#Flow - Born from an E5Pro
4925 messages · Page 5 of 5 (latest)
I'll get a manta or another skrat
(maybe manta with cm4)
more compact considering the printer footprint
wow! you are so efficient on cleaning up after yourself.. I was thinking to buy the frame for the beta enclosure and sell the old ender 5 if I manage to rebuild it somehow… but it looks like a lost cause… I’ll probably do it when I build the smolZG first
have you seen the new mellow board coming soon?
I made the commitment to myself of selling first then buying 😛
Mellow Fly-DP5
https://a.aliexpress.com/_mqCpWFi
there is also a version with regular drivers
with those drivers it's a no for sure
unless they are compatible with the ez (I highly doubt it) it’s not a smart move imo
i'll wait on reviews
but getting tied to a specific design...
I guess they want people to buy the version that locks them in with mellows style drivers
status: everything tested isolated (still need to home), endstops ok, bed ok, steppers ok, entire toolhead functions ok
next step rewire the TR for it's board
TR is rewired, need to flash the new board and test
@formal marsh saw your comment about the omnibox. Was it worth?
for me it is
Cool! Did you also make an enclosure for your printer?
Oh I thought you went and bought acrylic panels haha
haha nope, it's a giant MDF box
NEW THR ARRIVES TODAY
oh CAN oh CAN
love it when it works
hate when it doesnt
time to recheck all wires
120ohm is correct
ok wrong pins configured on skrat
now find why it shuts down
ADC out of range yet all temps seem fine on the logs
one thermistor wire just broke
so that can be it
taking katapult out to check
enclosure thermistor
now to find why its bork
fucked up a wire
at least everything else works
IT'S ALIVE!
show it man!!
one more thing to do, I split the dual 5015 in 2 fan ports so need to figure how to merge them in klipper
thank you, multi pin it is
stop, you're making me want to upgrade my horrible stock creality heatbed
I've only just got my printer up and running enough for me to be satisfied to print again
it might be cheaper 😛
I've been very tempted you seem to have great results with it!

Awesome thanks!
TODO 
Eventually reprint some omnibox parts as this PLA one is breaking (delaying this one until next release maybe)
Install Mellow Easy BRD on TR ✅
Print at least 2 ERCP for the TR/rewinders need to choose
Print SUAFT/something else with HEPA
finally getting smaller
FOUND WHY MY Z OFFSET FLOATS
so, doing a sexbolt repeatability test gives good results, same with the probe ones, since those are stable what's the remaining variable? the way klicky/unklicky attaches so I ran a different test, attach klicky do a probe repeatability test, dock klicky
on 10 tests got between 16.198 and 16.05, so a difference of 0.148 wayyy too much
accepting ideas
might have forgotten to glue the magnets, I don't think I ever did most of the times
yep, took them off, added glue now need to wait
I don’t think the glue would make a difference, if they stayed on, it means they were firmly attached to things
if they are loose, it may wobble?
that's my current assumption
meanwhile since glue starts setting pretty quick ran 2 tests
attach, 10 probes, dock - 10.times
then calculated range etc using the probes averages (final from each test)
both times range is around 0.01
need to run a few more
std dev 0.004
seems acceptable
again .01 range .003 std dev
I think I'll risk a print
Since you're running both a sexbolt and klicky, presumably you're running auto-z, so why does the z offset floating matter? Due to the way the calculation works, the only factor that should matter is the switch offset (which doesn't change) right?
the float is not consistent, however, so every pair of samples turns out wildly different.
the switch is not returning consistent samples
Ah that makes sense
Maybe I'll check the variation after docking/undocking on mine too, I just figured the slight differences were down to slight temperature variation within the enclosure, but I too don't have the magnets glued
because during the procedure you dock-undock thenprobe several times, every time you re-dock the probe you might have a slight different offset. Apparently glueing makes a big difference
to be fair, 0.2mm is a very small wobble for eyeballing with our senses. so it’s quite possible that without glue thing move slightly
in theory it should not matter indeed, even more because on a print I attach and run everything without reattaching, and all the measurements are made relative to nozzle+unbolt+unklicky, my suspicion is that sometimes during those measurements there is a slight change on the probe attachment, that is enough to throw off the calculations, remember we probe the bed a lot (z tilt, mesh, z calibration) and then klicky body also touches the sexbolt, somewhere in the middle changes may occur, 0.14 range variation is small but enough to mess a first layer, I'm expecting that now that measured range is 0.01 things get more stable
also I'm only probing the bed once
since the start print flow is already long and klicky repeatability tests showed good results I ended up setting only one probe
if it seems to show up again I'll try to set 2 probes and see if it improves
Ah I see, I take the median of 5 probes, same idea of doing everything without reattaching, 0.14 is indeed pretty high, I'm sometimes seeing variation of ~0.05
those 0.14 were between attach/reattach
I've been thinking about maybe getting a bdsensor so the klicky is only for auto-z but so far, with 3x3 adaptive bed meshing, unless I'm doing lots of short prints, the time taken is tolerable to me
Yes, I mean variations of ~0.05 in the z offset between prints
that is a good value
let's see, got chamber heating up
offset seems good now but was done cold, hot should not matter but I want to double check
so probe switch offset is wrong when the printer is hot, not making sense
That makes sense since the bed is possibly expanding when heated, no?
if it was really probe yeab
wrong name, its the switch offset
unless magnets are not enjoying the heat
as auto z should handle thermal expansion
My magnets (cheap amazon) hate the heat, I’m replacing them asap
Meaning within few months 😅
I always readjust the offset with a 1 layer 200x200 square. I do it once I change something on the probe
I’m waiting for people to adopt strain gauge in klipper
fuck me, the glued magnet decided to come off the duct
need to remove it from the toolhead
I had similar struggles, what glue are you using. I was using the wrong glue before, plasticglue. It only really stuck to the printed part and not the magnet.
