Itty Bitty Double Extruder

[beerdart]

Rob you can get a screen door spring and cut it to size.

Love the gold fan. Hey, what springs are you using on the idler? I'm using tiny rubber bands stacked on each side. It's adjustable but they stretch out an about a month and i have to replace them.

[RobH2]

Steve, did you have the belts walking off or are you just thinking it wouldn't hurt to prevent it?

I saw your cones too, I had the same issue. Where the filament changed and dove in I got a messy line of black. I've been thinking about doing a 'wiper'. Here's a link that shows one. I've seen a few variations that employ different mechanisms. We might need Roxy or someone to help get the code right.

https://www.youtube.com/watch?v=UF6Z...ature=youtu.be ...uses a guitar wire.




.................................................. .................................................. .................

[beerdart]

Yes I had a belt walk off during one print I moved the motor side pulley over to prevent it but adding a rim the the large pulley should prevent it altogether.

[beerdart]

For the wipe check out the way Cura makes a prime and wipe tower works good. But after switching to S3D I need a better way.

[RobH2]

Yes I had a belt walk off during one print I moved the motor side pulley over to prevent it but adding a rim the the large pulley should prevent it altogether.

Wow, really. The axis of the two shafts must be off just a bit or something softened and shifted. I might go ahead and print two gears with guards as a preventative measure so if it happens I can quickly change them. Although, if it's a chronic issue and the belt is pressing up against the guard, it might wear the belt out or cause other problems. It might be better to diagnose why it happened to start with and fix that.

I'll have a look at the Cura 'wipe.' It shouldn't be hard to do a good wipe system with the super-coders we have on this site. I'm just not good enough to write it. I can come up with the mechanics but not the code. To keep the travel time to the edge down I wonder if we could come up with a servo that brings a wiper in and then retracts? That would be fun. That might be difficult though in the first 5-10 layers when there isn't much clearance between the bed and the hotend orifice. Hmmm.... need to think about this.

[beerdart]

Dropped the pulley STL on Thngiverse.
http://www.thingiverse.com/thing:652354/#files

[paulpangrazzi]

So in playing with the Geetech one on D6/D11 I found I could just install it shifted two sets of pins to the right towards Aux-1 and now it runs on D4/D5, This also opens up D11 for the servo to connect to so I don't have to do any modding of the expander. Board type 34 works fine with it installed two rows of pins over. I would expect the RRD expander on D4/D5 might be able to be shifted one row to make D11 available to the servo and still have D4 for the fan.

This was a helpful tip, especially since I ordered the Geetech before realizing there were so many alternative flavors of fan extenders.

[clough42]

Yes I had a belt walk off during one print I moved the motor side pulley over to prevent it but adding a rim the the large pulley should prevent it altogether.

Interesting. SDP/SI has specifications for when flanges are needed, and this drive setup should be deep in the region that requires flanges only on one side. I generally turn the pulley in the normal forward drive direction unlit the belt settles in and adjust the small pulley until it stays centered on the big one.

I'd still try to adjust your drive so the belt stays in the center. Belts will try to climb to the highest point if they can. That's how most band saws control the blade position on the drive tires.

[beerdart]

Yes its staying in the center now it was only for one print that was sliced with a new slicer and it had wicked retract speed.

[paulpangrazzi]

Finally printed, assembled, and wired up on my Makerfarm i3V 10" with polulu 14 motors, hxt-900 servo, 1.75 hexagons, and a geetech fan expander. I have the heated bed on a relay. Using Repetier with the Cura slicing engine.
The hardware assembly all seems to work, (even after I briefly had the 12V jumper to the Geetech board flipped for a moment. Facepalm.) Extruders roll in the right direction, heat up and display temps. Print fan enabled at will. Golden.

But "auto bed level" printing is challenging...

1. What is a typical sample Start G-code sequence for this setup?
When I copied Clough42's sample provided earlier in the thread the ABL routine begins, but everything stalls when the LCD displays "Heating..." Nothing begins to heat*.
If I use Repetier/Cura dual extruder Gcode defaults, things appear to behave as expected, but after the G28 homing sequence (x, y, then move nearly to the center of the bed and z) the print begins at homing height.... in other words, above the bed. Confusingly, I was able once to print ON the bed, but now can't duplicate it. I need a proper procedure.
My probe offset settings in Configuration.h currently are:
X=22
Y=-7
Z=-2.11

2. I've disabled eeprom settings in Configuration.h based on other forum advice; mostly to eliminate variables in testing. Is this important?

3. I understand there are two ABL methods; grid vs 3pt... But I don't understand why or how they are used. The both appear to be enabled, but is grid invoked by G29? It appears to be, since G29 causes the probe to tap across the glass 9 times (3x3).

I realize there are other forums devoted to ABL; I'm posting here because there's likely to be specific solutions with the common parts/vitamins for this build.

Thanks.

*Edit: I realized Clough42's G-code samples from post #55 weren't meant for copy/paste. They were examples where the bracketed attributes [first_layer_bed_temperature] were supposed to be replaced with dynamic "S{BED}" or explicit "S110" temperatures. Sigh... brain working again.