Chrident (Chiron Trident mashup)

[west quest]

Hi,
This is my Trident inspired build I have started, using parts from old Anycubic Chrion printers. I have built a Trident and I have used a Chrion for a long time. There are things I like about them both, so why not Voltron them together!? ... Don't answer that.

My goal really is to reuse as many parts from the Chrions as I can (where it makes sense to do so), keep the budget "low", and still be Voron. Part of the reason I am posting this log is to help share my build with friends and other contacts. The other part is to get advice on some specific 'challenges' I get myself into. Also, I may shoot my eye out, so I am sure someone will find that entertaining.

So, first thing. Why am I building this at all?

Probably first thing, is that my Trident (SB), while a beautiful machine, has limited capability in printing large PLA prints fast. I don't want to sacrifice the quite stellar ability it has to print ABS/ASA in its current form. I did experiment with a remote cooling MOD using a turbine from a CPAP machine, but while it "worked", it was making sacrifices I didn't like. So, the remote cooling will be diverted to the new build.

Next, why Chiron(s)?

A few reasons. One, I am familiar with it. Two, it has a lot of extrusions. Like a lot. Three, I like the build plate.

Edit: Actually four, I just like upcycling things.

I got four non working Chirons and one working one (still in use) that I can use to cobble a lot of the base structure, motors, print plate, heater, power supplies, screws, wiring etc. This would save a lot of cost, which means I can divert more of my budget to the things that really matter, like decent rails, tool head and electronics.

After tearing everything down, I determined my longest extrusion length was 630mm on the 2040. I also figured, that this machine did not need an enclosure, so I did not need to worry about conforming to the outer limits what Trident would normally have, including the section that sits over the gantry. So I cut out the top side of the design, to save about 130mm of vertical length. So I used the design calculator to have a build volume of 405x425x500, which put the longest extrusions at 760mm - 130mm to get to my 630mm that I actually have. I then modelled a mockup of what the extrusions all would look like in this configuration, with some additional extrusions for rigidity.

Lots of bits.

[west quest]

I cut the extrusions a using a hack saw, and tuned their size with a bench sander. I then removed the burrs with a bench grinder. I also tapped the ends by hand. I almost screwed up one of the holes when I first stared, as I forgot to use a fluid (WD-40) to help the tap cut smoothly. Aluminium gets really gummy when cut without a lubricant/coolant. In the end though, the frame came together really well. In fact, the 2040 extrusions are great to square up. Getting my original Trident to square up took hours.

[west quest]

I did manage to assemble half of the frame with the M5 bolts from the Chirons, but I did have to get a fastener kit from Siboor on Aliexpress, which was my first purchase. The Chirons are filled with many M4 bolts, but there is very little opportunity to use them here.

[west quest]

The rear gantry parts are stock, despite the 2040 extrusions. I used Fushi bearings and pulleys from the Siboor kit. The two 48mm Nema 17 motors are from the Y motors that drive the bed on the Chiron. However, the one on the right I actually replaced later as I found it had a wobble in the shaft. That ended up being a chore, as I only had three 48mm motors, and the third one had a fixed pully. I had to cut it off with a Dremel and then cut a notch for the grub screw.

I will be using the inverted electronics MOD, and as a bonus, there is a remix on Printables for extra long ones for Z motors with couplers. So Kudos to this remix.
https://www.printables.com/model/782380-voron-trident-inverted-electronics-z-motor-mounts

While this does move the motors down quite a lot, the extra 20mms from the 2040 extrusions allow for plenty of room, so this model was perfect.

Also, I want to point out I printed all parts in Elegoo ASA. This is the first time printing with ASA, and they were the cheapest option. I quite like the finish over ABS, and it prints very well. Once I dialed it in, it has been very predictable and failure free. The only thing that would make me happier is if Elegoo would offer more than nior/blanc.