yep that thing
but I don't have any CA right now
https://www.revell.de/index.php this one I think
for plastics it's good, doesn't damage like CA
yeah don't think that will bond well with magnets, maybe rough them up a bit on the side there will be glue?
good point I'll try it out
might try a different one just for now maybe I can find one slightly larger
sooo none is larger
Damn.. that would drive me nuts
they seem to be good now
You wouldn't by any chance be interested in sharing this would you?
i can, if I find it 🙂
@serene relic
looks like it's back
saving these for reference accel_per_hz: 75
now its 100
and now 150
@raven cliff quick test done with the ADXL on the toolhead I didn't bother to get the other out of the box and set it up
it actually suggests lower accel
and lower frequencies
interesting
ressurecting thread with a small update
ignore warping, but extrusion is quite uneven, ordered integrated gear from fysetc, should arrive tomorrow.
finally building the SUAFT as well parts are ordered too
What extruder do you have?
Relisting so I don't forget
TODO 
Eventually reprint some omnibox parts as this PLA one is breaking (delaying this one until next release maybe)
Too many options on rewindera and buffers currently can't decide what to use on the TR
Print SUAFT 🕑 (parts ordered)
Model new sherpa body with fixed idler, filament sensor and ECAS for the new gears
the fysetc ones are quite good, I really like mine
At least with ercf, passive rewinders are not recommended due to being too jerky, I think there are various active rewinders in development
my initial plan was to try reprewinder and an ERCP (modded for ZeroG logo)
and decide from those
https://github.com/SkiBikePrint/ERCF_Mods/tree/main/Filamentalist
this one is interesting
Yeah that's the rewinder design I had been looking into which seemed to not be recommended
Oooh that's interesting
not understanding the jerky part really
I looked it up, the issue is potentially causing the encoder to mess up
well not an issue on TR,first no encoder
and I don't really see how the ratchet of the rewinder can put any feedback on the encoder if the encoder lives after the gears
Probably by causing slippage, particularly when not running the gear motor in sync with the extruder
But huh, wasn't aware the tradrack doesn't use an encoder, does it operate "blindly" wrt filament lengths to feed?
TR has a sensor after the gears, and can operate blind from there forward.
or you can add a sensor either before or after extruder
I like it before as its better to detect if an badly shaped tip got stuck in the extruder when unloading or if the filament reached the extruder when loading
Ah I see, seems a lot less overcomplicated than ercf lol, I guess that explains the inspiration for the belay
belay is to help keep TR and extruder in sync
a more brute and cheaper approach compared to encoder
Yes, what I mean is that since TR doesn't have an active feedback mechanism on lengths, it needs the belay to compensate, while ercf didn't seem to need it, with the software only recently having added support for it
I'll definitely have to build a TR at some point
Oh I see
an alternative is to every layer change raise and lower TR servo for example
while not even disabling sync
also works, but I found I don't need that
I see, I have the belay configured, but haven't really noticed any difference either, I see the rate adjustments in the log but haven't really looked at them properly to see if the adjustments are meaningful
new gears arrived, want to make a print before and one after, same gcode so not installing yet
testing before new gears
nothing which is weird
here they are
ignore corners that's too fast to properly cool
aaand I'm running the test the wrong way
@formal marsh are those pictures all from the same print?
no, every set of pictures is a different one
so what made the inconsistent layer lines to come back?
I'm also struggling with them but I'm also still running stock Z
from first to second, a little faster, thinner layera
got another on the bed to check that I printed yesterday late night
now here can be see properly, next step swap gears and reprint this one
Interested in the result with a gear change!
so am I, will disassemble sherpa in a moment
new gears finnally in, printing
Fystec gears right?
yep posting photos in a sec
fysetc integrated gears, not perfect but
and side by side
not perfect but better indeed
same exact gcode, same filament
does make me wonder how one might get it even closer to flawless
I'll take a photo of the same side of the piece side by side, some of the remaining artifacts seem to be the same on both parts which might imply the source is something else
yeah I had a feeling it might end up involving scraping the bottom of the barrel to get everything else near flawless
I'm willing to blame leadscrews next if there are artifacts that match on both, even more on a second print with the new gears
next step (and easiest) is to attempt changing to fixed gear sherpa, I'll use the opportunity to also add an ECAS and built in sensor, this one adds too much height even though I like it separate
I have an extruder design that is slightly better
But also wears down the gears much faster
Take a look in here
https://github.com/LynxCrew/LynxSherpa
But I really would not recommend using it
LynxCrew Sherpa Mini/Zero fork. Contribute to LynxCrew/LynxSherpa development by creating an account on GitHub.
