Trident 250 -- My First Voron

[brendanm720]

Update 13


What in the ever-loving !@#$?!

[two days earlier]

I began work on wiring up the inductive probe. I took a 3 pin JST socket and soldered it to the 24V line, the Ground, and the data wires, and put plug on the wires coming out of the probe. It's not perfect, but it detects my Leatherman, so I guess it works. It seems to work fine in Klipper also, so there's that.


Next, I attached the bed. The Manual calls out M3x16s for attaching the bed, but with this particular bed, we're looking at M3x20s. I ended up stealing one from the Z Endstop pod (I ran out) and replacing it with a M3x30 and a nylock nut on the pod.


It was then I tried to do a Home All...

...And the cooling ducts went Z-Negative-China. It didn't take me long to figure out why -- or what I thought was why.


It did take me a bit to figure out a work around. I tried a number of things, and ended up finding that a SHCS 3x30 mm screw was just about the right height (what with the socket head being around 5mm), but wasn't big enough around and didn't always actuate the switch. The M5x30 screws I had on hand are BHCS and are thus still not long enough (button heads are not as tall as socket heads). I need something right around 35mm long, and I just so happened to have some M5x40s.

Time for some "Custom Machining".


That's Better.


It was about this time I noticed (because the toolhead was now missing the endstop and still going Z-Negative China) that we were homing x, then coming off of the side 15mm (instead of 10) and then homing Y and backing off 5mm... but Klipper seemed to think it was at max travel (250, 250). After a quick question to discord later, it turns out I missed this very important step.

So that fixed things. Mostly.

During troubleshooting, I realized that the printer now thought that the back corner was 250,250... and the nozzle was off the bed. So I changed the max travel # to 255,255, and now we're back in business. 0,0 is right up at the corner of the bed where it's supposed to be, and 250,250 is in the back corner, where it's supposed to be. I did notice that I can't tell the printer to go all the way to the front (past the bed), but I can fix that later, if I feel like it.

I noticed that the nozzle wasn't quite hitting the endstop and instead of changing the settings in the config file like a normal person, I made a new and improved version of the endstop using the spare M5 nuts that came in my BDF Kit.


Anyway. I homed the thing about 10 times with no further issue (well, one time the bed was up too high and the gantry was all the way back and the nozzle snagged on the new-and-improved endstop pin and carnage ensued, but that was self-inflicted).

Then I tested Z Tilt. It worked. I did the probe accuracy thing. It was better than the numbers the Voron Pre-Flight checklist said I should be better than.

Then I hot-tightened the nozzle, and put the sock back on.

And broke the bottom zip tie loop thingie.

I PID Tuned the bed, and then the nozzle.

I calibrated the Z Height (it was waaaaaay off).

I did NOT calibrate my E-Steps. I too like to live dangerously.

I downloaded SuperSlicer and sliced up the obligatory and ubiquitous Voron Cube (v7). (I did manage to not put a solid top on it, but It's the first time I've used a Slic3r-based slicer, so... )

I sent it.

To Be Continued...

 

[brendanm720]

Update 13, Part Deux

Apparently you can only do 10 images per entry. So, here's another entry.

The print completed successfully, and judging by the dimensions of the cube (which were really close, acutally) I was over extruding just a teeny tiny bit.

This thing, without being broken in and without any tuning or input shaping or pressure advance, already prints better than my Flashforge does.

I adjusted the extruder's rotational distance. Twice. But now it's right on the money.

So, I sent another job -- some panel clips. They turned out pretty good. There was an adhesion issue with one of the center clips and it came off the bed and caused some inclusion of spaghetti in a corner clip, but that can be cleaned up.

And then I tried to send another cube (this time sliced with a top), and disaster struck.

It homed. It went to do a Z-Tilt. The probe was already triggered. MCU EBB36 communications lost.

Whiskey. Tango. Foxtrot.

So, I shut down the pi, and turn off the printer. I check my connections (one of the can connections on the MicroFit on the EBB36 isn't staying in its hole very well. [sigh] I unplug it, plug it back in, ensuring that pin goes in all the way, and turn on the printer.

