Probe Design Ideas and Discussion

[vagrant coral]

Designing a probe that can be stowed and retracted similar to any klicky probe, but with mostly 3d printed parts and cheap springs or rubber bands

current first prototype is printing, will have to adjust this thing once known what would be the problems in the design

[inner shuttle]

I think that without special care in design, repeatability of the latch might be too low

[vagrant coral]

I'll have to consider that then

[inner shuttle]

Like, the latch on the moving part will have a radius, the edge on the stationary part will have a radius. They'll be mismatched and with a significant friction, so they might not play well here

[vagrant coral]

I think I tried to acccount for that on the sketch

there's a miniscule space in there where the latch can hide when its sliding

I don't think that the latch being 1.5mm is way too strong yet though

[inner shuttle]

I wonder if this all could be done as a buckling spring compliant bistable mechanism, with switching force being significantly higher than microswitch force.
So, you'd ram it into something on top to deploy, then ram it into a bed mount for example to stow it

[vagrant coral]

I was also thinking about that

but I figured against it because it might be non-repeatable due to the stresses the plastic will encounter when its deployed

it might just deform it that results couldn't be repeated

[inner shuttle]

You don't have to let springs fully straighten out, you can make them hit a stop and be preloaded against it

[vagrant coral]

something like this?

will take note of that

[inner shuttle]

Yep

[vagrant coral]

I also have a slight idea modification of the design where instead of stow default, it is deployed default

then you ram it to the bed to stow it

which is kinda similar to this, just that it requires a spring/rubber band

my only problem with that is that the idea doesn't have a way to deploy it again by ramming it into anything

https://www.youtube.com/watch?v=qTOFFN_0idA this seems to be a greatish design for that idea

[autumn summit]

I think there was a mechanism you're looking for in that one sunshine vid where the printer electroplates layers after printing

it just bumped the top bar to deploy

https://m.youtube.com/watch?v=W1d36wbx_yg&t=253

maybe put it upside down and make it more compact

[vagrant coral]

looks too flimsy tbh 💀

[vagrant coral]

got the thing printed in PETG roughly, the force required to move the flexing hinge is great, so it might be a great sign

the thing acts like a spring when overloaded, pushing the "rod" back whenever its way too overpushed

the only problem rn is the clearance issues it has when being stowed, the latch is barely getting in the way. Need more clearance near the end holding point

[vagrant coral]

this only happens when I overpush it

the clicking force of the microswitch is lesser than the overload bending force, so it bodes well for the design

[vagrant coral]

wish me luck 🤞

[vagrant coral]

I got the thing to work (sorry for crappy petg print, but phili is too wet, dried this yesterday for 8 hours to get 2 hours of decent print. would have taken a long time if I really had to dry it again)

still have to fix the crappy retraction on the sleeve

the general concept works, but currently its wayyy too long

sending the video here, but would take a long time due to crappy internet lol

[vagrant coral]

SERIOUSLY?!

[vagrant coral]

oh hey

back

This is my current design for the probe

currently, my idea is to have two microswitch for this

1 for cutting off the pin to the motherboard for probe deploy detection
and 1 for bed probing

the design currently works with a bistable mechanism with 9mm of push down on the short sides will cause it to retract

and that is honestly not a bad tradeoff

the problem is that its made in freecad and freecad is just shit at this

oops I lost the data for the probe lmao

anyways yeah... the ratio here is 2:1, for 10mm push and then snap, it pulls it by another 10mm

can't believe my assumption though of the shallower spring leaf would cause a rigid-er spring is true

the deflection is easily doable on the loooooong side of the leaf spring

[vagrant coral]

this is the final design for the intermal mechanism, now lets find a way to mount microswitches on this

the great thing about the added stopper is that it engages the tab to the spring, which adds a force that pushes it downwards, which prevents it from getting slack from the probe's microswitch click

its also a satisfying toy now

[vagrant coral]

Updated Print Files below

[vagrant coral]