ooh interesting
so it 'fixes' the idler gear in place? (I'm not sure what you mean by angle adjustment)
that one due to the gear wear I'm not fond of trying it
maybe nextruder like gears 🤔
orbiter is closer I believe
or that one extruder that uses belts
papilio
Oh right, yeah, papilio, I think it had some experimental fully printed version, might give that a try for fun
Ah nvm, wasn't fully printed, what I was thinking of was a failed experiment to have some of its gears printable (from their discord)
I considered it but a few things stopped me, one of them being the gears prices
it has some printed gears but the ones that drive the filamennt are LGX
or toothed idlers or something
Yeah, bit more of an investment for just an experiment
It's also probably much heavier
also no sensor
Orbiter fixes the backlash between stepper and drive gear
Not backlash between both drive gears
Imo Galileo 2 with an integrated drive gear would be almost perfect
Planetary gearbox to reduce/remove any backlash between stepper and drive gear
Big Single drive gear to remove backlash between the dual drive gears
And integrated drive gear to remove offset
Nextruder is pretty good
Having 2 large gears in series removes a lot of play
hmm shaky photo
If the adjustment screw is adjusted well you remove basically any backlash between both drive gears
Though that causes excessive wear
That's why I recommend not using it even though it's my mod/design(I improved upon the Sherpa zero)
Ahh
The fysetc helical gears are pretty close to perfect imo
Low weight and low wood grain
there is also another reason for wood grain which even the best extruder couldnt remove:
inconsistent filament diameter
good point
so imo getting the fysetc gears is good (since the price is also really fair imo)
Personally I wouldnt get a genuine ridga again
way more expensive while actually being slightly worse than the fysetc ones (well with the LynxSherpa they were a tad better, like 0.1% but I guess they will wear down in one year and I'm not gonna buy a 40€ gear set every year for that little of an improvement)
for under 17€ the gears are damn great
I'm happy with the improvement, I want to try the fixed idler next because I also want to integrate sensor and ECAS into it
someone at Annex made one but uses switch with lever and I want to try without
especially since they should have more wear resistance (a mate of mine used their standalone ones in his vyper and he prints a lot of abrasives, he said the nano coated fysetc gears lasted way longer than the original vyper gears)
there is also a body for the normal sherpa for a collet clip
but the fixed idler doesn’t have the idler thumb screw (lighter) and it actually worked great for me. it will also remove some issues using the sherpa with manis style fan ducts
Looking at the fysetc gears, they don't have the "curve" on the filament grooves, like these have
https://a.aliexpress.com/_mLA3iK2 unsure if they're meaningful though
curve on the filament gears is actually not that good imo
Oh?
it twists the filament, causing other issues
Ah I see
Ordered
this discord os terrible for our wallets
Ikr lol, recently discovered the ender ng from here and now I have to convince myself to not immediately jump on building it
Better than another I am on, they do group buys and not just 3d printer stuff.
don't say anything about new printers, I won a set of panels for a RR V-Core... I don't have a VCore...
Haha yeah, you mentioned that dilemma
Yet, gotta use those panels you won!
yup....
it would be easy to wait if there wasn't any discount on the base kit
self sourcing it is actually more expensive 🤷
but the discount is roughly 150e
soooo unklicky started misbehaving
oh? in what way
probe accuracy results: maximum 16.050000, minimum 15.992500, range 0.057500, average 16.023000, median 16.022500, standard deviation 0.017529
usually there's another 0 to the right of the dot
is this the bfp version?
yup
huh, wouldn't expect that design's pin to loosen up
it's like 10 good samples stupid consistent, then one wayyy off
let me see if I can get a snippet of the log
maybe something else came loose?
oof
yeah that feels like the pin wiggle I had in my non-bfp unklicky
Do one as a bfp-hs
it IS BFP-HS 😭
Seems like it’s time for unklicky-00?
ehhh too bulky
probe accuracy results: maximum 15.800000, minimum 15.727500, range 0.072500, average 15.771800, median 15.772500, standard deviation 0.015887
Disassembled and reassembled same shit, wrap up this filament calibration and print a new one
I’m going to start stuffing with load cells
probe accuracy results: maximum 0.180000, minimum 0.037500, range 0.142500, average 0.093700, median 0.061250, standard deviation 0.054576 now the unbolt WTF
I'm so close to drop the money on the bacon
new test unbolt now good again, unklicky messed up, and I have no idea where I put the switch of the original klicky
from my experience this happens when the magnets are not seated properly and the probe can wiggle
have checked them, not that I can see
meanwhile found old d2f klicky will test it, can help print new ones
d2f one works good enough, I'll print new unklicky
I'm actually surprised this went so good, put the old klicky, heat up, adjust switch offset for this probe, send print and forgot
well I'm dumb
new unklicky:
probe accuracy results: maximum 15.510000, minimum 15.500000, range 0.010000, average 15.506620, median 15.507500, standard deviation 0.001605
the problem with the old one? has an M3x12 which is too short and was wiggling when probing... new has a longer one with 16mm a little too long but works damn good
old unklicky with longer screw:
probe accuracy results: maximum 15.985000, minimum 15.975000, range 0.010000, average 15.980050, median 15.980000, standard deviation 0.001968
you technically need m3x13
so basically get a m3x16 and cut it
yeahhh that's what I get for not reading the spec and thinking the 12 would work, 13 will still be short due to the amount of wire I use, I just hacked the back to have a through hole
will re-design later the STEP as I don't mind the little tip of the screw in the back and it actually secures better
for UnklickyNG BFP the doc says m3x12, which is actually too short
I don’t know why they didn’t put an hsi for that screw, it’s the only thing that I don’t like, bolting on a thread on plastic always gives me goosebumps
maybe a hex nut or something. but I understand it’s in theory a set and forget screw that one… If something is off, you print another one anyway
ehhh sort of, today building one of them I had to pull the screw out 2 times, needless to say it's deviation is a little higher
now on to print the suaft
@pale cloak finally
I wonder what happened there
ohhh warp 😅
well unless its not functional not printint another its a waste of filament
woop woop!
should be good despite quality
well SUAFT is printed, someone forgot how to mount it
hot glue, duct tape, and cable ties
engineers best friends
lol im tryna scroll up
yeah it's loooong
been tempted to pull my current bed heater out, drill new holes on the bed and put a new heater from Fysetc
Their heater goes close enough to the edge. No need to run heaters to the very edge
reason being my bed has the front holes in the wrong place and still have some Z atrifacts that may be caused by my adapters, drilling as is will destroy the heater so need another
yeah thats my reasoning as well
oooh I forgot you're using a voron plate
exactly
fysetc heater would have an advantage too, 600W vs my current 300
Oh yeah, I like my fystec heater. It gets to 110 quickly!
