiverse.com/thing:8252 hobbed with an M5-ish SAE 10-32 tap per my notes at Talk:Wade's_Geared_Extruder#Filing_and_hobbing_the_bolt
• A shorter Wildseyed_Simple_Hot_End, since the air-fitting's locking fin seemed small compared to Wildseyed's pictures, per Talk:Wildseyed_Simple_Hot_End#Air_hose_Adapter_machining
• Pololu_Electronics#Teensy_Breadboard using Pololu_stepper_driver_boards running ATMEGA32U4-ported Teacup
• Old ATX power supply
• Initial plastic: Natural PLA from http://thefutureis3d.com/node/113 through Ebay
• Pololu_Electronics#Teensy_Breadboard
  • Thermistor circuits modeled after Pololu electronics
  • Heater circuits modeled after Pololu electronics, using $3.34 42V 49A Protected MOSFETs from [[1]].
• Wallace, ATX and electronics mounted on a 18x24" piece of MDF
• Put the Z steppers in series, so x,y,z circuits all use similar current: 0.7A drives the axes with cool steppers and Pololus. Pololus tuned to Vref=0.28V for 0.7A peak current, 70% of stepper rating, for 70% of 1400 g-cm torque and 50% of motor heating.
• Setting microstepping for XYZE as 4/4/2/4 for noise and speed in Teacup on a Teensy
  • Determining max feedrates. (x,y,z,e)=6881,6881,233,582 with above microstepping. Teacup stutters on interrupts if you set the rates too high. See [[2]]
• A version of Wildseyed_Simple_Hot_End]
• Big delay due to several life events.
• Wired up the extruder and hot end assembly with a ribbon cable using Nophead's male D9 connector per Mendel90_Build_Manual#Extruder_Wiring and [[3]]
• Filed flats on my steppers & re-fit the connectors and pulleys for less slippage
• Teacup ported to Teensy 2.0 / ATmega32U4 on https://github.com/drf5n/Teacup_Firmware on the Gen7 Branch. Many changes pushed upstream into Teacup
• No endstops for now: I jog to the center until a piece of paper binds and then start with G92 x0 y0 z0.1
• Filament reel hanging from a long L-shaped stick wedged behind the desk.

Things to do:

• Tune things.
• Document Pololu_Electronics#Teensy_Breadboard as Teensy Breadboard.
• Test the hotend temperatures. My PLA seems too hot and liquidy at a setting of 185. It seem to flow and stick at 165.
• Assemble and install a heated bed

Mistakes

• Don't assume a metric rod in the US is 1m long.
• Don't overcurrent the steppers. Pololu's drivers can do a nice job of limiting the current, but can overdrive the 1A/coil steppers if you want them to. I think I had damaged my Y stepper earlier by trying it on a circuit for the higher current extruder, then since I was fighting too much friction on Y with the new steel LMUU6s, I upped the current until a new failure: odd pulsing due to thermal overload of the Pololus. By adding a heatsink I was able to increase the current further, compounding my problem, as I still didn't get motion on Y. Then I noticed that the holding torque on X driven by a much lower current was much stronger than the holding torque on Y. My Y stepper was weak and feverish, compared to my strong and cool X stepper. Bummer. I ordered a replacement and spare. In the future it might be good to Thevinin-ize the stepper and set the current with a dummy resistance load. A couple of fat resistors would be much cheaper than a stepper-shaped heater.
• Don't strip the nut-trap on the Z-axis couplers.
• Do file/use flats on your stepper motor shafts. They don't need to be much, but without flats the pulleys and couplers might slip.
• Make the hobbed bolt groove well defined so the filament does not ride up to different diameters and change the steps/meter extrusion calibration.

Calculations

# R code for estimating feed rates, etc.:

stepsPerSec <- 15570 *16/20; # per http://reprap.org/wiki/Teacup_Firmware#Distances_as_expressed_in_steps for a 16mhz chip
stepsPerM <- c(13576,13576,400000,160423)
maxFeedPerMin <- stepsPerSec / stepsPerM * 1000 *60 
safetyFactor <- 1/(4*2)  # divide by 4 axes, and factor of 2
maxFeedPerMin * safetyFactor    #  == 6881.2610 6881.2610  233.5500  582.355 mm/minute
maxFeedPerMin * safetyFactor /60 # == 114.687684 114.687684   3.892500   9.705591 mm/second

maxMMperMin = c(6881,6881,233,680) ## Teensy w/ 4,4,2,4 microstepping on a Wallace 2012-10-13
#Steps/sec at max feeds:
maxMMperMin/60*stepsPerM/1000  # == 1556.941 1556.941 1553.333 1818.127

# Resolution:
1/stepsPerM*1000 # == 0.073659399 0.073659399 0.002500000 0.00623352 mm/step