Self Sourced Voron Trident 350 - Yet Another India Build

[d4h]

I’ve parked the toolhead assembly for some time due to the struggles with the H2 V2X extruder. Dc3d was kind enough to reprint the entire lot of printed parts for H2 V2X using ASA. They seemed to be holding up better than the ABS ones until the extruder mount spontaneously snapped. May be I’m doing something wrong. Probably tightening too much. This time around I was more careful. I’m not very confident that the H2 V2X extruder assembly is not going to come apart mid print. I’ve called off the toolhead build using H2 V2X.

Page 23 of the StealthBurner Assembly Manual has this joke -

DON’T OVERTIGHTEN
Tighten until the plastic bends and cracks. Back up 2 turns, discard parts, reprint and try again.

I’ve already done that two times!

I’m considering four toolhead options
(1) Fallback to stock SB + ClockWork 2 - I’ve sourced some cheap generic gears to finish the CW2. The thought of crimping and running loads of wires from MCU to the toolhead is not looking very appealing.
(2) SB + ClockWork 2 + EBB - I’ve rigged up an umbilical by wrapping a piano wire and the 4 core wire from BTT CAN Cable in a heat shrink tube.
(3) SB + Orbiter 2 + EBB - I sourced the Orbiter 2.0. The extruder mount looked straight forward in theory. I tried it but it did not fit properly.
(4) Xol + Orbiter 2 + EBB - the Delta fans mentioned in the Xol BOM are still not readily available here. Plus I’d have to carry out mods to the Z spacers and the front idlers.

I’m inclined towards (2) as a now. Not having a working 3D printer at my disposal limits the choices.

 

[d4h]

I completed the A/B Motor and Belt assembly. The shipment from Biqu landed up just before Christmas up after about 2 months from order date. By that time I had already ordered a new set of 2x Nema 17 Stepper Motor 17HE19-2004S A/B motors and installed them.

I’ve installed the CNC Voron Tap Sensor V2.1. I had a potential clearance issue on the left side of the gantry where it looked very close to the extruders for the belt to pass through. I struggled with the issue for some time. De-racking the gantry by loosening the bolts on the D Aluminium extrusion as in Solving Gantry Racking video seems to have solved it. I struggled with the belts as I had cut them too close and there is not much spare length at the ends.

 

[d4h]

I’ve partially completed WIRING section of the assembly manual. I did discover some errors in the wiring diagram that I had created. I’ll update it at the end of the build. I used 1m cable duct chopped up in 6 sections (2x 380mm, 4x 60mm) to conceal the wiring. I could have done with a little longer duct. But I’m happy with the outcome. The motors are disconnected. It’s just enough to do the first power-on.

 

[d4h]

Software installation is partially complete.

• Raspberry Pi 4B is up with Klipper, Mainsail (and fluidd as well)
• BTT Octopus Pro is flashed and is visible in Pi.
• U2C is visible in Pi as a device
• EBB 36 is flashed with Katapult. I need to hook it to the umbilical and flash the Klipper on it.

Login results from Pi after first power-on. Fingers crossed - all boards are working as expected!

 

[d4h]

I’ve parked the toolhead assembly for some time due to the struggles with the H2 V2X extruder. Dc3d was kind enough to reprint the entire lot of printed parts for H2 V2X using ASA. They seemed to be holding up better than the ABS ones until the extruder mount spontaneously snapped. May be I’m doing something wrong. Probably tightening too much. This time around I was more careful. I’m not very confident that the H2 V2X extruder assembly is not going to come apart mid print. I’ve called off the toolhead build using H2 V2X.

Page 23 of the StealthBurner Assembly Manual has this joke -


I’ve already done that two times!

I’m considering four toolhead options
(1) Fallback to stock SB + ClockWork 2 - I’ve sourced some cheap generic gears to finish the CW2. The thought of crimping and running loads of wires from MCU to the toolhead is not looking very appealing.
(2) SB + ClockWork 2 + EBB - I’ve rigged up an umbilical by wrapping a piano wire and the 4 core wire from BTT CAN Cable in a heat shrink tube.
(3) SB + Orbiter 2 + EBB - I sourced the Orbiter 2.0. The extruder mount looked straight forward in theory. I tried it but it did not fit properly.
(4) Xol + Orbiter 2 + EBB - the Delta fans mentioned in the Xol BOM are still not readily available here. Plus I’d have to carry out mods to the Z spacers and the front idlers.

