3D Printer yet?

[mnementh]

Only if, once my junk stops aching from all the kicking,  I am somehow stricken by some sort of prescient clarity. 

I've literally NEVER had a 3DP go this far off the rails; every one I've owned since the Tarantulas has mostly worked pretty well stock out of the box and required very little dickery to go from "Okay" to producing decent, better than usable prints reliably all the time; including this one when I wasn't fucking it up myself.

Pretty much every problem I've had with the Diggro was resolved by figuring out what glaringly obvious mistake I'd made while tinkering, but overlooked for days on end afterwards...    By comparison, my worst headache with the Tornado was some ghosting when trying to do too-small print in sidewalls. 

Usually when they fail on me, it's a go/no-go fix. Maybe I just haven't suffered enough 3DP misery yet to know how to fix this. 

mnem
 

[Zucca]

My hot bed measures 2,5 Ohm.
@24V power 230W on a 235mmx235mm bed about 420mW/cm².

Interesting.

[SilverSolder]

OK I'm expecting a Master Class in 3D printing problem deduction and fix ... 

Machine the part out of aluminum instead? 

[Zucca]

Stepper motor current calibration.

According to here:

https://teachingtechyt.github.io/calibration.html#steppers

and



to calculate the Vref in Ender silent board  it is

VREF = (RMS current * 2.5) / 1.77

according to the TMC2208 data sheet, see here pag 50, it is



which means if you do the math Rsense = 100m .

Unfortunately on my board (Creality Silent) I see R150 and not R100 sense resistors 



So I am almost sure the formula above do not apply to my board, it should be:

VREF = (RMS current * 2.5) / 1.277

Now regarding my motors on my Ender 5:

Z Axis, X Axis
42-34, I max 0.84A

VREF = 1.163 V - 10% (to be safe) = 1.047 V 

Extruder, Y Axis
42-40, I max 1A

VREF= 1.384 V - 10% (to be safe) = 1.246 V

In the future I will have the two Z Axis with two 42-34 in parallel, so I need to adjust the Vref there to 1.047 V * 2 = 2.094 V

PS: Always double check.

[mnementh]

Done long ago on mine. A lot of the manufacturers of the cheapies like to configure the current pretty low to reduce operating noise; the Diggro was no exception even tho it came with TMC2208s.

Are you sure about the current ratings on those steppers...? I have yet to see one in the wild that wasn't safe to at least 1.5A; IME, practical limit is the capacity of the driver, not the stepper. Don't confuse max continuous/hold rating of the stepper with peak current, which is what you're setting on the VRef.

https://wiki.fysetc.com/TMC2208/

In my case, there is a functional issue with the way the extruder is working when the layer fan kicks on; I have confirmed this with manual feeding. Extrusion quantity and quality goes totally out to lunch when the layer fan kicks on, and a huge increase in backpressure. I'm just not sure what to do about it. 

My latest attempt was 220° first layer, 230° regular temp with layer fan set to 30% regular and 50% max speed. Not sure how much more heat I can add before I start boiling the PLA inside the heatbrake. And it's not like I can move the layer fan any lower... so I'm a bit stumped as to what to try next.

mnem
 

[Zucca]

Thanks a lot dragon. It makes sense.

Done long ago on mine. A lot of the manufacturers of the cheapies like to configure the current pretty low to reduce operating noise; the Diggro was no exception even tho it came with TMC2208s.

Are you sure about the current ratings on those steppers...? I have yet to see one in the wild that wasn't safe to at least 1.5A; IME, practical limit is the capacity of the driver, not the stepper. Don't confuse max continuous/hold rating of the stepper with peak current, which is what you're setting on the VRef.

Here the datasheet about some creality steppers:

https://gist.github.com/knoopx/e6c40a009e796203b93a75a3ed6a5ab8

Mine are for example BJ42D15 - 26V02 so no idea, I keep the above as reference.

After rereading the TMC2208 datasheet, with Rsense R150 the creality eng. set the max current on the steppers of 1.28A RMS (Vref 2.5V). So yeah, the bottle neck is the driver not the stepper, kind of, depends on which steppers are installed.

Anyway also here

https://wiki.fysetc.com/TMC2208/

the Irms/Vref calculatio is wrong.