Flow - Born from an E5Pro
not much updates here but here are some plans:
New bed to replace the voron one with adapters, no specific reason other than being tired of looking at the adapters and wanting a more powerful heater, Fysetc one is here along with hard magnet, bed will be done by @quartz tide just waiting on it (no rush)
new sensor on sherpa being designed (thanks @deft geode for the test prints)
designing buffers to attach to spool rollers (roller is @pale cloak design from his drybox) buffer designing my own to attach to it
to be done find a way to make a cutter for E34M1 (I don't like any current design)
buffer WIP
damn. that's a lot of material. buffers seem to require more space than one expects.
quite a bit, can be minimised but I like having the filament more constrained, makes for a smaller possibility of tangling. those horizontal arms might be able to do without them
still need some fillets and ECAS mounting to test print
discarded base hexagons, they don't reduce material at 20% infill (what I'm planning) make print slower and have weak spots
looks good!
one final check and I'll start printing it
also took pictures of the heater and magnet, this will look damn good
I just ordered 2 more of those magnets
After my printer ate one this weekend.... and I had to steal the one for my new build to replace it
So I will have one spare
I'm hoping I dont engrave any 😅
RIP
I saw it on #demolition_derby hurts even watching
Looks very good 
first buffer
seems to handle 3 loops fine, each loop roughly 60cm of filament, so I could get away with making it slightly smaller as I need wither 70cm or 90cm (from TR load and unload values which are not very correct)
next up a cutter
no but takes some space
taking notes on buffer:
- base can be 5mm instead of 7, just add some material on heat insert area where screws into the roller
- add cutout to pull ecas out more easily
- measure screws, attempt to standardize one size
- maybe make side "arms" a little smaller, sometimes the filament instead of making nice loops extends up and into the back, making it shorter might help.
new version 2.5mm smaller base (because shorter arms)
new sherpa sensor in
still need to organize that wiring and shorten the stepper one
Looking nice
cutter concept, not sure if this will work
I think that I have seen some eva3 cutter somewhere.. if I find it I will share
there are some on printables, I don't like them
wellll shit, after re-soldering the hotend fan sent a print... without testing the fan
CLOGGED RAPIDO!!
accepting tips before I even start
Eish... Nozzle out and with a "poke" with the size of the filament and long enough push the clog out
Something similar
turns out it was mostly heatbreak
rapido heatbreak is delicate, ask me how I know
if you heat it up, make sure the top didnt curl inwards, if it did, take the small allen key that came with it and use it to uncurl it
looked good, still need to print again to check it
just wanted to let you know
cause I didnt notice it curling inwards and that was another nasty clog, let me tell you XD
I didn't know that.... Maybe I would give another chance to rapido 😅
how's that going, btw?
good, still need a cutter, went through printables but I'm still not happy with any, not happy with my design either so there's that
turns out I need more buffer than I initially thought, and with this design filament might tangle a little, need to run some loads/unloads
I didn't want to but may need to further enclose the filament, want to test 3 loops too (which fit the buffer fine)
new bed arriving most likely today
it is here, damn it looks nice
@quartz tide thank you
gladly 🙂 anytime again 😊
mesh on the new bed, looks good to me, pretty sure my frame could get some further adjustment also this is still the bare plate with the protective plastic
Retries: 0/5 Probed points range: 0.011250 tolerance: 0.050000
this z tilt...
that’s pretty similar to my mesh (fysetc 255), [0.14-0.08]. I guess the protective plastic will give you some unevenness, so yours could be even better
still want to take it out and mesh
need to find my IPA to give it a good clean as well
also, didn’t you get the zeroG hard magnet 🤤?
Added some RTV today, probably didn't need it, but screw it better safe than sorry
new bed still not in, got an order and have been printing it, next need to tune the red ABS for the RedFox build
sooo not yet
meanwhile new sherpa had adjustments since filament sometimes wouldn't load, triggers sensor but then doesn't go down, need to find the cause
still not perfect, wish it was completely transparent
small update, swapped for tr8x4 (stolen from the rat build) one of the nuts was wiggling need new nuts for the rat
meh something not good
roughly 4mm spaced, matches leadscrews so need to find if its the nuts binding or something misaligned
Also make sure the couplers are tight on the motors and leadscrews
I had 2 slipping on my new build at first
I'll double check, but I find it odd matching the rotation distance, betting on something misaligned
aaand it's there 4mm
maybe the back hydra extrusion isnt squared
unlikely but I'll be checking everything at this point
replace every part I can, starting with the bed
if doesn't fix it swap couplers maybe try the 8x8 again to compare using the same filament
double check rail alignment and finally reprint hydra
I'll miss my PCCF parts but
if you loosen the motor couplers, do all 3 lead screws want to naturally want to line up with the motor shafts?
It kinda looks like this https://youtu.be/qNCDXdjJb90?si=YCrKkMRgwn4po5dG
Wobbly prints and prints with lines... How you fix it? It can be many things (trust me, every time is a different reason). That's why I loved the Book I showed in this video. It will help you a lot, specially if you are just beginning with 3D Printing:
"3D Printing Failures" Book
by Sean Aranda
https://geni.us/73Gd (Amazon)
More 3D Printing Bo...
yes but its consistent with the rotation distance
front ones yes, back one is a little off
well, first finding, back stepper screws were too long, 10mm instead of 8
might be the cause, might not
rat oldhams won't fit
back to 8x8
not sure if the filament is not properly dried or if it's due to having a little glitter
still some binding but screws were not even lubricated and they're probably still dirty
finally
sexbolt had to come off, the ground screw hits it need to see how to fix that
Retries: 2/5 Probed points range: 0.028750 tolerance: 0.050000
mesh on the magnet
tilt on build plate Retries: 2/5 Probed points range: 0.038750 tolerance: 0.050000
similar to previous bed
full bed mesh with sheet, I'll take it
frame might still be a little off
damn now this heats up fast
love it
first time I didn’t believe it was already at 90C after not even a minute, I had to touch it 😅
yeah 1 min 30 sec to go from RT to 100C
hmm looks very similar to the red one, this was printed on the new bed
filament is dried
https://www.printables.com/model/82829-oldham-couplers-for-ender-5-pro-plus-z-axis
printed one of these, added it to the leadscrew that looks worse, seems to print fine will print 2 more at least will help figure out if it helps
dumping random thoughts, cutter used works great but shaper suffers, I'm also using the thr adxl need to move my ass and configure the PIS.