I’m inclined towards (2) as a now. Not having a working 3D printer at my disposal limits the choices.

I finally figured out why I was having issues with the toolhead assembly.

StealthBurner Assembly Manual says.
FAN PREPARATION
Remove the front of the 5015 fan. Clip off and file down the stock mounting ears.
There’s a trimming jig STL included in the release to make this task easier.
PART COOLING FAN



The front view of the 5015 is not apparent from the diagram. What I was doing wrong is after clipping the ears off, I was putting front of the 5015 fan back again before inserting it in the StealthBurner. No wonder all the extruders were not fitting properly! After some measurements and reading through the manual again, I’m happy to have located the issue. This had caused almost a month of heartache.

 

[d4h]

It's been a slow but steady progress so far. I completed the heavily modified "VORON STEALTHBURNER" assembly manual.

As I had started out, StealthBurner assembly was the one I was least comfortable with. So had I had modified the it to have as much pre-assembled components as possible. Also, I liked the idea of having an umbilical cord running the CAN bus which can flex in the 3 dimensional space and not having to deal with crimping and running 7+ pair of wires through a flexible cable chain between the MCU and the toolhead that flexes in only in the 2 dimensional space.

I had read enough about CAN bus and felt comfortable pulling it off as my a part of my first build. The CAN bus uses a dedicated BTT U2C V2.1 between the Raspberry Pi and the EBB. I had initialled toyed with BIQU H2 V2X All Metal Extruder for Voron. But, given the issues I had with its printed parts, I settled for Orbiter 2.0. I picked up the ChaoticLab CNC Voron TAP V2 for the Z probe.

Electronic, Electrical and some Mechanical Hardware:

Component	Electronic Hardware	EBB Termination	Remarks
Extruder Breakout Board	BTT EBB 36 CAN V1.2(G0B1)	-
Hotend	Phaetus Rapido V2 UHF Black PT1000(HF config)	Hotend 0(*), TH0	Hotend 0 on EBB uses a ferrule termination. I changed the termination for the thermination to JST XH2.54 2 pin female connector. The thermistor wires have a braided teflon sleeve. I wasn't able to crimp the JST terminals on them. The problem being that the braided sleeves don't have a clean cut. There is no clean separation between the naked wires and the sleeve. After 6-7 tries, the wires were too short to terminate on the EBB. I crimped a pair of regular teflon wires and connected them to the original wires using heat-shrink butt connectors.
Part Cooling Fan	GDSTime GDB5015 24V	FAN2	The fan came with the JST XH2.54 2 pin female termination.
Hotend Fan	GDSTime GDB4010 24V	FAN1	The fan came with the JST XH2.54 2 pin female termination.
Z Probe/ Bed Leveling Sensor	ChaoticLab CNC Voron TAP V2	Endstop	The termination on the EBB is changed to JST PH2.0 5 pin female connector. Only 3 pins are required.
Extruder	Orbiter 2.0 (uses LDO-36STH20-1004AHG(XH) motor)	E-Motor	The printed parts for the mod is from elcrni. I had an issue pushing the PTFE coupler into the printed part. I've just made the hole bigger as no amount of physical force I applied was enough to push the coupler into the original hole of the printed part. The termination wires for extruder motor is JST XH2.54 4 pin female connector. Termination pins are swapped to match the EBB pinouts.
Umbilical cord	BIGTREETECH EBB SB2209/SB2240 CAN V1.0 for Voron StealthBurner, Umbilical Support Spring Steel Wire, Heat Shrink Tubes	CAN Bus + Power	BTT CAN bus cable and a steel wire is wrapped a heat shrink tube to create the stiff umbilical cord. The connectors at the cable has to be changed to match my assembly. Two sides of the 4 pins of the CAN bus are terminated on a Molex Micro-fit 3.0 2x2 male connector of the EBB and U2C boards respectively.

