trident Chris' 250 Trident Journey

[claudermilk]

I'll add my self sourced 250 Trident build to the diaries.

This is my first Voron, and second printer. I started out with a Prusa Mini+ kit and enjoyed the build so much that I started looking more seriously at the Vorons. I did look while doing initial research, but the full DIY nature was intimidating at the time; the Mini has been a great learning printer both for building and printing. In the process of researching I found the Voron Discord, this forum, and Nero & Steve's Youtube channels. Lurking, reading, and ultimately starting to ask dumb questions has taught me a lot in a short time.

So I started with the Voron configurator to get my BOM, then grabbed the matching info off the sourcing guide to start finding where to buy parts. I spent quite a while learning about many of the components to determine what the best cost vs quality balance is for me. In the process, I found the popular vendors supplying the hobby. I built myself a spreadsheet to track & manage the list of parts and my often-shifting plans of what to buy and from whom. I started purchases taking advantage of sales on Aliexpress, and just accelerated the process after witnessing the Voron Hug and general supply chain issues. If something is in stock these days, you probably want to buy now rather than wait.

So the list of vendors I've used are:

• 3D Maker Parts
• Amazon
• Aliexpress
• DFH
• Grainger
• KB-3D
• Mandala Rose Works
• PiHut
• Push Plastics
• Sparta 3D
• West 3D

I tried to minimise the orders and take advantage of free shipping triggers as much as possible, so I didn't get hit too hard with that forgotten cost. A few redirected sources due to out of stock situations; PiHut most notable--had to order the board from the UK after the things became unobtanium.

The last part to get is ironically the frame. I'm being picky and want the LDO blue kit which has not been available. As I was composing this I checked again, and unexpectedly got a hit on KB-3D, who now has them! So all parts are now ordered.

I've spent the last week getting my Mini prepped for printing up the parts for the Trident. So I now have a full set of spare parts in PETG, which itself was a learning experience. I have to move the OctoPi from a separate case to the integrated electronics box lid. That is to help fit everything under the original box for the Mini as my cheap enclosure for ABS printing.

Next step is to get some ABS test prints so I can dial that in, then start printing parts.

EDIT: I should probably also list the mods I am incorporating into the initial build:

• hartk SexBolt Z endstop.
• hartk Z endstop PCB
• randell XY endstop PCB
• jeoje 4.3" Waveshare touchscreen with Steve's modified skirt.
• hernsl magnetic bottom panel clips
• Steve's no tape skirt fan mounts
• chrisgonzales 270° door hinges
• Nero's idler bearing stack in place of toothed idler change

Edited February 9, 2022 by claudermilk

[Demosth]

Exciting!!!

3 minutes ago, claudermilk said:

Lurking, reading, and ultimately starting to ask dumb questions has taught me a lot in a short time.

What were some of the questions you felt like were "Dumb questions" - maybe something that once "ya know ya know" - but maybe just was an assumption that everyone should know?

Additionally thank you for listing those vendors, a great way to share the wealth with others!

[claudermilk]

I don't recall specifically. Just newbie questions. Nero and Steve have been great and patient with the questions. Things like why PTFE wire? What stepper is better? Basic stuff I now see they get nearly every stream.

So I pushed some ABS on the Mini and it turned out really well. I hadn't heard of Push plastics before a desperate search for the color I want for the primary color and they showed up this time around. They are a US based company and delivered quickly. The test patch and Voron cube both printed really nice and it doesn't smell--in fact I get more smell from Jessie PLA. I have some fine-tuning to try to do before getting into printing Trident parts, but I hope to get that started today or tomorrow.

[claudermilk]

So it has begun. I spent the afternoon dialing in the Prusa Mini+ for printing ABS. I'm using a Fysetc smooth PEI flex plate which is working well; the ABS sticks well but not too much, my only complaint is after getting heated a few cycles it's developed a bunch of little bubbles that don't affect the part finish. Not as nice as a Prusa bed, but it works well enough.

I chased Z height briefly and now have that dialed in to click-print-and-forget and get great first layers. I then chased top layer finish and a bit of corner swelling. Some pressure advance tuning and outer perimeter speed adjustment cleaned that up as much as I can right now. After getting to a workable point, I threw the Voron torture tests at it and they passed with flying colors. The bearing press-fits nicely and the threads work smoothly.

Since I saw a post from RCF from a while back that if those all work you're good to go, I'm running the first plate of parts as I type this. The Push Plastic ABS is running really nicely. Prints well on the stock Mini with cardboard box enclosure, and doesn't really smell.

