#Glizzy got an other industrial printer
1 messages · Page 1 of 1 (latest)
take more pics of the motion system 👀
Yeah, gets to 100C chamber and 200C bed
I'll try to take some when I get home in like 2h
Just got home exactly 2h later
mmm ok
and this shit gets them 100c huh
interesting choice with the bltouch
and that wiring
also, is that reverse bowden capricorn?
thanks btw
Yes
Yeah I am not sure if everything is original on there
Might have to reverse engineer things to put them back on metal parts
Also emailed them because there are free parts I can get apparently, hope they will give them to me even if I am not the first owner
Hmmmmmmmmmm
Yep yep yep yep yep
That would indeed be the issue right there
@crimson sluice @strange rose lmao wtf is happening there 
I mean my finger clearly closes the contact but wow didn't think it would be enough to turn that lil light on
I feel more adequate with the DupliHeater now🙏
Oh it's the 5V converter that is shot
And 12V too I guess
Aight so full electronics bay rework it is then
And then I hope nothing else pops up in the process
Oh I have a 24V meanwell
Amazing, didn't think I did
that doesn't look original
It was a bltouch originally
i know for a fact older version 22idex use bltouch
120V plug fits perfectly
surprising amount of printed shit on that printer
Yeah I am not sure if everything is original
although i guess it makes sense for such low volumes
that ebox definately didnt arrive like that
There have been at least 3 batches of the v2
Most of it did
damn. and it costs 15k? 😭
my stock ender 3 was better cable managed
huh
Oh idk about cable management
Mod #1: replacing some phillips with hex, this probably will give an 80% better print quality already
The bed is only 500W at 240V...
Wtf
And the chamber heater only 150w?
Huh?
Well ima need a new heater that's for sure
Probably a new bed entirely
Hmmmm
Wonder if it is not a v1
Adding it all up it doesn't make 2200W
Doesn't really matter, anyways, should head to bed
that's the same w/m^2 as an ender 3 lmao
Well, it's MOSFETs, there's no current requirement other than forced by the gate charge / discharge resistors
@median idol I think you will also enjoy this 
Yeah I saw it, custom made?
I'm a fan of low power stuff, I am curious to see the heating times. Is it working?
Nope, 22idex
But I was told it is a v2 but it actually is a V1
Vision miner emailed me back, hopefully they will send me free stuff to upgrade things 
Mmm
I can only ask how much was it?
Im not a 22idex fan, quite the opposite
but I am happy for you if you got it for a good price
1340$ CAD
Do I think it is the best printer ever? No, I am not delusional, but I do like markforged kinematic and i think it has good potential for upgrade
What exactly is it you do not like about it?
At that price I like it very much
It's not that I don't like it as a machine on its own, but I really dislike as a machine sold at the price they put
ah ok yeah
Does it make sense? I feel like I did not explain myself the best
you just think it is massively overpriced basically
Yeah
To say the least
Because when I put it next to a real industrial machine its embarassing
At the same price
How much is it retail, 12k?
Lol 15k
They need to cut the price 4-5 times lower as far as im concerned
Idk what your version is different in, if it has the same specs but just older/slower then you got a super deal, I would have bought it too in an istant
it is on BMG and mosquito instead of LGX and copperhead, printed motion parts instead of the metal ones in the newer versions and duet 2 instead of duet 3 and maybe custom boards for the V4, it also has a bltouch instead of the klicky style on the new ones
the bmg and mosq sounds like an upgrade
printed parts you can redo anyway and the duet 2 is still alright
sounds like a good deal
I'm curious to know if there is any particular things in that ht bed compared to a regular one?
And how is the chamber heater?
I prefer LGX personally tbh
mosq is an upgrade tho
not really no, but the bed connects directly to the kinematic balls instead of having a frame for it
and the chamber heater is just a PTC
I am going to need to get a new one because this one will be too weak, maybe I will get 2
I think I understood but im not certain 😂
And a alu plate?
or Z stages
yeah
it is a bit thin, probably will get a new one at some point
Damn 200° and a thin alu bed doesnt sound very good
Im not sure about the various alloys but I feel 200° is too much
Creep territory right
That's my preferred arrangement 
well yeah it is the better one
I want to see how long it takes to reach the max
I think @stone vault the other day was saying something about the difference from heating to 100 or 200 is not much or something like that.. I can't recall exactly
Like I am probably overthinking it but does that mean I can heat any bed to 200 as I can heat it to 100 with the same configuration? Ignoring the mechanical build
Tldr: what the fuck is the difference between your 200 bed and a "standard" 100 bed
Probs not much, considering a normal alu heatblock is good for 300ºC (that doesn't need stiffness, granted)
adhesive

also thermal fuse
So in printers they limit the bed temp to what the pieces its made with can withstand basically
Or well, they should..