[west quest]

One challenge I had with the extrusions is that they are all v groove T slots. They will not fit 9mm rails. I could have printed slot gap brackets to fit the 9mm rails, but that seemed more effort than it was worth. Instead, I found MGN12 models for both the Z sliders and the Y axis mounts.

https://www.printables.com/model/454583-voron-trident-xy-gantry-conversion-to-mgn12h-rails

https://www.printables.com/model/454574-voron-trident-z-carriage-for-mgn12h-linear-rails

So I ordered all MGN12 rails.

These are RDBB rails from the CNA store on Aliexpress. They appear quite decent. They are either carbon or bearing steel, and the cars have a grease port on the sides that actually work. My only complaint with them is the wipers have a bit of a gap that will easily let dust in. I do not anticipate that this will be an issue for me though. While I am not fitting an enclosure to the printer, I do have a dedicated spot for it which is enclosed that I built from a metal frame shelving unit which is wrapped in fabric.

They did not have an option for the exact sizes I needed, so I did have to trim them down. I used a reinforced cutting wheel for the Dremel, which was actually spectacular in going through this steel. I finished the edges with a fine grit grinding wheel.

Both the MGN12 mod and rail fit perfectly.

There were a large number of M3 8mm screws recycled for putting the rails and gantry parts together.

The rear extrusion is a 2040, mainly because none of the 2020s were long enough. So I offset the 2040 to one side and cut a notch in the bottom to fit over the plastic bracket, which ended up being about 8mm. Everything lined up perfectly. I did have to cut a groove to allow the Z bracket screw to be inserted and removed.

[west quest]

The T frame for the bed came together fine. I also used BFI idlers.

https://github.com/clee/VoronBFI

I printed the version that uses a BHSC M5 bolt. Incidentally the Siboor button heads were too big, but the salvaged M5 16mm BHSC had a smaller dome and fit perfectly.

However the bed mount I am going to come back to, as that is going to be a particular challenge, since I am reusing the existing Chiron bed.

[west quest]

I printed the feet from TPU using this model, but added some M5 washers.

https://www.printables.com/model/126659-rubber-feet

And belted. Pretty much where the build is now. Currently the belt is tied to a Voron Tap back plate. I do not have the rail yet, as it appears to be stuck in transit somewhere in the country. It left customs last week and no updates on the tracking since.

[sudden edge]

Isn’t that heat plate. Upside down?

[west quest]

In the mean time, I have received some parts:
Triangle Labs Rapido ACE HF
Fysetc Spider 3.0 H7
BTT TMC2209
Orbiter V2 with Nema 14
Mini 12684
POM T8 Nuts

I am not sure if I should have gone with the Orbiter, or just gone with an LDO Nema 14 and built a Sherpa Mini. In either case, I tried hard to find extruder designs that worked with a Nema 17 pancake that I already had, but would be compatible with TAP, but in the end it would have required me to redesign a lot more things that I would have liked. The orbiter was discounted a bit, so it wasn't too much more.

[west quest]

Indeed it is, and not fixed to anything. It is just being stored there right now. This is an area I have have yet to solve actually.

[sudden edge]

Gotcha. Just was slightly confused on how you were gonna print over wires.

[west quest]

The bolts that go through the plate do not line up with the standard T bracket, so I need to do some brain storming to figure out how I am going to get it attached with the current bolt arrangement, or if there is an alternate arrangement I can drill holes in that doesn't go through those thick traces.

[sudden edge]

You basically a need a hole somewhere along where those thermistor traces are. So you could run a different thermistor to the heater and drill through those traces or find just the right spot a hole will fit.

[sudden edge]

Anywhere along this line.

[west quest]

Yes, one of the things I contemplated.

Here is a better photo.

And lined up with the T.

[sudden edge]

Wonder if there’s enough space for a hole there

[west quest]

There should be. The bolts that go through this in particular are flanged and flat topped M5, which recesses into a countersunk hole. But the plate is quite flimsy, about 1.1mm thick. Even with some 468mp adhesive, I am concerned that the heater plate might bend at the anchor point this close to the edge. But some really wide washers might mitigate that.