 

[Demosth]

Nice looking prints! Glad to see you were able to dial it in so nice. 

How long did it take you to get the Prusa Mini+ dialed in?

[claudermilk]

Initially I spent a few hours over a weekend after building it. I tried to be careful and build it square, but didn't understand yet how sensitive these printers are to how tight you screw things together--I've been used to car projects which sometimes needs a several foot breaker bar to deal with torque required. So I had to go back and redo some of that. Once I got it squared up and properly tightened it's been running like a champ. For dialing in the ABS, it took a few hours of making test prints and tweaking settings. I focused on the live Z adjustment, external perimeter speed, and extrusion multiplier (pressure advance?).

It's not perfect, but plenty good enough to get going. I did have one full plate of electronics bay parts where some of the first layer is not PIF part quality, but they will work fine and will be hidden. Tweaked live Z down another step for some more squish on the next plate.

I've read so many trials people have gone through with printers, and either I've done better than I thought taking in those lessons or I've been very fortunate so far. Except a couple of nasty head crashes right off the bat and the squaring lesson, I've only had one partial nozzle clog in about 3 months of printing.

[claudermilk]

So this showed up yesterday and actual assembly can begin.

I've been busily printing trying to make sure I have parts ready, and the Mini just wasn't fast enough.  I have gotten enough printed up so I can keep going for a while, and maybe not outrun the printer output. I have also learned the Mini just can't successfully print larger ABS parts. I can get a full-ish plate of smaller parts reliably, but larger items are warping on me. So the huge feet (a 14 hour print for all four) will need to be redone. I also saw the same issue with the power inlet skirt and keystone panel skirt; I'll probably reprint those as well just because I'll know it's warped back there. For the rest of the skirt, I'm now waiting to print on the Trident.

I've currently switched over to printing up accent color parts, and that's about a total of 19 hours of printing. The Mini should be able to keep ahead of my building for that.

Grainger also got the last screws in just in time, so now the only part I'm missing is the hotend. Since it's a Revo Micro pre-order what was expected, now we will see if E3D ships before I finish building.

With the frame kit in hand, it's time to actually start bolting parts together. A couple of hours on the kitchen granite counter and I have a nice square frame (each side diagonals all measure within 0.6mm). I scared myself forgetting some extrusion don't need both ends tapped until I remembered where the pieces went; in the end, there were only a couple of tapped holes that weren't perfect. One I was able to carefully chase out with a screw, the other I simply swapped from a M5x16 to a M5x10. I do have two tap and dies sets...both SAE.

I was thinking about having it etched, but a quote request has gotten no response in a week, so I'm not sure if they are still doing it or not and don't feel confident in the  idea of sending my frame off now anyway.

One thing that has struck me is how small all the parts are, and how compact this printer is. The frame isn't that much bigger than the Mini, and with the paver and box over it, they are about the same overall volume. The Trident certainly seems to be using its physical volume more efficiently than the bedslinger Mini.

[claudermilk]

Moving right along, I got the A/B Drive and Idler section built despite multiple distractions and interruptions from life. Can't they see I'm building a printer? 

It went pretty smoothly and I now fully understand what Steve was doing during his build with the level and clamps. I used the same trick--the point being to ensure the rear Z motor and rail support (H extrusion) is in the exact same plane as the back uprights of the frame. 

 

So, with this set up, it's where it should be and shows the offset of the rear brace (Extrusion D). 

It's beginning to resemble a printer already.

I'm not completely happy with the level of the two Y-axis extrusions so I'll revisit their positioning before moving on.

[claudermilk]

It felt like a lot of progress was made. First, I double checked the positioning of the Y extrusions. I got out the level and grabbed a level app for the phone. First I tared out the counter then compared the main frame extrusions and then the Y axis extrusion. Everything came up dead level, so I left things alone and moved on.

First, I did some troubleshooting on printing large items and found a solution. I have found that cleaning the flex plate, then laying down some glue stick and adding a 8-10mm brim to the parts prevents the warping. So I reprinted the feet, and will move on to the skirts next.

Getting the linear rails cleaned and lubricated wasn't nearly as hard or scary as I had thought. Being forewarned, I made sure to have shop towels laid down which is a good thing since a couple of ball bearings did make a break for it. I picked out my best two MGN9 rails for the Y and left the rest for Z. Fortunately my MGN12 runs really smooth. I was wondering what Nero was on about with that until I actually got my hands on the 5 rails; sure enough some were smoother than others.

While printing up a few Stealthburner pieces, I was happy to see this Easter egg for the Revo Micro toolhead mount.