As one does, weakest link etc
Yeah but im not sure a bed plate, especially without a frame, at 200° won't creep
300mm also
Anyway thank you all the info will come in handy for my builds 🙏
Are you gonna test for creep or you gonna rebuild anyway?
vision miner says to heatsoak for at least 60 mins at temp
creep of what?
Bed at max temp
The alu bed
I dont think youll be able to break 100c with that
with a heated up chamber I might be able to
I want to get a convection oven heater I think
but monies
they are expensive af
I have an air fryer heater and fan maybe I will try that
What am I missing here
not enough watts/m^2
How is it sold as 200°
maybe it's meant to run on 220v
yeah it is
But from what we said before... it doesnt matter?
if it can reach 100..
Assuming you can contain the heat inside
160w is .13w/cm^2. it's gonna radiate the heat away faster than it can heat up
tbf as long as I have a chamber heater that allows the chamber to get to like 90C I don't really even need a bed heater
I'll watch closely 😄
usually adhesives for the magsheet and heating mat are lower temp
so ideally you sandwich the heater between the alu bed and a steel backer plate and use embedded smco magnets to hold the printing plate
or maybe get an e3d ht bed from somewhere
@stone vault would you put more trust in a genuine fotek 100A or a genuine crydom 50A?
How much load? Send datasheets of both
Oh well I guess it is not genuine, can't find it directly on fotek's website, or maybe it is an old one and they don't have it anymore
I guess it would be this for the fotek
And the crydom is D2450-10
does it even matter
Well I just want to put what is best on it
And I am rewiring it all rn so might as well do that rn
It's really weird, all the fans and it seems also the hotend heaters are wired for 24V but both PSU's were 12V
It's obviously a clone
We're swapping out SSR's then I guess
almost fully rewired it up, just a couple more things to do, I might be able to have it at least homing tomorrow
printing covers for the SSR's rn and I might see if I have some microswitches to add filament sensors
@strange rose do you know if any filament sensors that are better than others?
like I am looking at the duet one and that seems pretty damn expensive
but like just a microswitch with a ball bearing sounds fine idk if there are any other sensors that would be enough of an improvement to spend money on them
I don't use any
I'd rather just look at the filament left than gamble on having false positives
wtf
I'm playing pokemon with industrial printers 
@strange rose do you know if there is something that might be modifying the path of the homing file?
I keep getting this error when I try homing
and it slams the 2nd toolhead on the 1st one for like no reason
yeah that's what I don't get, it keeps trying to read it from there
not sure why
@tacit vine I also have a project thread here for the 22idex
Alright.
it be printing
making a purge bucket for it now
need to make tolerances a bit tighter
I love this
I am going to portugal for a week but once I get back I think I will start making macros for it, probably also going to try to do filament management
Yes
What was it in the end, I missed that update.
I had to redo the kinematic, it was done as if it was cartesian and it was wrong, the current config for the v2 is also cartesian but I wasn't able to get it to work properly
the legacy v2 thing you shared with me was incomplete for dual markforged. No idea what they are smoking at AddWise
the current v2 config is also using a cartesian like that, I don't get why
I mean, if you do cartesian, and then define the X,U,Y,Z axes completely, that will override it
Yeah I did override all the axes
could you share this?
It is not finished yet but I can send it in a couple hours when I get to the next airport
where you flyin?
To Philadelphia for a layover and then Portugal
oo nice
You don't have solidworks or inventor right?
nope
Aight, I'll try to remake it in fusion for you so you can modify it for your printer then
It also has a space for a silicone brush at the top
step would be perfectly fine
that's perfect
Would be a pain, I'll just remake for you and add parametric things so you can adjust preload
I have sketch and modeling tree already won't take too long
Alright, I am back in Canada so should make some progress soon™ I was just told by the rep at visionminer that the mosquitoes are indeed not stock so they don't have a duct for it so I am going to have to make my own
interesting
what will you do to line up the nozzles?
the copperheads can be adjusted but mosq cannot
probably just pray it is not too bad for duplication and mirroring and for idex prints I'll calibrate it
btw how should I be setting things up for IS on RRF? I forgot how to do it 
or I might not even have done it properly last time
same as klipper, mzv but make it so it covers both x and y
yeah but like I have to do the motion analysis
idk how I should set that up
should I be doing small or big moves?