However the other two anchor points down lower are just too close to the traces for comfort.

But I might be able to fit some small 2020 extrusions, bolt them down really tight, and then have the existing bolts go through those extrusions.

The seem to line up perfectly.

Placement and length not representative, but 10mm offset does seem perfect. But will it be rigid enough?

[sudden edge]

I think the best way to find out will be to do it.

[west quest]

lol

Well, eventually anyway. I am a bit of a ways from getting to the point to determine that.

[west quest]

I have four of these Cheng Liang 800W, and four of these TDX 91W, 24V power supplies. On the Chiron, the 800W power supply fed the bed through a very large mosfet. However, the control board used this 91W laptop charger.

The bed will be obvious, using the 800W and mosfet. The rest, I am not yet sure. I need to power:
Control board + possibly RPI via 5V rails
Fans
CAN Tool board - EBB36
Rapido ACE hot end, which I think is ~82W heater.
5 Motors
Turbine - 98W, though peak amps can be as much as 8a

It seems pretty obvious these little laptop bricks are... just that, bricks. Maybe I can use it to power some 24V LEDs later down the line.

So, I can bring in another 800W power supply to power everything else. I had concerns initially using this for the electronics, as the Chiron has two power supplies. But then I figured once the bed turns on, the voltage fluctuations are probably pretty gnarly. Also, Cheng Liang is used across many other printers in the industry, most which appear to power the electronics, so this should be fine? 800W does seem a little overkill.

[west quest]

Time for Pi.

I got the Spider V3 H7. I had looked for a V2.x board, since that is what I am used to, but the Aliexpress listings didn't have much. Also, the V3 has the CAN expander built in, so that was a plus. However some people had issues with the CAN connections on the V3, so I figured I would look at the H7, as it might have more improvements on that. Plus, it was only $3 more.

Turns out documentation is still a bit spotty with this one. The WIKI URL at the bottom of the box goes 404. The Github URL works, but information is bare bones, and the pinout diagram is barely legible. The 2.x boards had a lot better documentation and reference images. I know Fysetc can be inconsistent, but this is ridiculous.

[west quest]

Turns out flashing this thing isn't as straight forward as other Spider boards. After setting the U5V and 3.3v/BTO jumpers, it could be detected over USB on the PI, but not as DFU mode. After a day of periodic research, and trial and error, I figured it out. There is a BTO button that requires pushing while powering on the board. This sets DFU mode, but prevents the board from being recognised as USB serial device. Then running dfu-util -l will search for DFU devices, which it did find, and I copied the ID. Then I ran the make flash to that ID. This method seems to be similar to flashing a CAN bridge board.

[west quest]

As a sanity check, I checked the pinouts on the BTT mini12864 vs the Spider board.

Am I reading this right? Is the BTT mini12864 a complete 180? The cable provided does not flip.

[west quest]

https://wiki.fysetc.com/BIQU12864_On_Spider/
This does seem to be a thing. Also they note that the R1 and C6 SMDs should be removed. There isn't an R1, but there is a C6. They also mention swapping the cables as well between EXP1 and EXP2. The pinout does not seem to suggest that is necessary. It appears to be an error on the Spider V1 board.

[west quest]

Totally clean test setup.

Wait this worked first go?

Granted, I don't have any drivers installed.

[west quest]

Well, the Mini 12864 works, as well as the reset button. I wasn't prepared for this, now I need to pop a driver in and test some motors with stepper_buzz.

[west quest]

I have seven 40mm motors with lead screws attached, and four of them appear to have bent shafts. Fortunately, three are straight, and the driver_buzz worked fine on those. Even if they didn't, I have four other 40mm I could swap out if something went bad on them. I will need to test the X/Y motors later, since they are already attached to the printer, and my jank setup won't reach.

[west quest]

Revisiting the bed brainstorm, since the electronics is going so well; Eyeballing this from this angle looks promising.