[claudermilk]

Lots of progress. I now have mechanical basically done; all that's left is attaching the bed, but per Nero's recommendation I'm leaving that as long as I can. I also have the hotend and fans to do, I'm also leaving that as long as possible since I'm putting a Stealthburner on it and those parts are the ones still changing most often.

I did make my biggest assembly mistake, putting the front bed mounts on upside down. It was much easier attaching to the lead screw nuts the right way around. 

  

I got the belting right!

I ordered an "extra" meter of belt just in case (so 3m), and this is the leftover from trimming the two pieces to length:   Either I misread the configurator BOM, or it was incorrect when I ran it for sourcing.

Electronics mostly installed. I do have a Pi, but it's a 4B 4GB that I'm swapping with my sister for her 4B 2GB; each one will be more suited to its new use.

I was happy to see this little Easter egg in the Revo Micro toolhead part for Stealthburner.

Next up is wiring.

 

[claudermilk]

Got a little more done over the weekend. Other projects paused any major progress on the printer, and i want to take the next part very slow and careful; I'm now into wiring and am not really confident on that part, so I am reviewing docs, and Steve & Nero's applicable streams several times.

The big next step done is wiring up the  power inlet, switch, and power supply. That went almost without a hitch. Like Nero, I'm not an electrical guy, so I made the same darn mistake and had my multimeter set up wrong. So, yes, sparky sparks. No harm done though thankfully. I am getting 24.3V out, so it's good for the next step.

I do have some wire management to do later to clean that up.

I put off installing the bed as long as possible, but is was time, so I got that in. I also pre-positioned the cable chains to get the lengths right, and make sure the ends were set up correctly. That took a bit more effort than I expected, and I ultimately had to use a couple of dabs of superglue to make the toolhead end a rigid mount. NBD, it's all working perfectly.

I stole Steve's idea for WAGO blocks to attach the bed wires. I'm also using them for the thermistor wires too. I'll just use a ring terminal for the ground.

Next steps are running the bed wiring down to the electronics bay, and hooking up the Octopus to power.

[Simon2.4]

9 minutes ago, claudermilk said:

had my multimeter set up wrong

Trying to check the amperage of your local grid ? -- pouff --

[claudermilk]

LOL, I had the probes plugged in wrong. Just like Nero. So I got to join the sparky sparks club. I did make certain to probe the 24V outputs and not the 110V inputs though. My mistake didn't blow a fuse like it did on Nero.

So further progress with an assist from the Discord crowd (in particular BerzerkaDurk and Logan BC). I did have a couple of things wrong and they set me straight. I now have happy little LEDs showing on the Octopus, and the magic smoke stayed inside. So the scariest hurdle for me is now past.

Needs some cable management and I might rebuild some of them to route cleaner, but it works.

[claudermilk]

Not much visible to show. I have the motor wiring all trimmed and plugged in. I had forgotten the Z endstop wiring, so got that added; the trick Steve used with PTFE tubing is genius, it makes feeding new wires into the drag chain pretty easy. That is now connected up.

I finalized getting the bed wired up. I found that the pre-drilled and tapped holes in my Mandala bed are M5, and the shortest screw I had is the 10mm button head, which is too long, so I had to cut one down. Done, and ground now attached. 

The big step was getting the Raspberry Pi hooked up and Klipper/Mainsail running. As can be seen from earlier photos, I'm not using the RS-25 power supply. I am powering and linking the Pi via GPIO and the Octopus J26 connector. So now the printer is basically alive, though still missing parts.

 

Now that I'm in to wiring and firmware/software, there's not a lot visible changing.

Over the last couple of days I got the motors wired up. XYZ motors trimmed and plugged in, the extruder motor wires trimmed to fit within the wire cover. I finished getting the bed wired up. I found that the tapped holes in my Mandala bed are M5, and the shortest screw I have is 10mm. So I had to cut that down so it didn't stick out the other side. I also found that I forgot to run the Z endstop wires so I had to get those through the drag chain. I used the PTFE tube trick that Steve demonstrated on one of his streams and it works brilliantly.

I got the Raspberry Pi hooked to the Octopus using GPIO for signal and power. Just a simple 4-wire jumper and it's done. No RS-25 power supply needed--which is good because space is tight on a 250 build. With that set up I went ahead and got MainsailOS installed and flashed the firmware to the Octopus. While the printer won't start because it's missing the hotend heater and thermistor, I can verify the Pi is talking to the controller and start to configure printer.cfg.