oh, I just do very high accel, very high jerk, pretty high speed and do like a 10 mm move
ah ok
seems like the most readable output
why MZV over the other ones?
ah ok
as you see at 59 hz it covers both axes
Aren't they copperheads?
no
the guy apparently put some mosquitoes on it after buying it
thought it was standard but the addwise 22idex actually came with some e3d volcanoes
it's only an issue if you want to go "fast"
zvd starts rounding corners way before mzv
you can also do custom shaper
but mzv periods line up perfectly for me
I do 6k accel 180mm/s outer walls, top and bottom solid, 10k 300mm/s almost everything else
10k with zvd was awful for me
not super mega fast
ei is best like max 3k accel
zvd like 5-6k
but maybe your setups different, so worth a few tries
ah fuck does it home before doing the motion?
the adxl is pointed towards the bed and it would hit if it does home Z
yeah but it doesn't auto home then
yeah I set the Z axis height by hand and it is far from the bed
so just do like G92 Z100
but if it auto homes it will try to home into the adxl
ok I have no clue what I am looking at and what it should look like
like regular klipper shaper
disable the capture whole thing and enable wide band
then it should look readable
jerk of 900, accel of 25k
this is with no shaping enabled
yeah that's the one you should do
M566 X6000 Y6000
M203 X60000 Y60000
and accels at 30k for me
the initial spike is probably from the stepper starting to move, the docs mention something about ignoring the first 15hz or so
looks acceptable, the very low freq peaks just ruin it
yeah but if it is just stepper then it doesn't matter
not sure why yours look that high
but yeah give this a test
need to sit down again sometime and fix my X
first looks better
40 looks a bit better I think
any suggestions of what to print to test it out?
which one?
any
@strange rose is there not MPC now in RRF?
ah okay
dc42 told me their control algorithm is more like MPC than like trad PID
Are you using it?
yep
That's because MPC for hotend is so rudimentary, you take into account:
Constant heat loss (P term of PID)
Thermal inertia (I term of PID)
Heat loss due to parts fan (fan feed forward like in RRF)
Heat loss due to filament flow (RRF doesn't take this one into account)
The main thing that full MPC does better is lower settling time as thermal inertia is a bit more rudimentary than modelling thermal capacity
heatforward for flow I think
fan is different
it can be used for fan but mainly makes the hotend hotter when it's extruding more
M307 K is the fan parameter and calibrated during heater tool tuning
M309 S is the power scaling with extrusion rate
or feedforward
I forgot the names but it does scale based on flow
Well, both are feedforward in controls
yeah you're right
so many stuff to adjust
thankfully the fan and pid is handled very well by the autotune
the flow coefficient is a bit harder to do
The upside of MPC is that these values get physical sense like "average specific heat of filament"
bet it wouldn't be too hard to transform that into a config parameter
I wouldnt be surprised if there was a constant in the RRF source code that amounts to specific heat that could be pulled out ... nice starter project to acquaint yourself with the build environment and so on
Well I printed 2 parts in aero ASA so far, they look pretty good (especially given it is aero ASA) and I got the chamber to around 50C with the bed heater and the chamber heater, pretty impressive given that the chamber heater is only running at 188W
how'd you deal with the heating delay on the chamber?
I didn't really, I just kept resetting it until I got it to an "ok" temp and then left it at that, was able to stabilize just fine
the preview of the bed is flipped on the X axis I have no idea what is happening
the firmware on this printer is truely fucked
did it probed it in the correct direction?
ive discovered a bug with the bed mesh being weird last year or two years ago
what firmware version are you on? it should be fixed in 3.5.6 i think
3.6.2 rc-1
yea that one should be fixed there
if i remember correctly ive had that issue once too.
cant remember what i did to fix that
i think coordinates where swapped or smth
bed is fucked on there but when I try do check manually it is off by like 0.04-0.08mm maybe? so like it is good
I might have to just redo the whole config at this point idk
now it keeps going to tool 3 and there is nothing anywhere that mentions tool 3 in any macros that it could be running
usually a sign to just backup the configs and start over
yeah it is what I am going to do I think but like I really need to print some parts rn for a club
yeah their configuration for anything other than using 1 toolhead is god aweful. I am going to need to redo it all
well the club will just have to wait longer
@tacit vine do you have any macros for idex printers? I managed to find your cr-cast configuration but it doesn't have any macros
What kind of macros? It has everything it needs for mirror and copy mode.
Just general other macros like QOL stuff maybe
No, that machine was YOLO'd and manually tuned and bed plated shimmed
Also ran copperhead hotends so Z calibration was a breeze
I can confirm
only QOL macro i have is for centering toolheads on the bed for offset calibration
@crimson sluice that sounds interesting. Please share.