Drilling these three spots should be doable.

I will just need to drill through the extrusions at those points and have the bolts pass through, and fasten from the bottom. I have some aluminium spacers that should work nicely to keep things properly spaced, as well as some PC plastic washers that may help with preventing heat transfer.

[west quest]

And like that, it is done. However, I have sheets of 3M 468MP on order, so I cannot attach this to the glass bed right now. I also have sheets of under bed insulation, but I will not put that on until I am 100% sure that it is all working, so not to waste them if something is wrong with this heater plate.

[sudden edge]

You don’t want insulation on the bottom of the plate. You want that to heat up the chamber.

[west quest]

There will be no chamber.

This will be printing mostly PLA, TPU and maybe PETG.

I have things I print periodically that requires the build size of the Chiron, which my Trident 350 doesn't doesn't quite have the the dimensions for, especially Z.

[west quest]

Thinking on panels:
I don't need to enclose the printer, but I still want panels on the bottom, on the deck, and maybe on the back. The bottom and deck panels would be about 540x525mm. Voron specs suggest 3mm, but at this size, I wonder if that is enough? Maybe for the bottom it is, as I can just foam tape that to the DIN rails. But for the deck, because I am doing the inverted electronics Mod, I am wondering if 3mm acrylic is stiff enough for the deck panel?

[west quest]

Well, it turns out things were going too well. My Trident shredded a bearing, subsequently destroying the bearing next to it. Also dumping bearing grease everywhere. So a pair of the shiny new F695s I put in to the Chrident had to be reclaimed for the Trident. I still need to print parts and other things, so it can't be down. I have another 20 on order from Aliexpress, so I should see those in a little over a week.

However the adhesive arrived. While not the "right" size, there were not a lot of options that were not very expensive or excessively long sheets. This Gizmodorks was on one of the Amazon sales, which per square CM, was cheaper (relatively), than getting sheets that were "right" (500mmx500mm). Due to the unusual bed shape, 400mm square wasn't going to cut it... well, I would have to cut it, but even more awkward. The other options were to spend $250 on 55yards of the right size, or 5 yard rolls with a 150mm width.

Point is, it was cost effective, and someone hold my beer.

[sudden edge]

I actually thought about ordering a pack of magnet sheets to do that same thing on a bad 400mm printer I own. Lol.

[west quest]

I have considered that as well, actually. I will start off with seeing how I go with the Ultrabase for a bit. I picked the one that looked the least thrashed. Worse case, I stick a magnetic sheet on over the glass and spring steel.

[west quest]

Power switch went in fine. I wasn't sure if the Chiron had an odd shape, but apparently it is standard.

Z motors, lead screws and POM nuts installed. The right one has a slight wobble, and the rear every so slight. Left one is nearly perfect.

I also installed the DIN rails. I think I will need three due to having two of these power supplies. They are bulky, but fortunately, I think they should fit fine due to the extra 20mm depth.

The fan in these power supplies are loud. I am thinking of opening them up, or printing a vented enclosure. Just depends how I am going to mount external fans. One side of the PSU is solid, so airflow could be challenging if coming from another source, but it is not like they normally have a large amount of ventilation anyway.

Curiously, the positive and negative lines have loop wires on them. This makes me wonder if they have more than one rail. I don't know why they would do that otherwise.

[west quest]

This is all connected to the same power plane. It seems really weird to have this loop there.

[west quest]

Trying component placement.

[west quest]

I reused some brackets to attach to the standard LRS 200 DIN mount by drilling holes and tapping them.

I will need to mount the MOSFET somehow, but I think I am going to have to model a bracket similar to the RPI.

[west quest]

This should do it.

[west quest]

Electronics PSU installed. PI and MB clipped in. Drivers installed and tested all motors with stepper buzz. All checks out. I might actually run a line for the PSU fan to the controller so that it only runs when actually printing. Similar for the PSU that will run the bed, though it will need to have a different trigger and cutoff value. Two of these going is going to be loud. I am already going to be challenged working out a way to get the part cooling turbine noise dampened.