I also tackled most of the Stealthburner assembly. Because I'm me, I started with soldering the NeoPixel LED chain for the front panel. I haven't soldered in decades, and last time I tried was RC car battery cables--which were ugly but worked. So why not jump back in with AWG30 wires on tiny, closely-spaced pads? After watching a couple of how-to videos on Youtube University (tm) I realized my mistakes in my youth. While the joints aren't the prettiest, I am somewhat confident it will work.

After that, the rest of assembly was straightforward. It really is amazing how much thought has gone (and is still going) into these parts.

  

It will have to come back apart again when the Revo arrives, but with this much together, I can now start to run wires from the toolhead.

[claudermilk]

Lots more progress over the weekend, I've not got it as far as I can until the Revo arrives tomorrow. 

So, first I summoned my courage and pulled out the wire spools (well, after printing up some spools and winding the loose coils from Aliexpress). Then I built the gantry wire loom and four hours vanished in a zen haze. What a scary looking mass of wires. Fortunately, I was able to borrow a nice label maker and create wire labels as I went.

 

 

My wiring might not be a over-the-top OCD neat as some build, but it isn't too bad considering the limited space available. I stole the idea of using a spare drag chain segment for part of the wire management where the raceway wouldn't fit.

Next, double-, triple-, and quadruple-check that I have everything connected correctly. Then the scary power up. Again, no sparks of escaping smoke. I even got the expected magenta toolhead LEDs!

So, looking for more to do I mounted the skirts and screen since wiring is effectively done. I am using the Waveshare 4.3" touch screen connected to the Pi on a DSI cable and running KlipperScreen. Easy peasy install.

 

Yes, I know it's in an error state. On the web interface it's mcu shutdown adc out of range. That is expected since there's no hotend thermistor. So it's actually a good thing (tells me if there's a wire failure in a thermistor the printer will shut itself off for safety).

I also am able to do a basic test of the inductive probe and I believe it's working. Left is hanging in the air, right with a big chunk of metal next to it.

 

The little (barely visible) indicator light lights up and turns off as expected. I'm just wired straight into the Octopus PB7 port with printer.cfg set as "~!PB7" as in the starter file from the repository. Seems like it's working,s o good enough until I switch over to Klicky Probe.

Finally, to make Steve happy: the bottom panel.  I'm using the magnetic mount mod.

[Demosth]

That Stealthburner is just so nice - the build is looking fantastic!

[claudermilk]

It's alive!

..because this arrived:

 

Just fiddling with it, nozzle swaps are going to be as easy as E3D claims.

It just spins off and spins on. One single piece nozzle and heatbreak. On the 0.8mm nozzle I can lookup the bore and it's perfectly smooth for the entire length. The single piece nozzle/heatbreak extends all the way through to the top of the heat sink fins where I believe the PTFE tube meets it. The pre-terminated wiring is the perfect length for the Stealthburner install; literally all I had to do was change my harness plugs over to the right end (I guessed wrong and had the same end as E3D mounted).

With the Revo installed, I am now in preflight troubleshooting. There's a few issues that have to be dealt with before moving on.

First thing I ran into was Z2 didn't buzz correctly--it just vibrated a bit and stayed put. That turned out to be one wire hadn't seated correctly in the JST housing; I got that fixed and good contact on all 4 wires, now it buzzes fine. The base Trident config file from the Github repository has all the motors set with the "!" inverse flag on the direction pin and on my machine I had to remove them. No big deal. Now they all appear to be going the right way (buzz X & Y and it goes back & right first, Z goes down first).

Now there are three issues to address, all of which I kind of anticipated.

#1 is Y endstop is stuck at triggered. I have the XY Endstop PCB from randell (Linneo pre-built bought from KB-3D). The wiring confuses me a bit. X works, I have the X and GND wires connected to DIAG0/STOP_0 in pins PG6 & GND. I have Y signal connected to DIAG1/STOP_1 in pin PG9. I did connect and run the wire from Vin on the PCB, but it's just coiled out of the way unconnected for now. printer.cfg is set to PG9. I'm not sure why it's not working.

Edit: I triple-checked the PCB end wiring vs what I was connecting up at the Octopus end. I still grabbed the Vin wire instead of Y wire for the Y signal. Simplest solution was to re-pin the XY endstop PCB end connector. Y now triggers correctly. The lesson here, kids, is check your wiring again just to be sure.

#2 is very much anticipated. The probe also is stuck at triggered. It is in J40, pin PB7 with no diode or resistor. I hoped I would be lucky like some others. Looks like I might need to break out the signal to DIAG7/J34 on PG15 with the diode. I'll have to cobble some kind of little patch cable as this is just a temporary setup until I install Klicky Probe.

