Artemis - Build Notes

[wepollock]

Thanks to the SeemeCNC crew for shipping the Artemis.. First impressions are that it rock-solid construction and a leap beyond Rostock V2. The packaging is great as well. Compared to the Rostock Max there does not seem to be as much to assemble. I am going to take my time on the build, building is fun and I don't want to bang it out.. Today I did the unpacking, snaked the Z wires, and terminated one side of the Z wires.

I started this forum to document my build notes, mistakes I have made, and suggestions for documentation refinement.

Assembly Tip - make sure the Z column is free of packing material before you snake the wires through.

Assembly Tip - Bouncing a Meter to trace a wire

I think it is good idea to have a cheap ohm meter on hand.

I snaked the Z wires through the Z-tower column and realized that I had to pair one of the three white wires with the red and black wires (tri-pairing of red,white, and black). The other two white are thermistor wires for the bed.

The white thermistor wires have no connectors on the base-end, while the RED/Black/White wires have molex female connectors at the base-end. I am assuming these are FSR wires. It might be a good idea to supply different color wire in place of the FSR-white-wire.

To correct the mistake I bounced an ohm meter. I set my ohm meter to 2000K and crossed the leads, when the leads cross continuity makes the display bounce. I tested each of the snaked white wires to assure that I was pairing the correct white wire with red/black.

My Mistake in Assembling the First KK Connectors.

Instruction 11 might benefit from a few clarifications.

1) There needs to be a note earlier on in the instructions that PIN 1 is indicated with a "1" on the connector. It is difficult to see if you are not looking at it closely, but it is there. The picture in instruction 10 is not clear.
2) Suggest that Instruction 12 be broken up into two parts; a section for the four pin connector; a section for a two pin connector. Its not intuitive that three "paired" wires are split into two connectors.
3) Better pictures of the correct orientation can be found in Step 31 and other later steps.
4) Generally, people are "not going to look ahead" in the assembly instructions. This was also a concern with my Rostock V2 build, clarification was found in later steps. (a few steps later you would see that you made a mistake several steps earlier)

After I noticed I bolluxed the connectors, I was able to reorient them.. The spring in the connector can be gently pushed down from the reverse side. However, I did damage two connectors. Thankfully additional crimp connectors were provided; so I stripped the wires and used my crimping tool/die to replace the damaged contacts.. Most people won't have a crimp tool so Seemecnc might want to consider providing some spare Z wires.


Reading Ahead Instruction 15 and 16 would benefit from having an overview picture and then being split into parts.

There are six total connectors in total;
4 prong for the extruder
2 Prong - LED: Green Wire - 2 pin KK connector - PIN 2
2 Prong - HE FAN: Orange Wire - 2 pin KK connector - PIN 2
2 Prong - LAYER FAN: Orange Wire - 2 pin KK connector - PIN 2
4 Prong - PROBE: Red PIN 1, Blue PIN 2, Black PIN 3, White Wire PIN 4 - 4 pin KK connector
2 Prong - E0 TEMP: 2 White Wires - 2 pin KK connector - PINS 1 & 2 (Polarity DOES NOT matter)




Assembly Tip - Tools Used So Far

So far I have used an ohm-meter, wire crimper, wire stripper, and I used small zip-ties a 3mm threaded rod to snake the wires through the Z column.


Assembly Tip - Make sure the KK and Molex connectors are locked in place

KK, Molex, and Dupont contacts have pins which click into place (in the housing).

[wepollock]

This morning I am putting on the SL connector on probe, the stepper connectors, the tower wires, and the fan wire.

Screws were not holding on the top of the electronics box; so I removed the top easily. The metal is thick gauge! Everything is laid out brilliantly! I checked to make sure that the frame was in fact grounded to the external power supply case AND IT WAS! <-this is a very good thing that I have seen people omit in their own custom builds. I wonder if all these cheap Chinese 3d printers are grounded! I really appreciate the thinking that went into safety.


Suggestions

1) The instructions pictures are color coded (not to an outline but instead) to text within the paragraph that indicate the same eg. "in the picture with the green circle"

It would be easier if the legend in the picture matched up to the outline... Picture 1.1 Outline 1.1 1.2 1.3 Picture 2.0 Outline 2.1 Picture 2.2 Outline 2.2. ABC would be easier A in the outline A in the picture. It would be easier not to have to look into the text, and the color coding is not intuitive.
2) It would be nice to have a 12v power out 2.1x5mm jack CCTV jack or two in the top cover, or a USB power or two (or both). Or even an access hole with a plastic grommet of some kind. <Actually Belay That - I see a punch point on the safe side of the power supply, the SeemeCNC logo, and another hole that can be used to run 12v from the machine.