This is a snap I had taken to show the heat shrink butt connector I have used for patching the regular teflon wires to the braided teflon wires of the thermistor coming from the hotend. I likes the fact there there is no physical crimping required. Heat shrinking is enough to secure the wires in place. From my reading, the thermal characteristics of the butt crimp should hold up the heated enclosure temperatures. There is a solder blob inside the butt crimp which also also hold up as the EBB also has lots of the solder on it. Theoritically, it looks good enough.

 

[d4h]

I've rigged up an emergency stop switch using components bought from Amazon.

• Emergency Stop Switch Push Button: The top is press fitted to rest of the body. I pried open the top using a thin screw driver head.
• Extension Cord
• PG7 Cable Glands
• 18 AWG 2 core electrical cable - 1 meter: Bought from a neighbourhood electrical store. This is used to connect the emergency stop switch to the extension cord.

Edit: Added pic


The main driving force behind this is the warning note I read in Voron documention sensorless homing section. I'm reproducing it here.

Note

IF YOU DO NOT READ THIS CAREFULLY AND ADJUST THINGS WHEN NEEDED YOU WILL BE SAD. IF YOU USE A PHYSICAL ENDSTOP YOU MUST CHANGE MACROS TO HOME Z TO THE ENDSTOP PIN, NOT THE MIDDLE OF THE BED. YOU HAVE BEEN WARNED!

 

[d4h]

I've started the "INITIAL STARTUP CHECKS" according to the Voron documentation. I'm using the Mainsail UI for the verification steps.

#	Verification Step	Status	Remarks
1	Verify printer.cfg		The only real error I had after correcting the syntax parsing errors with the printer configuration was the following. Klipper reports: SHUTDOWN MCU 'mcu' shutdown: ADC out of range Sensor 'heater_bed' temperature -88.522 not in range 0.000:120.000 This generally occurs when a heater temperature exceeds its configured min_temp or max_temp. Once the underlying issue is corrected, use the "FIRMWARE_RESTART" command to reset the firmware, reload the config, and restart the host software. Printer is shutdown My heat bed sensor was incorrectly terminated on the Octopus Pro MCU. After correcting the wiring, Klipper loaded the configuration files correctly.
2	Verify Temperature		Verifed that the temperature of the nozzle and bed are present and not increasing.
3	Verify Heaters		Verified that the hotend heater and bed heater are heating to 50 Celcius.
4	Verify cooling fans		Not done as the cooling fans are not installed. The skirting is planned to be printed after the printer is up and running .
5	Stepper Motor Check		All 6 motors are now verified. I had a few issues with the A and B motors which took me a good part of the day to sort out. 1. Running STEPPER_BUZZ STEPPER=stepper_x gave me the following error. Klipper reports: SHUTDOWN TMC 'stepper_x' reports error: DRV_STATUS: 801900e0 s2vsb=1(ShortToSupply_B!) ola=1(OpenLoad_A!) olb=1(OpenLoad_B!) cs_actual=25 stst=1 Once the underlying issue is corrected, use the "FIRMWARE_RESTART" command to reset the firmware, reload the config, and restart the host software. Printer is shutdown This was because I had done the termination on the MCU for LDO-42STH48-2504AC motor. I had forgotten that I had changed the motor to NEMA 17HE19-2004S due to a delayed shipment. 2. Running STEPPER_BUZZ STEPPER=stepper_y gave me the following error Unable to read tmc uart 'stepper_y' register IFCNT Once the underlying issue is corrected, use the "FIRMWARE_RESTART" command to reset the firmware, reload the config, and restart the host software. Printer is shutdown TMC driver was misaligned on the MCU. The pins were right shifted. I managed to spot it visually.
6	Sensorless Endstops, XY Homing, Motor Configuration		I took a breather and double checkied on the sensorless homing configuration before running it. I had my hand on the E-stop switch all the time! Luckily I didn't need it. Home X and Y are working without endstops using sensorless homing. I updated the [printer.cfg] and included the associated configuration homing.cfg as per the Voron Documentation Configuring sensorless homing guide. I tuned the driver_SGTHRSdecrementing the value from a maximum of 255 in steps of 50. I hit 0 and still no success. May be I had the X and Y switched. Still it was the same result. I finally turned the printer upside down to check the DIAG jumpers. I had forgottem to jumper them! After that I was able to tune the sensorless trigger thresholds for both X and Y with some hits and tries. I was getting the following error even with the printer on standby. TMC 'stepper_y' reports error: DRV_STATUS: 80190103 otpw=1(OvertempWarning!) ot=1(OvertempError!) t120=1 cs_actual=25 stst=1 Once the underlying issue is corrected, use the "FIRMWARE_RESTART" command to reset the firmware, reload the config, and restart the host software. Printer is shutdown I've reduced the run_current from 1.4A to 1.1A. I'm not seeing the issue after that so far.