my buffers suck, that's the truth, good for short lengths for my use case not really
think the embedded design on EVA is too much work and not even sure if it will be reliable
also z offset is getting inconsistent
also a swap
yeah I was using flexible couplers between the steppers and lead screws, and realized the same issue, removed them because they were likely causing tilt correction to be imprecise back when using klicky, maybe they aren't as much of a problem with beacon, haven't tried
not sure but since I'm still with klicky
ah, yeah in that case I'd probably recommend against oldhams
since if you try to do tilt correction based on all 4 corners, the rear two corners might be able to cause the bed to tilt slightly when pressing down
@late veldt no issues yet but only have a printed one on one corner, need something better than the current one, might try to print something else just to check if they help with the remaining wobble
I think it looks ok, the peak at106 is the real one, you just have something else going on. in general the density is lower. I think once you remove whatever is going on in the x (I think the cutter might be the biggest contributor there) the 106 Hz peak will stand out
oooor make a new body for this one
it’s that a thing?
not that I'm aware but it's an idea, I can integrate the screw to attach on the enclosure
it's one of the issues, also add some chamfers to help filament go the right place
it’s difficult to get filament is also in mine, so it’s probably a design flaw
But I have issues mostly with pla, that tends to snaps instead of entering the SFS, I resolved by stopping using pla 😅
Petg for life!!!
meh used to be my first choice, getting tired of it, sticks too much on the nozzle
what do you print then?
pla abs and asa
petg is nice if you can’t enclose and need functional parts. it’s decent for hardware tools and maybe even printers parts, if not enclosed
I also find it easier to print than pla
yeah, like I said I used to think that as well
what did u do to ake it easier?
which?
pla
for me, pla became easier to print when I started using the tradrack motor to assist the extrusion, 99% of my issues with pla were filament path related. with 2 motors never had that problem again 😅
I have to crank up my cpap to 67% and print super slow
I feel like pla is converting this printer to a stock E5
WTF I print faster than that on a dual 5015
at what temperature are you printing
190
oh, I print pla at 220-240, even if the label would say 190 😅
higher temps were even worse
after testing of course
Some pla or quality pla should be printed a little hotter.
some PLAs I print go at 235, and faaaast
A lot of blends are designed for lower temps
are you enclosed?
how do you think I print plastic braces that can handle being tightened by a screw to death without breaking
this is a cheap pla that someone bought for me to print them some things I ususaly dont have this much problems with pla
Dry it
nah
I would just toss it and move on then
I have to print these 2 models in it
why so high cpap, can you see that it doesn’t dry fast enough? I have never seen this, only when testing high speed on vaseline mode, maybe
it just didnt like going over 100mms and cpap under 60 on overhangs
thats how stong it is when I print petg
vaseline mode, I'm going to start calling that to vase mode
hahahah
vaseline mode on top
damnit autocorrect
zesty merc
keep it
yes, not I can’t correct it anymore, it needs to stay
forever be on this server marked in history
9 lanes should be enough, but I need better buffers or rewinders
filament got stuck some times
Ive seen someone use one of the big plastic boxes from ikea and put some rollers in there with holes for ptfe
And he printed deviders for it so the filament doesnt tangle
leaves spools too loose not a fan
Also some time ago I think there was a printed self revinding spool holder
But I forgot where I saw it
winchrack has motor, filamentalist uses the filament being pushed back to rewind the spool
did it warp and snap off the bed, or just didn’t stick enough?
You got an alchemy smooth plate?
If yes try that
warping on back
yes, already tried the textured but didn't stick
Textured can be tricky
Sticks better than any other I have tried but smooth is still way better from my experience
well and now z tilt doesn't work how fun
been fighting some artifacts from the leadscrews, added oldhams and fixed the top of the leadscrews not perfect but way better
enjoying the new toy?
What is your probe? beacon?
cartographer
running a first layer
aaaaand?
weird, half good half bad
but this is ABS, and I didn't heat soak shit
so need to properly test
also I calibrated cold, so might try hot
did hot calibration and fine tuned z again, looks good
now need a good way to prevent oozing
and guess who found a broken 5015 fan wire broken fixed it but forgot to test the fan
nope dead mosfet, must have shorted when the wire broke
Nice.. what are your thoughts?
Was thinking if it was an option for me to run it by can bus.. so that I could free one USB in my pi
so far has been good, I'm running through USB though, couldn't be bothered fiddling with my THR to splice CAN
but the tap functionality works as expected and for less than half the price of Beacon
which my wallet loved
tempted to get another for the rat
Uhm you convinced me 😅
wasn't even trying
I’ve asked about the IDM scanner at 3do, and they said it works with carto firmware (survey tap)… so maybe idm and carto are indeed the same thing?
That would be awesome
I will try that
Does anyone knows how to remove the Idm firmware from Klipper?
carto has the open bed scanner thing so could be based off that
what do you mean?
Forget it 😄 it's fixed
what were you even talking about? 😅
so did you use the survey touch firmware on the idm?
Changed feet today, now it's a small printed piece with a small adhesive felt pad.
Shaper before:
shaper after:
slightly better, X still a little messed up
but I blame the cutter for that one
not yet
finally setup shaketune
still looks like crap
I need to improve the toolhead, with the cutter things got very out of balance
Yeah mine looks similarly awful, but haven't had time to look into ways to improve, kind of thinking of getting the cnc toolhead, I figure its rigidity would help a lot
same thought crossed my mind, but price and having to go back to single 5015 or move to CPAP (more money)
and you’d still have the cutter to cause vibrations, probably. and arguably metal might transmits even more that kind of vibration
that cutter is going to change at some point
@tired pike @proud marsh IDM won’t get proper support on the carto discord. Richard doesn’t offer support for IDM but there are a lot of helpful people there that might help.