Well I guess I'll have to go mad and make a bunch myself
Probably not a whole bunch, most likely purge, wipe, custom pause that account for the fact there are 2 toolheads and can be using something that is not just t0, which it doesn't really do rn, also I need to modify things so when I home when I am using a tool it doesn't fuck up, rn it does for some reason
I also need to make macros for chamber filters, leds, add materials and such
you could use a lot of non-idex macros then
Yeah, I'll probably base the config off of the troodon 2 from teamgloomy like I have done for my other printers
@tacit vine do you think I should make the coordinates based off the corner of the bed or should I leave them as is with the origin at the center of the bed?
not sure if there is that much of a benefit to having them on the center
also not sure if having centralized variables that get called is better or if putting variables in the different macros and files is better
I did move from center zero to normal bed corner because everything else broke my brain. It should not make a difference function wise unless you have prints you want to print both on a delta and this machine without realizing the file. Probably not likely.
funny macro to purge things has been made, now going to be doing the wiping macro
claude is pretty handy to comment it all
not sure if its worth sharing, its just a file that checks if the printer is homed and then moves the active toolhead to the center
got another one that does the same but for the ABL probe
and now the wipe is done too, hopefully those will work well and I won't get issues, but with my luck I'll have a lot of troubleshooting ahead
I think I got a little carried away with getting help from claude AI, my print start macro is now 280 lines long 
I have material checks, temp checks, sets up idle and regular temp
Do share
it is not fully finished I think, still need to review it and confirm it works just like all the other things but I think it will work great when it does, I do think I'll be changing the globals to variables tho since that does not make sense with what I saw in the RRF server
pretty slow, what kinda heater?
oh, yeah that sounds a bit weak
yeah it is
this machine was on 240V
just don't have the money to get new heaters for it rn
also a 750W heater in the same size and in 120V is really hard to find
well maybe not really hard but hard
can't you use 240v?
my dad said no to me running some 240V to my bedroom
we aren't going to be able to get it to our house, too far from the industrial area
it is about like 1.5-2km away
Run fiber instead of copper, lower losses 
I don't think the power company would appreciate
and my mom works for them so she might get in extra trouble for it 
ah fuck I just realized that there can't be filament assignment to the mirror and duplication mode tools
I was making them on a break at school so now I need to modify them I guess
@strange rose do you know if there is something that is being called when a stepper gets enabled? I would want to have the electronics fan kick on whenever that happens but not sure if there is any command that gets executed I could add that to
hmmm adding it to daemon.g is probably not a good idea right?
not sure how you'd do stepper energized -> fan on
You can probably look at object model
I love object model and abusing the plethora of options
yea i use something similar to turn on the chamberheater fans on my k3
yeah not really finding anything about steppers being enabled, I just set it so if an axis is homed it will be kicking on the fans
just used the things in jay's config and put it in daemon,g
if move.axes[0].homed || move.axes[1].homed || move.axes[2].homed ; check whether X, Y or Z drivers are active
M106 P5 S1 ; if they are, turn on the electronics bay fan
else
M106 P5 S0 ; if not, turn off the electronics bay fan```
I think that object model lets you check pin state
move.axes[0].machinePosition might work
its 0 when its not set or unknown
issue is that after a print is done it will not get reset so it will keep the electronics fan enabled forever
then i would just add a global variable and set it in the homing macro and unset it in the print end macro
well then if I am just doing things on the printer it won't activate, I think jay's solution was good with the daemon.g
alright, remade the config.g, just need to finish the cancel.g, and the pause.g I think
You can just look at printing state
In v3.7 you can do OM based triggers which should reduce daemon.g abuse
the printer is on a duet 2 so I can't go to 3.7
bummer.
alright got the whole config loaded up
I don't get it, I have been fighting with that thing forever, never had to do anything like that with any other printers, the Y axis just refuses to home properly and goes in the wrong direction
and it only goes in the wrong direction when homing not when moving normally
I just ran it line by line and now it works? what the fuck
Put M400 between the lines in homey.g to see if it works.
Also, try swapping X and Y steppers to see if that stays with Y or moves to X. The. You may want to inspect the Y driver, depending on outcome.
it ended up working after I simply ran it once manually and now it is working every time, I am really lost on why it didn't work
okay so I am not sure how that can happen but when I do my pause.g I have a line that is putting the toolheads in their parked positions, which is G1 H2 X{move.axes[0].max - 2} U{move.axes[2].min + 2} F3000 ; park X at max - 2mm and U at min + 2mm simultaneously
for some reason the U axis is going to U-145 when it is not supposed to be going there since the minimum for the U axis is -25mm from the config.g```M208 X0:370.4 Y-6:362 U-25:350 Z0:450
so now I am really confused as to wtf it is trying to do
I am just slightly loosing my mind I think
What do you get when you echo move.axes[2].min?