Edit #1: Updated for check #6 and reformatted the table.

 

[sanketss84]

fantastic documentation. keep sharing your build progression.
Its great to see such a detailed log which would definitely assist future Indian Voron builds.

 

[d4h]

fantastic documentation. keep sharing your build progression.
Its great to see such a detailed log which would definitely assist future Indian Voron builds.

Thanks your your encouragement @sanketss84. Your posts have gotten me started me started on this journey. I've been tied up with work. Hence, not much progress on the build. Also, I wanted to re-crimp the CAN bus umbilical. I think the Molex Micro-fit 3.0 female crimps that I have are not meant for thicker gauge wires (probably AWG20). I've ordered some that go up to AWG20.

 

[sanketss84]

Thanks your your encouragement @sanketss84. Your posts have gotten me started me started on this journey. I've been tied up with work. Hence, not much progress on the build. Also, I wanted to re-crimp the CAN bus umbilical. I think the Molex Micro-fit 3.0 female crimps that I have are not meant for thicker gauge wires (probably AWG20). I've ordered some that go up to AWG20.

take your time no need to rush the build. I also split my build over 6 months as I was reading and analyzing various solutions and mods to add to my build straight up.

 

[d4h]

I felt that the umbilical that I made by running a piano wire along side BTT CAN cable was a bit too stiff. I made an alternative umbilical by pushing 4 wires and a filament through a PTFE tube.

• 2x 20AWG PTFE wires for 24V power
• 2x 24AWG(twisted) PTFE wires for CANbus signals
• 1x 1.75mm filament
• 4mm inner diameter x 6mm outer diameter PTFE tube - 1 meter

Another motivation for switching to PTFE tube based umbilical is that there is already a PTFE tube running to the toolhead for the filament which has stood the test of time. I'm quite happy with the feel of the new umbilical. I've used the Voron 2.4 Exhaust cover with Umbilical (PG7) mod to secure the umbilical. The umbilical secured at two ends using two PG7 Cable Glands (without reliefs). I had bought a 10 pack gland earlier which came in handy.

I had to recrimp the Molex termination on the EBB as one of the pins was loose and CAN bus was losing connection to the EBB. Stating that my skills with micro-crimping had not been very good would be an understatement. My success with crimping Molex Micro-fit 3.0 female pin had been 1 out 4 at the very best. Some of what I called "successfully crimped pins" came out quite mangled. Then I thought that there may be other factors than my crimping skills were probably to blame.

1. IWISS IWS-2820M micro-crimping tool: I bought the top rated and rather expensive Engineer - Micro Connector Crimpers - PA-09.
2. Molex Micro-fit 3.0 female pin: I thought that may be pins I have are meant for thinner wires only. I bought a new set of pins which go up to AWG20.

I'm very happy to report that my success rate with the Molex pin crimping has gone up to 100% by just using the Engineer PA-09. I cannot stop raving about it. The pins that I had been using were just fine. The crimped pins come out almost photogenic. I would whole heartedly recommend the PA-09 to anyone facing micro-crimping isssues. I had earlier posted about the issues I had while crimping the hotend thermistor teflon wires. I suspect that it was also due to the IWISS crimping tool that I was using.

 

[d4h]

I ran a few more startup checks. X, Y and Z kinematics are checked out now.