Nice good to know
Thanks
I have switched to Cartographer on my machine and am really liking it
The Idm doesn't have much support anywhere
yes, I ask at 3do why they provide the IDM and not carto, Thomas told me they are the same. But than I asked Richard if he knew why 3do only stocks IDM, and he told me they reached out to him, he replied but they never replied back… unfortunately the IDM seems more expansive and doesn’t have any support
Yeah probably because Idm sells in AliExpress and its more available that way..
even with the weird spikes, it seem sthe shaper will clean it quite well
finally a version I'm happy with, might not even need buffers
well shit, 75mm spools don't fit, only without the rollers
also I'm an idiot, was going to make a 80cm shelf, but the boxes (8) take a little more than 10cm each, almost 11, so need a 90cm one, at least I did remember that before I made it
not that I like it but I guess I'll respool the 3 problematic ones
ordered the wood ready, just need to cut down from 120cm to 90, I could make one but don't feel like having the extra effort
let's see if this becomes #demolition_derby material
could be worse, out of 10 6 survived, and if it wasnt for DK z hop none
and yes I tuned speed down on the tops
wood for the shelf arrived time to get the saw out and chop this down
hey... this is starting to look like a carpentry log.... 🤣
just one shelf, I promise 😆
there's logic to all the madness
I promise it's not going to become a woodworking log, here's how the boxes will sit on the shelf, might even add the buffers screwed under the shelf if there's room for it
shelf ready, just need to put it on the wall, the holes on the edges are a little too much the edge
shelf in place, need to print the remaining parts for yhe dryboxes, print 2 more pika lanes, and a support to tilt the TR
that looks clean
thank you, still a little annoyed that the buffers will be a source of moisture
thinking of adding something inside to prevent the filament from going to the sides inside the boxes so I can use them as buffer
enclose the buffer system?
ehh didn't really want to do that
not with a box
I'm thinking of making the sides acrylic
There is that one buffer design that uses acrylic panels
I want to reuse the pika
not too bad regarding looks
Is that a custom CNC toolhead?
yes, which utilities only alu plates and standoffs. I really like the concept
Yeah that's pretty neat, plus, it would feel very appropriate with the full metal gantry
Stylistically
it was one of the inspirations
it's a mix of VZ, RR, E34M1, and inspired by the full metal gantry
it supports any hotend, extruder, etc combo as long as it's a Rapido UHF, sherpa micro, and dual 5015s 😅
lol
cutter and sensor are optional
are you using the full metal gantry? I missed it and only realized it existed today 😅
got the parts not installed yet
ah
might install the toolhead first then gantry, slowly changing to see if I see improvements or not
cutter really messed up my IS
I'm holding off to do a full enclosure + full metal gantry + either this toolhead or the official one upgrade sometime next year
this? you mean mine?
yeah, assuming you'd be open to sharing it eventually?
I'll release it sometime but would like to test it first
it's my plan to release it yes
yeah, no rush, my wallet would be happy to wait 
this whole toolhead adventure only started because I'm designing the smaller merc
ah I see
And now precurssor comes to tell you its slow lol
nah
it's as fast as I need it to be right now
estimating 50g worth of printed parts not considering the sherpa
well crap now I'm thinking how to make it as light as possible and that includes aluminium heat inserts and Ti screws 💸
RIP wallet alu HSI and Ti screws are on the way
also finally built the roadrunner sensor
but it's binding
I'm suspecting the gears are to blame
hey @plucky plume found you mentioning you were building the roadrunner sensor, any feedback on it
tightened up the lever a little, better movement but a little resistance
need to test the TR pushing in
I still need to solder it together 😅
heh mine has a little resistance on the gears
How's it working?
still need to make it tune properly before I test it
can't really calibrate it properly
Failed to find well-fitted linear equation, the measured fitness was 0.9868171626589851 (min. 0.999). [1/3]
the BMG gears are not rolling smoothly, no wonder I had quality issues printing with it
and I can't really figure out the cause
looking for a 0.999 fit seems really tight, maybe there's an option to loosen that requirement?
there is, managed to get it calibrated but
sometimes it will say it did not detect movement
retrying now
and the default calibration speed is too slow to get anything
might get good results now, next step is check if the TR can properly load it otherwise it's useless
Most frequent reading: 21.56 with 8 measurement(s), which implies a 'correct' rotation distance of 28.4375000.
let's hope
detectable change of 0.060mm this is nice IF IT WORKS
also re-calibrating TR e-steps
in a 50mm extrusion it was doing 47 😬
ohh wait belay
crap
well TR managed to load the filament through, printing today to test it
after managing to tune it, been printing fine with it
still need to test if it detects properly but TR has no issues loading
finally tried sensorless on X seems to work
for a first printed test looks good
next is smolzg cad cleanup
@late veldt here it is, I abandoned the metal cutter for now
What toolhead board is that? Just curious.
Oh I think I found it.
THR36
Can you run a Beacon through it?