The A motor cable is also too short. I managed for all motors except that one. I have a short length cable I will sacrifice to solder on as an extension.

Chrident (Chiron Trident mashup)

[west quest]

My rail and car have finally come in. I installed the rail, but unfortunately there is no point proceeding further as it appears the car obstructs the four screws that fasten the back of the the TAP to the MGN12 rail.

And since I am still waiting on replacement F695 bearings, this is now on pause again.

Shifting focus. I opted to go for the XOL PAP toolhead. I had originally thought to go with the Rapido v2. But Triangle labs had their Rapido ACE on one of many sales that Aliexpress does, though this time it was actually notably cheaper, even without the bonus discount code.

It was a gamble to get, since the XOL only has models to support the Rapido, and the ACE is 4mm shorter. However the is rectified by the an included spacer. Also the hole pattern in the heatsink is interesting, as it allows for more flexibility in the install rotation angle. I don't know if the Rapido from Phaetus has that, but it helped here, since the XOL frame is very tight.

I also spent a long time trying to figure out a use for my Nema 17 pancake motor I had left over from an Afterburner toolhead. I couldn't find a good model that would fit the XOL extruder mounts, and support TAP. I am not skilled enough in any CAD program to mod something which were already elegantly designed for a Nema 14 motor. I gave up and decided just to get the motor, and then gave in and got an orbiter v2, since it wasn't a whole lot more.

I am still not sure I made the sensible choice on that, but it is here.

[sudden edge]

You could go Bowden with a M4 extruder with that pancake motor.

[west quest]

I wouldn't bother with the pancake in a bowden setup, simply because I have eight other 40mm motors. But I want to do direct drive on this build. I have an upgraded bowden on my working Chiron, and it is still awful to tune the retractions, and TPU is impractical anything below 95A.

[west quest]

I had looked around a lot on toolhead designs that supported remote cooling with a turbine. There isn't a whole lot of choices unfortunately. I had tried modding the stealthburner toolhead on my trident previously. It did work, but I think there were problems with flow restrictions around duct inlets for the hotend mount. There just wasn't space to keep it in the footprint and make the pathway less restrictive. I did try, but I could not get FreeCAD to behave with the model and make it reasonably printable.

XOL seemed to be the best option. It appears to be well designed and supports the extruder, hot end and TAP options I had in mind. My only gripe is the use of M2.5 screws and they strongly recommend a particular model of Delta fan. I can understand why they recommend that fan, but I think the lack of modularity in the hotend heatsink fan means you are stuck with only one viable 2510 model in existence.

The Delta fan also isn't readily available in my country, and has to be ordered, and imported from the US. I am going to try a local alternative, but I may still have to order a Delta fan for it, or make a mod for it. I searched for a mod that supports a larger fan option but mostly came up empty handed. While they do exist for the standard frame, they do not exist for the PAP version. Bummer...

I might be able to make an extension piece that can extend in front further to taper up to a suitable 30mm or 40mm fan. But I am not yet any good with curving CAD work, so it is something I have to work at before I can make something usable.

On the bright side, designing a bracket for the MOSFET PCB was easy.

[lucid horizon]

btw if you are looking for like minded we kinda gathered in a thread in voronmods for Ender3 Trident conversions but we are build other stuff too#1296518261566996491

as for the delta, they ware worth it (iirc 6 for free shipping with 3 days from US to Germany)

btw people like the 2510 because the dead zone (center) is smaller than a 3010 for example which helps big time for things like a dragon heatbreak

[west quest]

The Delta fan is by no doubts an excellent fan. One thing that gives me pause is that it is an oil filled sleeve bearing. Bearings of this type will always experience dust intrusion eventually, though I expect this one will prevent it far better than typical "cheap" fans. It is of course one way they reduce the hub size, since ball bearings take up space. Mostly my complaint is the lack of options on XOL for the hotend fans. And I know that some options would mean it compromises on features that the XOL is trying to excel at, especially size and weight, but it would be nice to be able to make that choice on tradeoffs myself. The X axis assembly is 560mm across, I am not going to get the full benefit of that weight and size advantage anyway.