#	Verification Step	Status	Remarks
7	Bed Locating		Not required as the build uses Voron TAP V2. There is no Z endstop.
8	Define 0,0 Point		Homed X and Y axes using G28 X Y. I used the Mainsail control to move the nozzle to the front left corner of the bed but still on the bed. I could have claimed a few milimeters of bed. But I let it be for now. The Main sail GUI and M114 command reports the current location as X=0.000, Y=0.000. I did not update the position_max for [stepper_x] and [stepper_y].
9	Z Endstop Pin Location		Not required as the build uses Voron TAP V2. There is no Z endstop.
10	Voron TAP V2, Z Home (was Probe Check)		1. Updated the printer.cfg [stepper_z], [probe], and [homing_override] sections according to the instructions. [safe_z_home] is commented out as sensorless XY homing is used. 2. Checked that the states of "Endstop Z" and "Probe" on the Mainsail Web GUI are reflecting correctly by lifting the toolhead and then by releasing the toolhead. 3. Moved the probe to a safe location (X=175, Y=175) using G90 followed by G1 X175 Y175 and homed the Z-azis using G28 Z. The bed moved down instead of up. I pushed the E-Stop and inverted the signal for dir_pin on all three Z motors. Z homing ran correctly after that. 4. Ran PROBE_ACCURACY samples=100 two time to break-in the probe. - Standard deviation (σ): 0.001412 (expected less than 0.003mm) - Sensed distance is not trending (gradually decreasing or increasing) with a median of 0.0275mm.

 

[d4h]

I was able to complete the rest of Initial Startup Checks.

#	Verification Step	Status	Remarks
11	PID Tune Bed & Hotend		PID tuning of hot bed: 1. Moved nozzle to the center of the bed and 10mm above the bed surface using using G90 followed by G1 X175 Y175 Z10. 2. Performed the PID calibration routine using PID_CALIBRATE HEATER=heater_bed TARGET=100. 3. Waited for ~7-8 min. I observed that PID tries to maintain the temperature between 95°C and 100°C and then saved the configuration using SAVE_CONFIG. PID tuning of hotend: 1. Set the part cooling fans to 25% using M106 S64 2. Ran PID_CALIBRATE HEATER=extruder TARGET=245. 3. Waited for ~3-4min. I observed that PID tries to maintain the temperature between 240°C and 245°C and then saved the configuration using SAVE_CONFIG.
12	Bed Leveling		Bed levelling with cold printer: 1. Bed levelling using Z_TILT_ADJUST macro. I got Retries: 2/5 Probed points range: 0.002500 tolerance: 0.007500. That was amazing! I didn't expect to adjust it in one try itself. I was prepared to run it atleast 20 time. 2. Re-home Z using G28 Z. 3. Run SAVE_CONFIG to update printer configuration file. 4. Reun RESTART. Note: Performed this before step 11 when the printer was still cold. Bed levelling with heated bed and chamber: 1. Re-home the printer after a restart using G28. 2. Set the extruder to 245°C and heated bed to 100°C and let the cold printer heat up for at least 15 minutes. 3. Ran PROBE_ACCURACY with the nozzle at the center of the bed till reading are stable. It gave a message Extruder temperature target of 245.0C is too high, lowering to 150.0C. It looks like the instructions didnot consider Voron TAP setup. I ran PROBE_ACCURACY samples=100multiple times at 5 min interval to see when the readings stabilize. It took almost an hour. I don't know what to make of it. The printer is not enclosed and the components are being for probably the first time for elongated duration. 4. Ran bed levelling Z_TILT_ADJUST macro again. It probed 3 points on the bed and responded that there is adjustment needed. There is no need to run Z_ENDSTOP_CALIBRATE. Observation: I don't quite understand the point of this as we're not saving the printer config. I guess the Z tilt base lines are saved when printer is cold. Maybe, this is just to make sure that the Z tilt adjustment can still be be performed when printer is hot.
13	Z Offset Adjustment		My understanding is that manual adjustment using 'paper test' is not required as the build uses Voron TAP V2. There is no Z endstop.
14	Extruder Calibration (e-steps)		1. With the cold hotend, ensured that the extruder is running the correct direction 2. Heated the hotend to 220C, for eSUN PLA+, and extruded 10mm filament. I used G91 followed by G1 E10 F300. It was retracting instead of extruding. I removed the signal inversion in the [extruder] dir_pin of the printer settings. After saving the configuration and restarting the firmware, the extrusion worked fine. 2. With the hotend at temperature, measured and make a mark on the filament between the roll of filament and your extruder, at 150mm away from the entrance to the extruder. 3. For extruding 50mm 2 times, (for a total of 100mm) at extrusion speed to 1mm/s, I used G91followed by G1 E50 F6. Nothing seemed to be happening. My mind was racing to figure out what I had missed. Then I saw the extruder slowly extruding a string of filament at 1mm/s. 4. Measured from the entrance of the extruder to the mark made previously. The reading was exactly 50mm. Really! Is that even possible? I think I have the Orbiter 2.0 to thank for it.