Gotcha. Thanks!
re-tightened the belts, and put a rubber band on the current cutter arm, things look cleaner
Y is weird no rubber band and now it's better
Oooh nice!
is that a built in cutter?
somewhat yes
toolhead works without it as well
cutter is just 2 pieces under the extruder, there is also a sensor under the cutter
and printing a PoC of arm
oooh interesting
so 2 problems
this section is too thin, and the plan of printing a .2mm wall as support doesnt go well either, need to split in 2 to print or put the blade on a side like the one I have now
new plan
the arm itself at 10x4mm seems to be strong enough, the weak spot now becomes thicker and the tip of the blade is supported by the pin in the pivot point
I'm just unsure if that tip is enough or if it breaks the blade
or I can just move the blade in a little more
if this fails next step is a proof of concept with an old hacksaw blade, grind whatever teeth remain cut to size, attempt to drill holes including some for a sort of handle, and it becomes 2 alu plates as well
idea for the printed cutter, make 1x4mm pins, piano wire or even paper clip should work, use 2 or 3 around the blade as needed and make a printed sandwich, 2mm pieces with a .2 or .3 recess for the blade on each side
if the pins fit tight enough they should help keep the arm together along with the pivot point and the screw on the blade hole
finally printing parts for the toolhead
and for metal gantry just need to finish sanding the plates and paint
on the XY joiners I'm saving 23g by swapping all steel screws with Ti, and the pins by CF, I'll first install everything with steel give it a test run and then replace to compare
and because I don't have all the Ti screws
just to keep the shots with the build
morr Ti screws on the way, last parts printing
had forgotten how might the LW Gantry was
also 1.43g less with Ti screws not much 😅 but I do feel fancy
plans for tolhead plates
plans for v2 IF, v1 works
hmm I should double check belt position
sooo metal gantry is in
now align everything
wondering if I should run E34M1 for a test run
maybe test fit belt positioning on the new toolhead first, then E34M1 back in until new toolhead plates are done (alu ones)
quick test fit I'm actually surprised things fit so well, bigger belts are on the way
going back to E34M1 for now
might cleanup the cad and share soon
hindsight, the sensor needs to be wbove the cutter
otherwise when I cut the filament stays on the sensor
E34M1 is back, the cutter arm on mine can also be shorter
shaper is still crap but I was expecting it, I still think most issues come from the toolhead (cutter etc) so until new belts are here I'm not going to dig into it
but I do need to redesign to have the sensor above the cutter
sweet jesus
come to think of it, the feet I'm using now might not be helping either need to decouple the vibrations from the box too
just so it stays here
also big shout-out to @proud marsh for putting up with my endless rants, sharing and comments, and his feedback as well
and we have sensor and cutter
wrenches because I'm lazy to put screws
one handed is hard the other was holding the phone
cutter measures 8mm tall, sensor adds another 8 or 9
no choice toolhead doesn't sound that good anymore
technically you can choose if you want cutter and sensor
and also made this
weight breakdown
56.00 - printed parts
25.68 - alu plates
24.92 - Ti Screws and Alu HSI
59.47 - dual 5015 blower
07.36 - 2510 fam
15.76 - toolboard
45.28 - hotend (rapido UHF)
20.31 - sherpa micro gears
70.41 - sherpa stepper
Total: 325.19g
5015s to blame now
Cpap
I know....
in time
meanwhile, steel and brass comparison wirh Ti and Al inserts
19.57g lighter, not much but
need to check the weight of the E34M1 I'm running
wow, the carbon dowels are so light, I’d say 1 pin might be 5 gr
no carbon there yet
that's toolhead
aaah makes sense
gantry is next, it's currently screwed in with steel the M5s are already Ti but I'll compare both BOMs as well including dowels
last Ti FHCS have arrived
waiting on a saw to cut the plates
and belts
the wait is killing me
Carbon rods are flexible though
Not too sure whether that's a good jdea
I'll find out
I believe @coral basin has been running CF pins
Yeah iv been running them for well over a year now and they have been great. I checked them a few weeks ago when I switched to the newest printed x joints and they are still fine. No signs of wear or anything.
Have not tried them on the metal gantry yet.
I need to find where I stuck the other part that I was cutting off and was going to test the metal gantry with them as well
from 9 to 1.9
Yup, about 7g per side saved.
Tbh I dunno how much the Cpap will take off the tool head. You remove the 5015 then add that hose on top the flail around lol. Iv always wondered how much are you actually gaining that way. I guess it makes up for it with ultra cooling.
it's a good point but I still think it takes weight off
the hose weights barely anything
also it removes vibrations
you also have the tube that helps as strain relief and removes a lot of the flapping around of the various cables
and you gain something like 6~8x the static pressure 😅
the weight at this point is more of a stupid goal
E34M1 175+132.16+45.28(rapido)+59.47(dual 5015s)
total 411.91
86g lighter
already an improvement
XY jointer HW, steel vs Ti+CF common stuff not weighted
34.19g - steel
13.47g - Ti+CF
so 20.72g lighter
had the urge to put it together even though the toothed idlers will be changed
need to redesign that endstop piece for the toolhead
guess that's next while I wait for the belts idlers and remaining tools to attempt cutting my own plates
might make 2 versions, one solid and one with the weight reducing holes
need to write up the BOM
first drill attempt, I rushed what can't be rushed, didn't properly punch the hole centers which was the first mistake
second is the lack of the correct drill bits for some holes, m3 requires 3.4 for clearance (might do 3.5) M2.5 is 2.9
and need a 4.2 hole for ptfe
tried to adjust and thread the holes
rapido only 2 fit s
the others are too misaligned but not too bad, confident I can make a proper one tomorrow
Do you drill pilot holes first as well?