Issues and Concerns

Two of the contact-connectors had to be pulled out a bit.
A single FAN wire was missing a contact connector - NP I have a crimp tool.
I figured out why/how you are running one wire to FAN; but it would be nice to spell it out in the instructions. Three less wires is a good thing.

[wepollock]

Assembly Tip

I decided to run the X tower stepper wires, and the extruder stepper wire Under the Duet Board and then snap into place. As a result there are eight less wires that have to run over the board. I will tension them so they don't fall through the vents on the bottom side of the enclosure.

For the Z Tower I used the adjacent channel with zip ties to consolidate the wires and to allow them to drop into the Duet Board, as opposed to running over the board.

In Step 26 I placed the wire holder in an open space.

Wires feed into the board from three directions, and mostly downward. If you lay out the wires before plugging them in you can plan out how you are going to run the wires. Its a real advantage NOT to have a spider web running over the board. The expansive space in the tower (and the how Seemecnc laid the components out) makes wiring, servicing, and upgrading, the printer a dream. Additional "probe" wires also speak to the planning that went into this machine!

My Mistake

I forgot to run the Z wire needed in step 28 that hooks into the solid state relay. No problem because there is so much room in the extrusions I was able to snake the wire through without having to disassemble anything. The channels cut through the tower also was a help

Concerns

I Re-crimped the Spade Connector to the rocker switch in step 33. The original wire was on firm to the crimp connector but but some unstripped wire extended beyond the insulation on the spade connector.

YES the wires that connect to the WAGO are very short. It would be great to have wires long enough so they can be managed better.

How long is this taking in actual work time.

The answer is NOT VERY LONG at all.


Suggestions

It would be great if the pictures in the documentation had the same orientation where it was possible to do so. Or put a compass on the pictures to X-Y-Z.

It would be great to have more instruction steps and sequential pictures. Step 27 for example, replace the arrows with individual steps and pictures after each KK connector is attached.

Another tool I have been using is my label maker. The first thing I did was label the towers.

Rhetorically, how many people are going to have the Artemis as their first build? I think first-builders and non-mechanical or non-electronic people might be mind-boggled. If you built a printer before this is a cake-walk.

Thinking about stuff for later builds

If you need to manage a large amount of wires why not just run a multi-pair wire and then double (or more) up to support the amperage you need. Or alternatively shrink up a wire harness.

[img]https://i.imgur.com/IHvi7ks.jpg[/img]

[wepollock]

Last night i decided to connect the base wiring with includes hooking AC up to the bed, the female-plug-socket, and the bed thermistor. I have the printer upside-down, on the floor, sitting on a large green drafting-table cutting mat; next to the printer is a folding table which I used for parts and assembly.

Experiences

WAGO Lever connectors - I never used these before and they are GREAT! I just purchased a few on EBAY and they expensive @ 40-50 cents plus! The clips are totally useful to me especially the five gang ones. The use of these clips speaks to the build quality on the machine.

I wanted to check the FSR before sealing up the base. I wanted to test the FSR because two of the three FSR KK crimps were redone. Once the wiring was complete I plugged in the unit with the covers open (the base open on the rubber cutting mat and powered on, the top enclosure open to view). The FSR module has an blue light on, I was able to push the bed and see additional blue lights blink. When I look at a picture of the board it looks like it has four leds. (power and one for three FSR's?) Is this correct. I guess the board was oriented (awkwardly) to allow folks to see the leds flickered. The location is fine but it needs an acrylic window so you can see the leds.


Wiring the FSR ST John Board was not difficult with a standard needle nose pliers and a worklight Here is a shameless plug for my 3d printed worklight https://www.thingiverse.com/thing:2321559. I guess the board was oriented (awkwardly) to allow folks to see the leds flickered. The location is fine but it needs an acrylic window so you can see the leds.


Being variable resistors the FSR's would be analog devices, you press them and a resistance level changes. but the board has a single Signal pin, Ground and VIN. There are leds on the board for each of three FSR's. An FSR connected directly to an Arduino would be analog.