Also, I need one fan, not six. Part of this build is trying to be conservative on the budget (and the exchange rate from USD is pretty terrible), while reusing what I have as much as I can. There is no getting around getting a 2510 for the XOL right now, but I definitely do not need that many spares. Or at least I would hope not. For now I got a rebranded fan, similar to the Honeybadger (possibly same OEM). And since I am using a Rapido Ace, I am less concerned if it will be sufficient.

I have a dragon on my Trident, provided by Fysetc. The fan provided in the SB kit was useless, since the dragon needs a blowy-matron but I was given a mild draft fan. Even with a decent fan, I couldn't print TPU on that hotend, since it would just gum up in the heat break. I have a spare V6 hotend that I swap in if I need to do TPU jobs, until I finish building the Chriton. This is one reason I went for Rapido.

I will check out that thread.

[torn otter]

This is cool, but there are a couple of errors. The delta is a sleeve bearing, and is 15,000 RPM (not 10,500). Also the Honey badger is more like 100dB-A at a really horrible pitch that you can hear from the neighbours yard. At least that's what I've seen in other peoples videos 😂

[west quest]

Honey badger (clone?) spins at 20k RPM. Maybe in my old age, I won't hear it. 😛

But then, my part cooling turbine is going to trounce it in every metric, so probably a moot point on the noise department.

[torn otter]

oh yeah, external cpap type fans are going drown out pretty much everything else.

[lucid horizon]

👀 yeah mine is 15k rpm too. not my sheet tho, blame Wight554 in armschair discord

BlobBlush

[west quest]

Bearings arrived, so replace that idler that was missing a pair.

I want to test the heater bed today. I am just not sure how to yet. I have not added the adhesive to glue it to the glass, as I want to test it before I do so. But that means there is essentially no thermal mass to absorb any heat. I am considering grabbing some clamps and gently clamping the heater plate to the glass bed for some testing.

Maybe some wood to keep them from damaging something.

[lucid horizon]

measuring resistance should be enough for testing purpose. there is no real magic in it, just a long ass wire

[west quest]

The thermistor line is built into the heater plate, I want to test that too.

Part of it is stability, as I want to see if the wires that are already there are good enough to keep.

Then there is the MOSFET.

First thing is to test to make sure that when provided power without a signal, it output stays unpowered. I will do that with volt metre first.

Well, 26v is a bit high. Glad I tested that first.

[west quest]

Getting a stable readout on the resistance of the bed is a bit difficult, and inconsistent. Depending how solid of a contact on the wires and my probe, it ranges from 1.8 - 1.5 ohms. In either case, this seems to suggest that the bed outputs no more than 400w, which is surprising. I had expected more, though I could never get an exact value from anywhere so nothing specifically said it should be. If this is accurate then I may only need one power supply, given that the voltage stays stable when heating the bed. Another thing I need to test. If the voltage isn't stable, I may still use both so I don't have voltage swings on the RPI and MCUs.

[west quest]

MOSFET checks out, no surprises. PWM 24v signal works fine on it. I set the max power to 1.0 and let the bed go to 60. While it climbed the PSU voltage went from 24.04v to 23.89v.

150mv should be fine for the MCU and the RPI that it is powering?

[west quest]

It is curious, since the Chiron has an independent PSU for the control board normally. However, the bed heater PSU is specced double what the bed really requires. So if there was that much headroom, why go with two power supplies for the Chiron? There must have been more to this, such as transients from the high power draw, but I wouldn't be able to measure that without special equipment. Perhaps I will stick with the two PSU setup. It is in there, and working fine.