0.0 for some reason
even right after setting them up it is not giving me the right value
the Y axis gives me -6 like it should tho
move.axes[2].min is Z
try move.axes[3].min
It'd XYZUVAB .... and I don't remember beyond that.
And the position it shows may be for tool 1, so you'd have to apply tool offsets for it to make sense. Possibly. I don't remember.
Well I am an idiot, that would explains it, had it defined as X,Y,U,Z,E so thought it was setting the axes in that order
Ah yeah, that would cause some confusion if you're not aware that the output of the OM doesn't care about the order of configuration of axes.
pretty sure you can lookup the definition in the machine model
just sent a 13h (I think) print to it, one plate for all the tradrack parts out of PET-CF and the accent color will be purple ASA-CF
also need to do filamentalist
Time estimation is based on current speed and filament used. Time will go down towards the end a bit.
Well it is telling me 20 minutes now so it wasn't the best slicer estimate but the printer estimate was decent I guess
The entire plate is looking good
Came out perfect
I shall take some when I get back home
@strange rose you got any ideas of what to do for cooling fans on the 22idex?
Kind of seems to have to be mounted to the extruder
Thought about it but the fan would either need to be at the top of the chamber or I would need more holes in the back coming from the electronics bay
The gantry doesn't have any space at the back for a tube to pass
hmm, then keep the stock config and add sheet cooling
Oh yeah that could do
I think the 12032 can't do high temps though so you need something else
Just need to make new ducts because originally it was for copperhead or for v6 not mosquito
Yeah I can probably find something, I should have a couple people I can ask
berdair
Last time I tried it I was melting the parts more and not cooling them 
6mm should be enough breathing room for a 2510 right?
aight that should work
It just releases so perfectly
@strange rose here you go
getting to 40mm^3/s on infill, 2k accel printing and 4k on travel
thats why i like my real CF buildplates
just the right amount of adhesion for PETG and PCTG
also PET from how it looks
might grab a thicker one for my k3, but idk if i want to go with binderclips
could go with another custom made 0.2mm sheet but idk if the transfer tape survives the bed temps
yeah that's a bit annoying
I'll probably be putting holes with thumbscrews on it
but I can't find my taps
time to Hightemp silicone a sheet to some blank springsteel
I am really annoyed because they are really nice taps
👀
It looks so good
some purple ASA-CF, it is my new favorite filament
clean
And all of that is with no part cooling fan
damn didnt even know u got an idex22 thats clean, hows its vfa profile? wondering because it seems to use the same movement mechanics as the q1/p1s etc
The calibration for toolheads is taking a lot longer than expected 
The positions I had were way out of whack
Initial offsets I just eyeballed using a piece of paper with a dot on it
Yeah I did the same but I think I went in the wrong direction at one point and made it worse and then fixed it ish
XY is done but Z needs a bit of calibration, going to use a bit of percussive maintenance to do it lol
I percussive maintenanced too hard and now the left tool is a bit lower than the right one
I fixed the tool issue but I have an issue while changing tools where it doesn't wait for M116 to heat the selected tool up it just goes back to the resume position and then heats up, I posted on the discord and forums so I'll see if we are able to find the issue
Also posted the config to a github page https://github.com/Toaster0042/22idex_config
RRF config for my 22idex V0. Contribute to Toaster0042/22idex_config development by creating an account on GitHub.
Tool changes are also really long rn because tools need to heat up in the tool change, I'll dig around to see if there is a way to do pre heating
Orca has a preheat feature but it used to be broken and used a deprecated command for that
@flint sphinx you could probably try the second gcode stream to interleave the pre-heating. But that would probably require a post processor
Is that 3.7+ only or is that already in 3.6.2?
that's already in 3.6
This page documents the support for multiple motion systems provided in RepRapFirmware 3.5 on Duet 3 boards.
oh
you have Duet2 boards. So no can do
hmm might try that out on my idex
not sure if it will work well as its on a mini5+
Pain
I can't afford to put some duet 3's on all machines lol
And mellow boards are not allowed in my house
I need to learn to make pcb's
Could MPU's work with RRF?
Could have some really good controllers if we can use MPU's I think
What do you mean with MPU?