Legend:
: Done
: Not required
: Not done

With that a major intermediate milestone is completed!

 

[d4h]

Moving on to next step of SLICER SETUP. Of the two popular slicers: Cura and SuperSlicer. Cura is recommended as more beginer friendly. Hence, Cura is what I'll be installing. This is going to be a new topic for rampup for me. I've downloaded it and will need to figure out the profile parameters.

 

[sanketss84]

Moving on to next step of SLICER SETUP. Of the two popular slicers: Cura and SuperSlicer. Cura is recommended as more beginer friendly. Hence, Cura is what I'll be installing. This is going to be a new topic for rampup for me. I've downloaded it and will need to figure out the profile parameters.

try orca slicer its what I use with my voron and what I would recommend.
you can use cura to understand various settings and its options. it has some of the best docs.

 

[d4h]

try orca slicer its what I use with my voron and what I would recommend.
you can use cura to understand various settings and its options. it has some of the best docs.

Thanks for the pointer @sanketss84. I'll checkout Orca.

• Cura: I was able to set it up, connect to the printer and transfer the sliced Voron cube to the printer.
• SuperSlicer: Doesn't install on my Macbook. I don't want to hack around to get to work when other options are available.
• Orca: Installed. Still need to take it through its paces. From their feature comparison, AI error detection seems to be standout feature. I had been looking for a camera. But, I didn't want to get side tracked before I got the printer into a functional state.

 

[d4h]

I had a look at Cura and the Slic3r family of slicers. Based on the development support and feature support, Cura and Orca are best bets.



From the initial look, Cura is is definitely looks more beginner friendly. Orca seems to have more features for printer and filament tuning. User interface of Orca is not very intuitive and takes some getting used to.

I'm stuck with a clogged hotend/nozzle. Trying to sort it out using the recommended methods.

1. Cold pull: didn't work as the filament doesn't extrude beyond a point
2. Nozzle change: didn't work. Looks like filament is stuck somewhere inside.
3. Needles: Have ordered some.

Might have to take the toolhead and hotend apart as the last resort.

I'm planning to start with Cura and then move to Orca for tuning.

 

[d4h]

I opened up the toolhead as I anyway had to perform another modification to the StealthBurner.

The filament feeder PTFE tube was not very secure with the existing arrangement. I put a M6 heat set inset on top and screwed a PC4-M6 Pneumatic Connector to secure the 4mm PTFE tube (pics here).

Filament clog was both in the nozzle and inside the heating block. I used a soldering iron to loosen the filament and pushed it out with a thin Allen key. I have used an alternate nozzle as the cleaning needles hadn’t arrived.

It took me a good part of a day to get the printer back to the condition before I had dismantled the toolhead. I managed to nick the PEI coating on the steel sheet in the process. The Voron TAP v2 sensor wiring had come loose and didn’t trigger during the Z homing.

The filament extrusion is working fine now.

 

[d4h]

On to the FIRST PRINT.

I had already done the printer settings for Voron 350, material setting eSUN PLA+ white filament, sliced the Voron cube from Cura and upload the file to the printer.

I fired the print. Klipper couldn’t find the START_PRINT macro! I had missed putting it in the printer configuration. The bed was cold but luckily the hot end was still at 220°C from my previous extrusion check. Some part of the first layer stuck to the PEI sheet holding up the print. The nozzle is occasionally scraping the print, removing previously printed material. I let the print run.

It took an hour and half to print the cube. Actually, about 7 months from the start of the idea of building a 3d printer. Not perfect but it’s a cube none-the-less.



You can see the scraped off material on the plate and the on the cube.

I am hoping the the PRINT_START macro should be able to fix most of the issues.

Yesterday was Maha Shivratri (Shiva's night) in India. I'm not that religious. I cannot but help think of the transformation power Shiva wields.