[west quest]

Plate and heater plate scrubbed to get the old glue off with acetone, washed and dried with IPA. Applied the adhesive sheets. While I did remove air bubbles from the glass side, there is no way I can ensure, or even see any bubbles that may have formed on the aluminium side.

[west quest]

Print bed insulated and wired up. Cable chain installed for the bed. I am actually quite annoyed about this part. I lost my cable chains, and I have done a lot of tearing out and re-organising my boxes of printer parts. Nothing. I ended up grabbing the chains off my working Chiron for this. I will manage, as I can just suspend the cable harness. But on the bright side, the chain appears to be the perfect length.

CAN and tool board connected up and talking. Webcam set up. The find devices button in Mainsail is a really nice QoL addition.

Mostly what I need to do now is assemble the tool head. Well, first I should do the neopixel wiring.

Hmm, it calls for 26-28 awg, but the smallest I have is 24. But it is all FEP, so it will fit in the channels in the tool head.

[lucid horizon]

just cut a old usb cable 👀

[west quest]

USB cable would melt that close to the hotend. Or at the very least, after a while, it gets sticky and gross. I have plenty of FEP, and it worked.

[west quest]

Unklicky mod for TAP is working.

[west quest]

The toolhead is coming together. I had a couple odd inconsistencies from the XOL documentation. First one is that that the bolts that go into the stock TAP R8 are 16mm, with a few mm space once bottomed out. The XOL also documents 16mm bolts, but the spacing inside the toolhead bolt hole is at least 12mm long, leaving only ~4mm to screw into TAP. But TAP needs ~12mm to screw into both lower and upper parts. After some eyeballing and experimentation, I found that the 25mm M3 bolt works perfectly for this part.

[west quest]

Another was with the PTFE tube. The hole for it is just a touch too big. It drops in under gravity and the tube has a little wiggle room. With how spot on and tight everything else has been, this seemed odd. Also, the PTFE tube length suggested is quite a bit too long. There is a disclaimer that the lengths were determined by the extruder CAD models. Apparently this adjustable metal tube in the bottom of this thing isn't modelled correctly, or this is later revision of the V2? At least too long is better than too short and wasting a length.

[west quest]

I found some 3mm acrylic in the shed. Hacked it up with some hand tools. Not really pretty, but I only had to cut each section once. Note: peel is still on.

I found the first thing I overlooked somehow. There is no room here to put a nozzle brush.

[west quest]

When the limit switch is actuated, the nozzle stops here. Maybe a 1mm more. But this could be enough to put a wiper along the edge, so it can wipe the nozzle there. Maybe some PTFE tube attached to something. I have plenty of 468mp to stick it to the side of the bed.

Another thing I need to work on is this lead screw length. With the XOL ducts, it will collide on either front sides. It should be incredibly rare to need to print on either front corners, so in actual printing this may not matter. It is more a risk during homing procedures since default behaviour is to home X then Y. Homing Y then X would prevent that issue, or just limit Y by 15mm, but that doesn't eliminate the risk.

Otherwise, I can cut them back. I would just need to be more careful to ensure the bed is mostly level before homing Z.

[west quest]

If I remove 25mm from the rods, that will clear the ducts by about 1mm. I might remove the anti backlash part of the POM nut, which will leave me 20mm play on the rods.

[west quest]

Cut and ground back to the same shape as the original. Now no chance of collision.

[lucid horizon]

i put the ugly cuts into the coupler at the bottom so the nice factory one it at the top 👀

[west quest]

The ground ends are in the couplers. However, they look worse on camera due to being close up, so it wouldn't have mattered.

[west quest]

Now to start figuring out wiring, placement and test the turbine fan and controller. I had this working on my Trident previously so not much to figure out here on the electronics or configuration. But, one thing I need to work out later is a better enclosure for the turbine that also incorporates some noise management. The current one is just making sure there is a filter in place to prevent the motor from sucking up animal and people hair. There is no way to take it apart to clean it. It screams at 100% duty cycle.

