Solved-More detail on hub cutting config

[weak nova]

@runic raptor-you have any insight? Can I get some clarity on what some of these values do and when they are used -AFC_hub-move_dis- not sure what this is used for if the system knows your bowden length and lane hub distance unless it's the amount to retract from the hub switch so the next lane can pass but it doesn't act as expected to changes -cut_dist=the length to extrude from the hub switch which will be cut by the cutter? -cut_clear=the amount to retract the filament after cut to clear the hub for the next lane, how does this interact withe the lanes dist_hub on the next load since using one distance on cut_clear will leave the filament at different distances between the inner and outer lanes -cut_min length-what does this do altogether? I have updated since the recent changes to this section were made, still not getting the expected results when changing some of these values

[runic raptor]

Can you reformat the text? This is really hard to read.

[weak nova]

@runic raptor under hub what is move_dist supposed to do and when during a cut

@runic raptor Any idea what cut_min_length does?

@runic raptor cut_dist seems like distance from the hub switch to cut I think that is working as expected now

@runic raptor And what is cut_clear supposed to do when

[runic raptor]

move_dis is the length of the steps in which the filament is moved towards the hub until it triggers. Just leave the default of 50. Changing that too much will most probably make everything slower.
cut_min_length is unused.
cust_dist is, exactly as you said, how far the filament is moved before the cust.
cut_clear is how far the filament is retracted after the cut. This should just move it back before the hub, so that another lane can be loaded. Too short => collision inside the hub. Too long => wastes time.

[weak nova]

@runic raptor Is there somewhere I can see the coding so I can understand what it's doing?

[runic raptor]

Sure. It's all in this file: https://github.com/ArmoredTurtle/AFC-Klipper-Add-On/blob/main/extras/AFC_hub.py

Unless you really want to spend lots of time, leave everything except for cust_dist and cut_clear on default.

Increase cut_dist until the tip that is cut off is long enough. Then play with cut_clear until (after cutting) you can see the filament come out of the hub (between extruder and hub).

[weak nova]

@runic raptor Aww so move_distance is steps not mms

[runic raptor]

No, that should be length in mms.

[weak nova]

@runic raptor So does it pull back to the extruder when unloading then move 50mms at a time towards the hub?

[runic raptor]

Yeah. It retracts quicky for the configured bowden length and, if I remember correctly, it continues in small steps until the hub is cleared. At that point, the filament is between extruder and hub.
When cutting, it moves in steps of move_dis length until the hub is triggered again. Then it goes back in smaller steps (10 mm), until the hub is cleared again. Then is once more moves towards the hub in 2 mm steps, until it is triggered.
This back and forth in ever smaller steps is an attempt to quickly get to a good reference position (directly when the hub triggers).

You could achieve the same by moving towards the hub in 1mm steps, but that would take forever.

[weak nova]

@runic raptor Ok I understand, now will the hub_clear affect the next time that lane loads or does it start counting Bowden length when it hits the hub switch?

[runic raptor]

When loading, it moves in small steps until the hub is triggered, then moves faster for the bowden tube length.
If hub_clear is too large, then it just needs more small steps, costing you a few seconds. That's why you want it to just clear the hub, but not more.

If it's too small, you'll have a collision inside the hub.

[weak nova]

@runic raptor Ok, so I'm trying to not use poop and have it change filament very precise to reduce oozing, so I see now I need to be careful with these numbers because if I change lets say hub clear by one mm it may produce more ooze than that since many of these moves are incremental they are not directly related to switch activation

[runic raptor]

Most small moves have a length of 10 mm. So that's the amount of error you have in positioning the filament in the toolhead. hub_clear does not really change that.

Also, I don't think that you can avoid purging filament completely. If you don't want to poop (and I dont' do that either, as it costs print volume), then use a purge bucket and a nozzle brush afterwards.

Don't forget: you wouldn't have to just tune one lane, but all four of them in tandem. I don't thing that's really feasable.

Of course, you could reduce the length of the small steps from 10mm down to 1mm or so, but that'll slow down loading/unloading immensly.

[weak nova]

@runic raptor So but it's making more sense now, like if I load a fresh spool and that lane hasn't been loaded yet, when it does the first time it will sometimes purge a bit, but after the first change it acts as expected, but I understand now that if I change something by even 1mm it could cause it to purge by several mms which I experienced but didn't understand why

[runic raptor]

Yeah, with the default config you'll have an error up to 10mm. Would be nice if that were smaller (e.g. just 1mm), but that would make everything so much slower.

[weak nova]

@runic raptor I've made an ooze stopper/nozzle wiper to help control any oozing, so I just need to be mindful when making changes so I don't get 10 mms of unintended purge

[runic raptor]

Which hotend do you have?

[weak nova]

@runic raptorI Set it to park there and a quick wipe after changing, I'll make a tray to catch any falling pieces that should handle many many color changes, you loose some bed space no matter what so a purge tower is no big deal

Dragon UHF, running with the extender at the moment as well, oozes a LOT

[runic raptor]

Yeah, me too. No matter what you do, it will ooze.

It took me some experimenting to find settings so that it makes its way from the park position to the slicer's wipe tower without creating a blob on that tower.

Large retraction, nozzle wipe, and fast travel moves 🙂

[weak nova]

@runic raptor This is helping drastically, I get very little unless I change something and it squirts out 10mms lol

It uses those cheap silicone bracelets

[runic raptor]

That's my setup. I've remixed something from Printables to add the brush and replaced the bucket with a chute.

[weak nova]

@runic raptor It's coming along, yeah whole different story on a Trident

[runic raptor]

Using a non-UHF hotend would probably make life much easier, but who wants to print slowly?