Not 1 by 1 with the original leaf spring, I had to modify the thing to get it to work with freecad's dumb constraint system so its not fully 1 to 1, but it should be close

Red is new, Yellow is Original

but does it really matter, its really close anyways

found a much closer fit, 30 degrees

[vagrant coral]

1.5mm is perfect as a leaf spring

its light, yet its NOT that hard than 2.0mm

1mm is just a joke

it just boings with relatively light taps

I haven't tuned the probe yet, some slack would be evident on the bottom

[vagrant coral]

updated the spring, noticed it doesn't constrain with the peg

should now hook better to the peg

[vagrant coral]

oh shit the spring has no clearance lmao

it just sheared off one of the pegs

[brittle inlet]

Gonna try my version now

[vagrant coral]

yeah, I'll wait

but I think I'll modify this again to just use 1 leaf spring, so one surface is a reference surface

that allows for more stability

[brittle inlet]

interesting

[vagrant coral]

and I'll probaby optimize the design more to NOT use pegs

pegs are a nightmare in FDM

[brittle inlet]

oh huh, why?

[vagrant coral]

shears off

[brittle inlet]

fair enough

[vagrant coral]

i'll replace it with a thingy that just uses a hole instead

then the leaf spring is the solid one

[brittle inlet]

curious what you come up with haha

[vagrant coral]

its just a minor modification

cut it in half

and optimize for FDM

that reclaimed space can now be where the wires can be routed 💀

[brittle inlet]

hahaha

[brittle inlet]

extended

mretracted

limit switch registers whether or not it is retracted/extended

pointy end gets hit by the bed

@vagrant coral

[vagrant coral]

Okay that's actually decent

I'm yoinking that

[brittle inlet]

please do!

[vagrant coral]

Nope, too loose. Made it clearance 0.2, too tight

im validating it first by printing

I should probably chamfer that other edge

lmao

[vagrant coral]

Added a slot on to allow for flextures

The top hole was made 0.2mm, shit too hard to push in

[vagrant coral]

i got the microswitches, and a hall effect sensor with magnet

two methods of probing

[brittle inlet]

why hall effect?

[brittle inlet]

this design can't really flex as much as the last one, right? missing the pegs

[inner shuttle]

Because it drifts and that's fun

[inner shuttle]

I still wanna see buckling springs here

[vagrant coral]

yeah, which I had issues with in the fabrication

the side walls interlock, adding a ton of friction

[vagrant coral]

I just bought jt because its cheap

$1 hall effect sensor and $0.5 per magnet

but I intend to do microswitches first, those are really easy to work with

a hall effect sensor would actually be really trivial to implement, I could go really thin leaf springs which has a slack at the end

(its honestly just the bltouch but without the microservo lmao)

[vagrant coral]

my question here is how could you implement it

[inner shuttle]

Exactly how you have yours right now, have these buckle and that's how you get a bistable mechanism

[vagrant coral]

what

huh

[vagrant coral]

so those are rigid pins and the walls flex a bit

quite interesting

my question would he how would you control the amount of distance travelled

mine currently can do 10mm which is enough to lift it off the bed I hope

I've just noticed I need either preload to get that thing to pull back the probe harder

or the tab has to get longer

making the tab longer is easier so I opted for that

probably close to the final design of the internal

ooo nice, it requires 4mm of push to travel 10mm

that is certainly workable

damn someone should really teach me how to properly design shit for 💀

trial and error is fun and all but it takes time lmao

[brittle inlet]

Good call!

[inner shuttle]

There's a bit of math involved. You just can steal existing design as well

[inner shuttle]

I've been lobbying for these since the beginning

[vagrant coral]

well.... need math lol

[vagrant coral]

currently I'm sticking with that design

not final though, still missing the other microswitch

[vagrant coral]

got it working, albeit stupidly

I don't have a rigid mount for it

[brittle inlet]

Nice!

[vagrant coral]

oh shit

remember the hook design I was saying fir corexy?

Why don't I just do that

that removes the travel required

oh, yeah I forgot, that... needs an endstop

...if it needs an endstop, that's basically two switches already, so why not

[brittle inlet]

yeah why not

[brittle inlet]

why?