#3 is possibly the more difficult one--at least for me. When running stepper_buzz on the extruder it shuts Klipper down. I am guessing that the wiring is completely wrong on this one. I have the Trianglelab LDO clone pancake stepper LDO-36STH20-1004AHG(XH) and the wire colors are different from the real LDO Z motors and Stepperonline AB that I have. Spec sheet shows coils with A(red), C(grn) and B(yel), D(blue). So I need to double check that I have my plug wire correctly and figure out how to continuity check it. 

Edit: Issue #3 was bugging me so I couldn't leave it alone. Turns out the wiring had indeed crossed coils even though I carefully checked, double-checked, and triple-checked my harness followed the pigtail out of the motor. Flipped two wires to get the coils on the correct pins in the Octopus socket and it is now buzzing correctly. Naturally I had to change the "!" flag again (I think I removed it, now restored). Extruder is happy.

Edited February 23, 2022 by claudermilk

[Simon2.4]

@claudermilk I can almost smell the molten PLA! Good Job! 

[claudermilk]

Poked at it some more and I got the probe wiring figured out. I looked at the darn guide on the probe body a couple of times, ran wires and checked them at least three times. Then hooked the 0V (gnd?) wire to signal and signal to gnd. So of course it didn't trigger. I did build my little patch cable with the BAT85, so the probe is connected per the Voron docs wiring guide. Poof! Now it works! That resolved the last wiring hiccup and everything checks out.

With all wiring checking out, I continued through the Initial Startup process. It moves! It homes!

I got everything calibrated, z position set, and it's time for a print.

Oh, and while I was having a think about the goofed up probe and endstop wiring, I played around with KlipperScreen customization. It's pretty neat, and easy since it uses CSS. I might go back and tweak some more, but this is an initial setup.

So, without further ado: the first print start.

 

I did have to cancel after about 10-15 minutes because the cube broke loose. I didn't clean the plate well enough it seems. But that isn't a big deal, now it's to tuning and customization. Oh, and soon .

 

 

[Demosth]

it's alive!!!!

Congrats!

[claudermilk]

I am super chuffed. 

Now I get to transition from Voron wannabe/observer to Voron builder/owner.

[claudermilk]

Oh, almost forgot. I also had my first minor mishap with it. Trying to reprint the cube I noticed Z endstop homing was behaving a bit odd. Turns out the Sexbolt pin had gotten stuck, so wasn't pushing & releasing the microswitch. I'm not sure why it didn't just fail to home, but it made the moves, slightly flexed the toolhead and went on its merry way.

I then ran into the bed being a bit too racked and the z tilt adjust failed. Turns out the back was a bit too much higher than the fronts. A manual adjustment there resolved that error.

So with the Z endstop malfunctioning it didn't home right and I got a mild head crash. What an awful sound, your brand new Revo nozzle dragging across the textured bed.  I canceled that right quick. Then disassembled the Sexbolt assembly to see what was going on. Somehow there was some kind of film between the Chicago screw and bronze bushings. I also noted some ABS stuck in the threads. Weird, but likely due to me filing the housing a bit to get things fit where I wanted. I cleaned it all up and put a tiny dab of PTFE Super Lube to assist. Seems to be working smoothly now. I do have a spare flex plate knowing it was likely I'd have a crash or two before getting all dialed in--don't ask how I knew to anticipate this.

[claudermilk]

Next challenge is the extruder. I can get about 30 minutes give or take before it start printing air. Runs fine, then starts clicking and grinding for some reason. I printed up wile.e's latest parts (new body, motor plate, guidlers) and will give another try. Upon initial build, it seems to run smoother. Once I get that figured out it's off to the races.

On the plus side, homing and z tilt adjust are working reliably now.

[Simon2.4]

3 hours ago, claudermilk said:

starts clicking and grinding for some reason

I had problems with this and my problem was that retraction must be less than 1mm for direct drive. Cura had it set at 6.5mm for bowden setups...

[claudermilk]

Retraction was at default of 0.75 for the available PrusaSlicer profiles.

So, I reached out on Discord with the issues and got a bit corrected in SB beta, but also got assistance. wile.e pointed me to his latest guidler parts and the main body & motor plate parts. Logan BC also provided great help when I got stuck in these last phases. I went ahead and printed those then rebuilt the CW2. The guidler changes are nice updates and work well. I got started trying to print again today and....

Behold the output of my now fully-functional Trident!

New CW2 works perfectly, now for tuning. It is under extruding a bit and I struggled to get the prints to stick.