[weak nova]

@runic raptor Right! See others with Revos, man these prints are slow enough even at 30mms^3+ couldn't imagine at less than half that

[runic raptor]

I had a Revo (with HF nozzles) before I switched to the Dragon UHF. At least oozing wasn't a problem with those.

[weak nova]

@runic raptor That's funny great minds think alike thought the red in your picture was circling something, but you're also using a silicone bracelet, nice!

[runic raptor]

Yeah, I basically combined https://www.printables.com/model/692141-trident-silicone-nozzle-park and https://www.printables.com/model/882364-adjustable-gantry-mounted-nozzle-seal-parking-and.

[weak nova]

@runic raptor So are you parking on the silicone bracelet or using for a final wipe?

[runic raptor]

During toolchange I park over the bucket/chute. To clean, I use the brush in the middle. The bracelet is used to park the toolhead after a print, so that the nozzle stays clean.

[weak nova]

@runic raptor So far I've been parking on the bracelet even during tool changes and how it's oriented I wipe on both sides after, just put it on last night though so not quite dialed in yet

[runic raptor]

My setup isn't much older. Maybe 2-3 days now. So I'm sure I'll tune a bit more in the next days.

[weak nova]

@runic raptor Wondering how long the bracelet will hold up, we'll see, what size is your Trident?

[runic raptor]

Run part cooling fans at 100 % while parked there, and it should hold up forever.
The Trident is 300x300x300.

[weak nova]

@runic raptor Is there a way to have the fan ramp up for just parking, I usually run my fan about 40% after layer 5 for the entire print

@runic raptor I do during normal printing park, but during tool change park?

[runic raptor]

Ramping it up is the easy part. Restoring the previous value is more complicated. Does your macro just park, or also move back to the previous position?

[weak nova]

@runic raptor I'd have to check, didn't remember seeing that I could control the BT park, it appears to go from print to the corner of the purge tower then to park location and back same order after change

[runic raptor]

I've replaced nearly all of the default macros with my own ones. I just can't stop tinkering 😉

[weak nova]

@runic raptor I totally understand, I'm really good with the mechanical parts, but not so strong with the software side, so I usually can get the macros working but it's not something I'm great at, there appears to be more than one computer language involved and people call things differently and sometimes I just can't get it, like my stealthburner leds they work when I do a fresh klipper install then I get to messing with the macros or even just playing with the manual controls in mainsail and they stop working even after a full power cycle, I have no idea why, I'm running the standard Voron SB configs

[runic raptor]

That sounds crazy with your LEDs.

[weak nova]

@runic raptor FYI on the dragon UHF if you use the the HF heatbreak from the original Dragon and a volcano nozzle it'll give you an extra about 8mm of meltzone, since you got a 300 you may be able to use the extra flow

[runic raptor]

Yeah, to be precise, I do have the UUHF configuration 🙂

Not because of the higher flow, but because I wanted a ceramic heatbreak. But I'm not sad about the increased flow 😉

I print abrasives occasinally, and I didn't trust the Dragon UHF's default heatbreak.

[weak nova]

@runic raptor Sorry with the extender on the bottom you use the volcano adapter between the heatbreak and the extender

[runic raptor]

Exactly. That's what I have.

Such an easy upgrade.

Totally worth it.

[weak nova]

@runic raptor I really like the dragon, takes a minute to heat up even with a 70w heater but it's very versatile and with a decent fan I've not gotten any clogs like others have said happens

[runic raptor]

Yeah, hotend cooling is very important with the Dragon. But it's doable.

Never had a clog either.

[weak nova]

@runic raptor Think my next printer will be a 250 Trident with probably a dragon burner on a single mgn9 with the Dragon UUhf and 48v I think that will be a really quick prototyping machine

[runic raptor]

Really mgn9, not mgn12? Is that stiff enough?

A 250 certainly makes more sense for a quick machine than a 300. Less mass.

My prototyping printer is a Bambu A1 mini. No ABS, but for prototyping PLA is totally fine usually. And that thing is fast.

Also no ABS == no heatup time

[weak nova]

@runic raptor I've seen people use them them with real good shaper numbers, I don't think it would be enough with a stealthburner but with something lighter yeah

[runic raptor]

Yeah, personally I like the Xol and A4T toolheads.

Light (depending on the extruder) and very good cooling.

[weak nova]

@runic raptor I just can't get into Bambu, would take all the fun out if I could just hit print and it works lol, I like the tinkering part too much, I'd probably go orbiter or G2SA running a honey badger 5015 4wire like 9600 rpm PC fan on my SB it's a beast, might even make my own toolhead stripped down with two of those, hotend, extruder and board should be light enough, I like the idea of CPAP, but the hose flopping around in there bothers me

[runic raptor]

I wouldn't want a Bambu as sole printer, as I love the tinkering as well. But as second printer it's nice. It just works. For tinkering there's always the Trident.

Time for bed now. Was fun talking to you, but my bed is singing its seductive siren's song 😄

[runic raptor]

@weak nova: as you want filament loading to be as precise as possible to minimize oozing, this feature might be useful for you: https://github.com/ArmoredTurtle/AFC-Klipper-Add-On/blob/main/docs/Buffer_Ram_Sensor.md

Apparently this is much more precise than the usual 10 mm error margin.

[weak nova]

@runic raptor Yeah I was using this already, but turns out there was a mess gathering in my toolhead extruder and I got 2 bad jams in less than a day, so I'll be switching to a toolhead cutter and while I'm there I'm adding dual filament sensors in the extruder, so I should have way more control and cleaner swaps

[runic raptor]

I have filament sensors in the Extruder but am still using ramming, as it positions the filament more precisely.

But a toolhead cutter is certainly a good idea. Will go the route too in a few days.

And I certainly would add sensors when building a new toolhead

[ripe lark]

is this resolved?