I did not, something I might do on the next one today
drilled bottom plate now, everything at 2.5 then 3 with small hand drilled 1mm starts just to avoid things moving
taps need to be drilled at 2 and hand drilled to 2.5 the press wiggles a little threads end up a little loose
other than that looks good, now need to cut the damn curve
don't
oo jeez I realise I hate gifs, I’m going to delete it
since the saw I got kinda sucks I think I'll run one some holes
and file/sand it
or get better blades
screw it, both
I can’t believe I didn’t know anyone near by that can do a little milling for cheap, I’m looking at 80~90€ to cut my plates, to have 2 or 3 sets, because cutting only one costs more, and than maybe there is some stuff to modify
well, not bad at all to be honest!!
ehh it's terrible but if it works it works
hopefully next one comes out better
might attempt to cut another one
That looks great for doing it by hand.
no way this will work 😭
This is why I might 3d print parts to be laser cut for a test assembly, but then I just order the parts since the services are fairly cheap.
yep printed test worked out fine, but it's more of a pride thing at this point
if I factor in time as a cost it's already more expensive than ordering from somewhere
@formal marsh when I have fabricated mounts for things that needed holes to line up, like alternator brackets and stuff. I would start by drilling a single hole, bolting the two plates together, and drilling all the other holes at the same time through both plates at once. After that you can start to shape the part. At least then, the important parts line up good.
yeah that will be my next approach
but still in this case the rapido mounting holes are not aligned
@formal marsh it's definitely doable, and if you are determined to do it by hand with some more tinkering I bet you'll get it right.
determined is not the best word 😛 more obsessed
so, which is the correct plate that aligns with the 3d printed part? or are both somewhat off?
both somewhat off, being the top one the worst
I wonder if you could attach a scalpel or just a pen or something like that to the toolhead, to sign the shape of the plate directly on the alu part to cut… I mean, hypothetically
because it’s clear that you really need super human precision to even glue the paper on the part, let alone punching holes or cut
paper on part is easy, next plan is something I actually tested on this last plate, I cut around the lightweight holes and after removing the paper it's all there, so I can cross mark the holes as well, then the issue is punching them, tempted to test a brad point bit to punch them even if that means ruining it for this, heck might even drill a pilot hole with a small one
at this point I'm willing to try anything it's a learning experience
last resort well just order them laser cut
scribed everything on the plate, brad point drills smallest I got is 3mm so no good for pilot holes, but the tips are very good to manually poke in the right spot, then properly punch them, measuring with calipers things look good enough
next up actually drill some pilot holes and measure AGAIN
WIP, that damn bowden one ran away but I think I can save it, 4 remaining but looking good so far
let's hope
others look good but that bowden one is a mess, still deviated, luckily it's not too critical and I can widen it if I need
so, rapido attaches without the front screw that one is too off 😭
the worst problem being that my M3 tap broke doing the last tap on the bottom piece, meaning the top one is still untapped
standoffs were drilled together so they're looking good but didn't assemble shit
guess who's an idiot did a test tap with the broken remnants and thought, this will work
proceeded to do 4 more and none is good the first one was pure luck
and now I have an otherwise good plate fucked
@formal marsh it's a learning process! You'll get it! Each time you try, and each learning experience you have will get you closer to.your outcome! You get better every attempt.
new taps are here, I'll cut the round on the bottom one test fit the whole build, if it fits fine, drill another top one
new thing for me, 2 taps for M3 apparently a rough one and a finishing one
I've said it before, but I prefer spiral flute taps for hand tapping aluminum. They are typically used in cnc machines
They create a long chip that gets pushed up like a drill bit chip
I know and the one I broke was a spiral flute, couldn't find an M3 alone just full cheap sets, I have done taps with this kind years ago, takes more patience but it still works so I'll give these a go this time
Finally pre-assembled the whole thing
2 of the taps are completely messed up, one of the rapido mounting screws is off and the bowden is misaligned, other than that everything seems to fit, next up I'll put it on the rail to test the belting properly
if the belts turn out good, another top plate will be drilled
Looks good man!
i'm so far behind 😭
this looks great. i kept thinking the cutter was huge, but i realized the rest of the toolhead is just smol
it is indeed smol
wiring still needs to be adjusted, Ti screws are not in, and the plates don't have the cutouts to reduce weight but as a test this will work
didn't check belts yet but damn X looks clean compared to previous stuff
Y looks like shit
I think its your cables on top of your toolhead
I'm suspecting the same need to clean those up now
but damn it's looking fine and I need it printing like yesterday
small update
this is Y on carto adxl
this is with the THR
x barely had difference
theory is that carto mount allows for some wiggle
meanwhile Z is a little inconsistent because I have oldhams but removed the top bearings, I'll raise the bed with higher spacers on the arms and later redesign for smaller support bearings, also one minitank is dying time to get the PCCF spares back in
other than that printing great, so after I dispatch the giant orders I got I'll make the final cuts on the plates to make them lighter
and see what improves
oh and the 2510 I got is stupid weak hotend is also passively cooling on the alu plate
wait.... did I lubricate the rails
ok XY were lubed, since I was swapping the minitanks decided to take out the tank rails
dryer than a desert
relubed and they're in now need to cut 3 M4s down to 25mm
back to printing, no issues so far
@coral basin thanks for the support man, I got it printing
still need to make the plates lighter to compare
I think I'm ready to prepare an initial release
Good job dude!!!! Happy for ya man!
not done yet need to cut the holes but the parts fit and the thing prints
What version carto do you have? If it’s got the lis2d instead of adxl you will get different results.
ADXL
had to take out the rapido yesterday due to a clog on the heatbreak, turns out not so hard with this toolhead, sure need to remove the extruder and fan but no big deal
hmm the printer in the older posts looks familiar to me
i wonder why lol
i should start my own build log here
@white crane why would that be 😅
@sick bridge somewhere here I think
completely offtopic but is there no way to scroll immediatly to the top in discord mobile? Forum threads are impossible to read without 2h of scrolling
nope
@sick bridge toolhead starts here #1076124931517055026 message
it will have a post as soon as it gets released
waiting on @ornate flame to finish the new plates (no pressure)
same with you @formal marsh waiting on everything (no pressure lol)
did what you told me and created a build log #1362724839638696006
slowly coming back to this, realised a few days ago I could just make an entire lid for my dryboxes, needs more filament but could fix space issues
https://a.aliexpress.com/_Ext6ns8
first I'll test a roller concept with MR63 bearings