My guess is (having worked with capacitive touch on Arduino) that the resistors are a bit fidgety and the board has to be located near them. I am thinking; Does the board send a high or low if any of the FRS pads is activated, or does it send an analog signal? Or is it an endstop sending OPEN/CLOSE like a contact switch; is it bounced in software?. The red is getting power from the 50 pin connector on the Duet. The black and white are connected to the Z-probe pins; ground and Zprobe in. Is this IC2? which each FSR resistor is addressed?

The FSR pads are not located at the towers, they are offset by about 65mm to center, clockwise from Z- Is the FSR checking for a hit at a location or is it just checking for any-hit. Either way you need to see those lights. A clear panel is needed at that location, It looks like there is 5mm clearance for any cable running over the extrusion.

To be continued.

[wepollock]

With 5mm clearance there is not much room for zip ties; I made sure the cables (and ties) were not going to interfere with the FSR. It would be great if there was a small length of extrusion channel cover provided.

Power to the Duet.

I noticed when I powered up near the stepper motors there were three/four red leds lit. The Extruder stepper did not have a red light. I swapped cables and still no red light. (I am assuming the stepper for the extruder is not activated in software.

[wepollock]

The plastic rivets suck. They really DO Suck... Press fit these forget about it.. I have a rubber mallet... I just love banging a device that is a precision device... Come on.

[dc42]

I noticed when I powered up near the stepper motors there were three/four red leds lit. The Extruder stepper did not have a red light. I swapped cables and still no red light. (I am assuming the stepper for the extruder is not activated in software.

Those LEDs indicate the endstop input status, not the motor status.

[wepollock]

I noticed when I powered up near the stepper motors there were three/four red leds lit. The Extruder stepper did not have a red light. I swapped cables and still no red light. (I am assuming the stepper for the extruder is not activated in software.

Those LEDs indicate the endstop input status, not the motor status.

Very Good to know, thanks. VERY COOL... I tapped the contact switch for each tower and you get a led blink..

I should have reviewed the Duet board a bit more, and double checked my cables. I will start that now.

I checked the duet map and the endstops signal for the appropriate column.



Add to Tools

5.5mm Combination wrench
Rubber mallet.




PS plastic screws would be great instead of rivets.

[wepollock]

Also I am wondering where the missing (omitted) KK contact on the top fan is.. Did it creep into the power supply when shipping? It is probably just missing.

[wepollock]

I am going to power off and put on the effector.
I also purchased a TP link access point, not knowing how this device will/should connect to wireless.

[wepollock]

WOW... the effector.. the solid platform, linear rails. modular connections WOW!

[wepollock]

I had to uninstall existing teensy drivers to get duet drivers to load on windows 10.

[wepollock]

I was able to connect to the printer, I created network, uploaded the seemecnc zip file and started testing. I did not clamp down the Black connector to the effector probe fully into the WAGO connector. This was obvious because I did not have power to the extruder, extruder fan, or hotend. It took about 15 minutes to figure the problem out.

Mistake

There is a test script provided which does a conservative bed level. Instead I just clicked the bed level button and the printer homed properly and then did the first bed level. I was very happy, but the dog was WTF.. and the dog is providing a soundtrack

https://player.vimeo.com/video/250033926

I would like to thank the guys at Seemecnc for designing something that is totally outstanding.
1) The quality is off the charts - frame, effector, design and layout
The build was extremely easy with most of the main components already assembled.

I will try to print tomorrow.

As a note build tools (plus label maker not shown).
http://i.imgur.com/UbcGnux.jpg

[dc42]

I had to install teensy drivers to get duet drivers to load on windows 10.

The Windows driver is at https://github.com/dc42/RepRapFirmware/tree/dev/Driver.

[wepollock]

Sorry for the typo.. I had to Uninstall existing Teensy Drivers assigned to com4 to get the Duet Board Driver to load. Initially the Duet board was trying to use com6 (which it did not like). Once uninstalled my existing Teensy drivers the Duet Board Driver kicked in...

[wepollock]

THIS ISSUE AND CONCERN WAS SOLVED IN A "FLOWCHART" IN LATER POST


I ran into a low G0 Z0 situation after G32. Zero was contacting the bed exactly! the effector was exactly at zero, hitting the bed
I followed the configuration instructions (repeated several times) resulting the same concern.
For that reason - I am not quite sure if the instructions in Step 5 calibrating your printer are correct.
In the instructions the same command for "too low" is "too loose" in alternate documentation (link from siblimelayers) For me this was a too low problem.