It boots up. It talks to both of the MCUs. The probe isn't lighting up when I stick a 13/16" wrench under it and has a bad solder joint in that 3 pin socket (the ground is wiggly). Also, this is apparently a thing:

Being a dumbass, I used my finger to check if it really was that hot or if it was an error (it was hot enough to give me a minor 1st degree burn on my fingertip from just tapping it, so yes, it was probably somewhere near that hot). In retrospect, it would have been better to go and grab my wife's soaping IR thermometer, but now I have a nice callous on my fingertip. As the Scots on Outlander say, "Dinna be fashed. I amno'"

After a quick consultation in discord, we confirmed my thinking that this board was probably a lost cause and I ordered a replacement. I received the replacement yesterday and it is indeed the newest revision (I was slightly concerned as it was Amazon), which is nice.

So, my thoughts as to why the board failed are thus:

1. It's just a bad chip, and it was on and working long enough that it failed. If there's a fault in one of the components, it'll fail pretty quickly (like within the first few days to a few weeks).
2. I shorted something at some point, and managed to short something internally in the chip.

I'll have to be careful with this new one, and we'll see what's what.

The new board comes with new MicroFits, so I'll reterminate that end with the loose CAN pin. I also picked up a 5 pack of the same MicroFits (but also with the other end) and I'll use those for the probe. It's a four pin instead of a 3 pin, but it won't matter if one of the pins isn't populated and it'll be 100% less prone to bad solder joints.

In other news, I ordered a Tap Kit (Finally) from West3D and grabbed a Koil of KVP Galaxy Blue while I was at it. My next print, after I get this thing printing again, will be a cube with a top on it. The second thing I print will be a Slant Spool for that Koil that's coming. BrewWeasel used MatterHackers Build Series ABS for the black parts; I find that they're pretty inexpensive per spool (which is probably why he uses it), and I can get free shipping if I buy 2. The plan is to reprint all of the required chamber parts (gantry, Z, Bed, et cetera) in the event that something breaks on me (as this will be my only printer for the foreseeable future that is capable of printing ABS), as well as skirts, clips, Tap, SB and CW2, et cetera. The spares will go back into the PIF box.

I also need to have a look at the ERCF repo -- my next project after the printer is "complete" will be an ERCF (and probably also an ERCP or some other buffering system) and I need to print the SB/CW2 parts for the Hall Effect sensor (and also pick up a hall effect sensor). I figure if I put that in when I switch over to CW2, it'll be less dicking around later when I go to install the ERCF. Of course, it'll be more dicking around now... Anyway. I'll add that hall-effect sensor to my next DigiKey order. I've already got one started.

So, next steps:

1. Flash CanBoot on the new EBB36 via USB connection to RasPi. Remove USB Connection and move the jumper from the 5V pins to the 120Ω pins.
2. Re-Terminate the Umbilical with a new MicroFit.
3. Re-Terminate the Probe Harnesses with a new MicroFits.
4. Reconnect all of the things to the EBB36 and mount it to the rear of the toolhead.
5. Reconnect the Umbilical to the EBB36.
6. Shout "Fire in the hole" and turn the printer on.
7. Watch/Sniff for smoke
8. Flash Klipper on the new EBB36 via CanBoot
9. Print some !@#$.

 

[brendanm720]

Update 14

So, I got around to flashing CANBoot onto the EBB36 last night, and reinstalled all of the things. I mixed up the fans -- when I go to SB, I think I'll need to put the PCF pins in a white connector so I can keep them in the correct spots.



I reterminated the umbilical end that plugs into the EBB36, and the loose pin issue is now a thing of the past. Probably.

Bad Crimp! Bad! [rolls up newspaper]

I also reterminated the PL08N2 with something a bit more suitable for an inline connection. Seems like it'll hold up well now.


Homing and Z Tilt work much better now.

After that, I printed the cube again and went to bed.



It has a top this time!


After cleaning up the bed, I figured out how to enable bed mesh, and did one. Looks like I may need to tweak the Y Gantry a bit.



Now it's printing stuff. It made some clips this morning.



And it made a ball of death this afternoon. I caught it in time, so everything is okay. Might need to wash the PEI in soap and water, though.


I did order some Black Matterhackers Build Series ABS also.

Next Steps:

1. Get the toolhead wires cleaned up.
2. Wash the PEI.
3. Get some temporary panel clips printed.
4. Put the foam on the panels, and put the panels on the printer.
5. Find my roll of blue painter's tape for the doors (Because I'm not going to fart around with mounting them with PLA or PETG).