[west quest]

I decided to move the RPI and put the BLDC controller in that spot, so the DC wires didn't need to be too long.

BLDC and turbine wired and hosed. Hose "mount" is temporary. The hose, cables and PTFE tube will need proper support eventually, but I am thinking of supporting it using a retractable key chain suspended from the underside of the shelf unit this will sit in. I will see what works.

[west quest]

I am still stuck on the nozzle cleaning problem.

1. Given that this is a glass bed, I can have the nozzle temperature up high enough to easily squish any filament out of the way of the nozzle, but that could mean I could end up with blobs all over the bed from probing. I could just have a sequence that probed one part of the bed out of the way while hot, then cool it down to ~150c and then probe. But this would still require manual cleaning occasionally no matter what.
2. Another thought is that I could stick a silicone strip somewhere on the edge of the bed. Either home made or something like the Bambu A1.
3. I might be able to stick some silicone to the right side of the bed on the edge. The nozzle might overhang just enough to make that possible. I could clean it at a higher temperature, then park the nozzle there until the hotend cooled down a bit before probing.
   (Silicone scrub and park ideas might require a two part silicone bit, one with rough edges for scrubbing, and another flat one for parking.)
4. Probably the thing that would be the most work would be to design an apparatus that would extend from the frame to wipe or scrub the nozzle, but would flex enough to move out of the way of the bed.

There is also the possibility of using PTFE tube(s) as a scrubber, but I have not though of a probable way to implement that yet.

[west quest]

I wired the fans from the PSUs to the control board instead of their internal header. I now set them to turn on when they power the respective heaters (Bed/Hotend). Printer homes X, Y and Z (probe). Took a little bit of time to get the bed parameters configured as I need to identify the real travel distances so I could plug them in to the Tilt configuration.
Total usable volume:
X=408
Y=424
Z=485

Ran a probe accuracy for a couple hours as recommended by the creator of Unklicky TAP. From that I figure I can use a .02 probe accuracy retry and a tilt retry of .01. I have run a few tilt checks and they complete with a few probe retries.

[west quest]

Did a shaper run just out of curiosity. Honestly expected worse.

[west quest]

Quick test print done. It lives.

Not wrong about that 25mm fan, it is a screamer. My turbine enclosure actually dampens down the noise from the part cooling fan, so the print head fan is probably the loudes part now.

[west quest]

Got tired of the toolhead fan real quick, so I created a mod to fit a 40mm. Running a flow ratio test while I give this a try to ensure it functions well.

[lucid horizon]

2510 is good because smaller hub means less dead zone behind it. Important for dragon btw

[west quest]

We have been over that, this isn't a dragon. Also, the air has to be funneled into the 25mm circle anyway, so the dead zone doesn't matter. The model has a cone to help prevent turbulence around the back of the hub similar to the SB design, and this fan is about 18mm away from the original inlet.

[west quest]

I have been tackling an issue with the bed for the past week off and on between holiday festivities.

There is a significant amount of deflection in the bed, which comes up with some really bad looking results in the mesh that are not anywhere as bad in reality. But because the bed moves from the TAP probe, it appears lower than it actually is. After a bit of an examination, the bed has a lot more give then I initially expected, and the 2020 extrusion also twists slightly.

To counteract this behaviour, I replaced the 2020 with a 2040, and added an under-bed brace. This changed the mesh dynamic significantly, except the rear left, mostly unchanged.

If I can shore up that corner, then I can work on a more permanent solution.

[torn otter]

Couldn’t source the delta 5v?

[west quest]

I could get it from a few places that ship internationally, but it just isn't practical from a budget stand point. I have plenty of 40mm fans on hand, the modded mount/ 40mm fan works fine and it is much quieter. Since this will never print in a heated chamber and isn't a Dragon hotend, I don't have any concerns. I have not had issues with heat creep, and I have checked that when the filament when pulled out does not have an extended swell area as I have seen on a dragon hotend. I will have to try it on TPU later, but I need to resolve the bed issues first.