Test your calibration by putting a sheet of paper on the bed and running: "G0 Z0" in the gcode console. Your nozzle should just touch and pinch the paper so it's hard to move. If it's too loose, run the following GCode: "G31 P500 X0 Y0 Z-0.25", then run "G32" to recalibrate, then "M500" To save all your current calibrations and settings.


I raised the offset with the following command.

http://www.sublimelayers.com/2016/01/fs ... -duet.html

Once you have FSR based auto calibration working you may find that it needs a little tweaking to set Z=0 properly. FSRs require force to trigger. Depending on how rigid your bed and FSR mounting system is, you may get a little flex before the FSR triggers. This can result in the nozzle being too low - the nozzle will drag on the bed on the first layer or the layer height will be too thin. The G31 line in the config.g file can offset this. I have a very rigid bed and mounting system and find I need -0.05 mm compensation using the Z parameter for Z probe height.

G31 X0 Y0 Z-.05 P500 ; Set the zprobe height and threshold (put your own values here)

I have a .15 gauge thickness gauge.

G31 X0 Y0 Z-.35 P500 worked

I am moving the X-Y axis here and there and there is some variation.. So far it is not that significant..

[geneb]

FYI, the FSR board isn't sending an analog signal. It translates the FSR changes to a switch event. In the case of the Artemis, the board is configured as normally closed. This helps prevent false triggers due to induced signal noise.

No window is needed - you can clearly see the FSR LEDs blinking through the ARTEMIS logo cutout.

g.

[wepollock]

Thanks. I am just looked at the chip its an ATTINY chip

http://www.mouser.com/ds/2/268/doc8197-1066118.pdf

So the board is taking three analog inputs, figuring out a trigger point and then converting to open/close.

So if you see a blink through the metal vents you are covered.

So it's basically the FSR board is programed to provide a single open or close. Its acting like an endstop contact switch.

"Each FSR is monitored independently of the others, so the output will trigger when at least one FSR triggers an input."
By default, the end stop output will be normally open (NO), which should work fine with any printer. However if you want to change this to normally closed (NC), connect pin NC/IO3 to GND.

Perhaps, it would still be good to have a window. 1) it would make plugging in the board easier. 2) you would have more granularity as to which FSR resistor is triggering. On the other hand folks might ask why is x FSR not triggering, not realizing it is a single switch-like signal.

[wepollock]

Here is a big tip.. bed offset - PLEASE VERIFY THIS SO I don't mess folks up if I am wrong here

I was having a concern whereby after I did G32 the hot end was pressing exactly into the bed after G0 Z0, with zero room for paper or a gauge. So my offset from the bed was not correct. To correct the problem I did a rinse and repeat flowchart. To fix a problem like this I used the following flowchart.

FLOWCHART

HOME
G31
G32
M500
G0 Z0
Repeat

G31 Towards and Away

This is conceptually a bit difficult to gleam, in terms of offset.

The minus sign (negative increments) is going away from the bed
The plus sign (positive increments) is going towards the bed

So a command G31 P500 X0 Y0 Z-1 goes one millimeter away from the bed
Therefore a command G31 P500 X0 Y0 Z1 goes one millimeter into the bed.
The adjustments in the G31 command are not cumulative.

Progression

I am too close to the bed without an offset - so I need to offset away from the bed in the negative direction.

Iteration One

Home
G31 P500 X0 Y0 Z-.55 ; so i started here and worked my way down to -.40
G32 ; now that we have a trial offset lets autohome
M500; now lock this test case
G0 Z0; now lets see if the offset is good or not

Iteration Two
Home ; Repeat the process, take it from the top
G31 P500 X0 Y0 Z-.50 ; a little bit closer to the bed from Bed-.55
M500
G0 Z0

....

Final Iteration
G31 P500 X0 Y0 Z-.40 ; so i started here and worked my way down to -.40 with seven iterations
G32
M500
G0 Z0

Guess what; every time I did a G32 the calibration got better and better.

On another note;

I monitored the wattage the machine uses.
7.5W turned on in idle
450W when first heating the bed and extruder - 1/3 of a good hairdryer
190W where it levels off after most of the startup heating
115W-142W at temperature and running<this is very standard for the 3d printers I tested - its prob less than your computer gaming rig
30W the extruder only.
The BIG BEEFY power supply is not doing very much work; the mains power is doing the heavy lifting. Before I closed the case I ran some wires out from a punch hole in the cover..

PS the fan is quiet! The machine is quiet.

Peak is at 4 amps+/-
A good desktop might be 150-250W or more!
Remember voltage down Amps up.