[vagrant coral]

that needs an endstop precisely just so the probe doesn't crash on the bed

you can still set the Z axis home to the probe, its just that the endstop should also be wired to the probe switch

[vagrant coral]

Probe Design Ideas and Discussion

[vagrant coral]

OH

WHY DIDN’T I USE MAGNETS

That would make that thing compact as heck

There's also this:

It wants to deploy but it can't because its blocked by a pin

To trigger it, you need to send the probe at the edge of the printer, that will deploy because the pin gets lever'd away by the push block

now the issue is that it might not be repeatable, so I guess we need to add magnets on the underside of the spring

That will be hard to push back by the probe considering its going to be resisted by both the spring and the magnet

but if we do full send the probe into a special spot in the gantry, it can now stow away since the gantry is heavier than the spring and the magnet force

but then again this is just the same thing as the original spring loaded thing, just inversed

[autumn summit]

you almost have a strykepoint like that

[vagrant coral]

at least it works with bedslingers I guess

That thing doesn’t seem like it does

time to make it really compact by sticking a pen shaft on it lol

[vagrant coral]

now I just need a probe stow retention

which should be similar to the first design

[vagrant coral]

I know, we steal the tube of a pen, and then add the spring in line to it!

[autumn summit]

maybe magnets?

also make the bottom tip stylus shaped like unklicky or use a screw/pin imo

more accurate, less heat transfer

[vagrant coral]

I will do that, but I did the flat thing simply because I was trying to limit the motion of the thing

[vagrant coral]

the magnet spring type sounds like a nice idea lol, but idk. I'll try the spring being in between of the thing first

its nicer to have the forces be in line with the probe so it doesn't induce a rotation

[autumn summit]

you could use a pin and brass bushings

[vagrant coral]

er what

brass bushings?

this?

if this, nah

[autumn summit]

the ones with graphite

spotted

[vagrant coral]

couldn't you just use a PP straw

[autumn summit]

idk, just thinking about repeatibility

[vagrant coral]

quite fair, but I was thinking it would be funny if you can source almost all of the parts off a 3d printer and a pen

a PP straw usually comes with the pen well, and it has quite low coefficient of friction

its also quite flexible if the PP is sufficiently big

will it last though? Probably not. It's not as good as PTFE

I love how this is kind of stupid though

[vagrant coral]

OOPS WRONG SPRING DIRECTION LMAO

[vagrant coral]

this should be the final prototype I think

that mid point is cursed

2.23 mm

I’m worried about assembly but seems like if I put it 45degrees, it fits

[vagrant coral]

this test was made on this tho

[autumn summit]

looks pretty nice

I'd worry about the fan mount blocking flow

[vagrant coral]

temporary mount

I have no place to place it yet

it’s meant to be attached probed and then evaluated

it not being rigid mounted really close to the vroller + tremendous spring force causes repeatability to be low

but good enough for the first test

uh what

fully lifting the probe (NOT FAST) causes good accuracy?!

truly an engineering tri ven diagram

okay how many samples should I do

I’m gonna let it run for a while

f it, 500 samples

[raw edge]

my 2 cents this looks to be a bistable bed probe project, similar to https://www.thingiverse.com/thing:4736440

[autumn summit]

not bistable tho and a lot simpler

[vagrant coral]

we did go that route, but my main issue with this design is that how would you undock it

[autumn summit]

behind the X extrusion

[vagrant coral]

probably

[vagrant coral]

yep he has to push it down

[autumn summit]

he could go to the top of Z to activate it

[vagrant coral]

its possible to send it upwards towards the Z to undock it

the issue with that is that you have to sense though if the probe is attached

possible with klipper

every 3mm move upwards, you check probe pin, then push up more if not attached

hmm I should probably find a way to make this run typical pen springs to make it accessible to people using household items to make a probe

2 pen springs side by side should give enough force I guess

currently the spring there is harvested off some phone mount

good enough for a demonstrator tho