Oh and I stopped grousing about the plastic rivets.. I have a 3d printer; I will just print something to hold the top on!

[wepollock]

So far CURA sucks. My first print was absolutely scrambled. Absolutely.
The later version of CURA looks much better than the obsolete version you are suggesting we use. NO NO... NOT GOOD

To see if my problem was Cura, I used an old profile in Simplify 3d with all the wrong settings and wallah! the first layer looks awesome. THIS IS A WOW MOMENT

The machine is quiet, I can run the machine and watch TV or even read. It does not even sound like it is on!

I am getting a tiny little slipping-skipping on the extruder when it is doing infill. All the setting I am using are wrong!

It is also great not having to use a SD card! or USB cable. Compared to the web interface, the only really useful aspect of an onboard control screen is to monitor temperature. If you think about it a $100 chromebook can be your screen.

I would suggest, NO urge you guys to take the time to publish machine settings and profiles for Simplify 3d.

[pouncingiguana]

Just have to put my $0.02 in here. Our test print benchy was sliced with the Cura version we recommend and the Cura profiles we provide. Setup correctly, Cura makes great prints.

I've got new profiles and the procedure to get around the bugs in 3.1 on deck, just need time to finish them up. Look for them later this week.

[wepollock]

pouncingiguana Choice of slicer might be a "religious" type of discussion (one to avoid). Strictly from a business standpoint I could see some issues using a costly third party slicer - and then having my customers have to pay even more; but I can also see a concern using a competitors software (even if it is open, would Coke Use Pepsi's bottle?). I might actually try Kiss Slicer (which is anything but KISS). As to religion. What I am seeing in the Artemis is a Industrial Strength Hardware Platform. It is a tank! I am getting religious about this platform! great job.

The first print shines (again I am using random slicer settings, a really cheap PLA NOVAMaker - is resold and rebranded into slightly more expensive product Orb Polymer, 210 degrees )

[img]https://i.imgur.com/7jm6xQ7.jpg[/img]

First Print A little too close to the bed (there is a .08 shelf on the first layer of print, easily fixed)

[img]https://imgur.com/1YSp9yQ.jpg[/img]

This is what the print is for; I printed a clip to hold bins under my desk;

[img]https://i.imgur.com/GXTv09u.jpg[/img]

Silicon 14ga wires waiting for external accessories and a Grommet. See the Axis labels.

[img]https://i.imgur.com/raecpdA.jpg[/img]

Build Tools - Label Maker Not Shown.

[img]https://i.imgur.com/LhCGhxU.jpg[/img]

[wepollock]

Here is a macro for setting your bed offset. You can change the Z parameter in the G31 line as needed. -.48 happens to be my new setting.
I could not find how to input a parameter from the screen in Gcode. I did not use Pause and Resume I just waited 30 seconds; this way slip a piece of paper between the bed and extruder. Any help is appreciated.


G28
G31 P500 X0 Y0 Z-.48
G32
M500
G0 Z0
M117 "Measure Gap From Extruder to Bed - waiting for 30 seconds"
G4 S30
G28

[wepollock]

Running Cura Now... (I might be a convert) because I have never seen 1st and 2nd layers as nice!). The extruder has the correct settings thanks to the default profiles and it is silent. (spoke to soon.. there is a tiny little bit of grinding.. the filament is 1.71 presently, now it is at 1.78 no sound.)

To business. I figured out the mistake I made in Cura. I did not change the GCODE flavor. If I was going to rationalize it I would blame it on eyesight.

X Width 290
Y Depth 290
Z Height 540

Build Plate Shape - Elliptic
Machine Center is Zero - Check
Heated Bed - Check
Code Flavor - Repetier

Start G code
G28 ;Home
G1 Z15.0 F6000
G92 E0

End G Code
M104 S0
M140 S0
G28
M84

Number of Extruders 1
Material Diameter 1.75
Nozzle Size .5

Xmin 20
Xmax 10
Ymin 10
Ymax 10

The default profile is 5 degrees lower that I usually print
Its an hour long print of a small parts box - so I will see how retractions work.

[geneb]

You don't adjust the filament diameter to fix a grinding problem, you either increase the hot end temperature or slow down the print speed.

To get a good idea of the filament diameter, pull off a meter or so of filament and measure 10 spots with a digital caliper and average the values out, then write that average value on the side of the spool. That's the filament diameter you'll use in the slicer.

g.