The maximum settable motor current is 1.77A RMS (0.11Ohm sense resistors), but the SilentStepSticks can only be used up to 1.2A RMS.
Irms = (Vref * 1.77A) / 2.5V = Vref * 0.71
Vref = (Irms * 2.5V) / 1.77A = Irms * 1.41 = Imax

with R110 it is not 1.77/2.5 but 1.64/2.5 in legacy mode. Maybe they are meaning UART control. Meh 

Now I am tempted to swap the X,Y creality steppers with a stepping angle of 0.9°.

In my case, there is a functional issue with the way the extruder is working when the layer fan kicks on; I have confirmed this with manual feeding. Extrusion quantity and quality goes totally out to lunch when the layer fan kicks on, and a huge increase in backpressure. I'm just not sure what to do about it. 

My latest attempt was 220° first layer, 230° regular temp with layer fan set to 30% regular and 50% max speed. Not sure how much more heat I can add before I start boiling the PLA inside the heatbrake. And it's not like I can move the layer fan any lower... so I'm a bit stumped as to what to try next.

mnem
 

I would measure the Extruder current and voltage when the layer fan is on/off. Just to make sure it is not an electronic problem. Yes in theory it should not be the case.

[Zucca]

FUUU I couldn't resist

New X motor, here

New Y motor, here

Still looking for a nice direct extruder with hotend, candidates are

https://www.bondtech.se/en/product/bondtech-ddx-for-creality-ender-cr-10s/
https://e3d-online.com/products/e3d-hemera-direct-kit-1-75mm

[mnementh]

Yeah, that datasheet looks like the nominal settings FROM CReality. As soon as you go TMC2208 you get a lot more headroom; I have mine set for 1.1A.

You're right; fysetec's calculator is configured for R100 shunts, not R150. you'll have to do it the hard way with calculating-ily.

So... CReality's documentation blows... you don't know what the real current ratings or °/step on them are.

Primarily, you're ordering these steppers so you can have some specs on them...? That's really sad, CREALITY. 



I've rechecked my VREFs; they're all right where I had them. I figured "Maybe while I was in there fixing that toasted FET on the hotend..." but no. Still dead on.



Currently, I'm following through on my earlier threat to build a cardboard box enclosure... figured cost is zero (well, okay... a few glue-stix), so can't hurt to try.



That Hemera direct-drive is tasty looking... but I was never a huge fan of the V6. 

My biggest gripe with it was always the goofy collar mount, which the Hemera doubles down on their "special mounts for special people" stupidity by making the stepper motor a special part with T-slots buit-in. You'll need to buy/build/print an adapter, or buy a roller backplate to fit the thing to your E3. KA-CHING!

Aside from that, and the ridiculous price, I like that it's all CNC... but you'd think for that kind of money you'd get some ready-to-use-ness... and you don't. You still have to figure out your own layer cooling and heat-brake cooling solution yourself. Seriously E3D... not even the heat-brake fan is included for that price...? 



https://www.aliexpress.com/item/1005001525240599.html

https://www.aliexpress.com/wholesale?catId=0&initiative_id=SB_20201011074350&SearchText=bondtech+extruder

I like the BMG direct-drive extruder; it uses standard parts... except the cloners have produced a nice CNC version for 1/2 the price. And it's not very 'fang-friendly due to tension lever placement.



https://www.aliexpress.com/wholesale?catId=0&initiative_id=SB_20201011072619&SearchText=Titan+direct

Then of course there's the clone Titan direct-drive extruder, which can be had for ~$50 plus you figure out the mounting. I've toyed with that one for a while myself, but it is only single-side drive on the filament and... *head explodes*

mnem

[Zucca]

BMG Extruder Mosquito is very very sexy indeed.

[mnementh]

Yup. Only problem with it is single-side filament drive. Seems with the current crop of clones you can have gear-reduction for more torque or dual-cog filament drive, but not both. Brain fart; I was thinking of the Titan geometry. The BMG Mosquito extruder is both dual cog and 3:1 reduction.


I dunno... perhaps my Diggro needs the enhanced torque when stepping up to the 'fang, even if I don't want to print faster. 

My red dual-cog extruder has smaller, hardened pinions... maybe the increased torque from that will be adequate.

mnem
*fulminating at the mind again*

[mnementh]

If it actually makes a difference, I'll build something a little less tinder-ish. Back to basics; control the environment, control the process.

mnem
Or lose your mind trying... 

[mnementh]

Welp, that didn't take long to crash & burn.

mnem
Time to drop back & punt...

[mnementh]

Red dual-cog extruder fault-fixed (came with wrong hardware, idler lever/cog required realignment with hand-made shims  ) and installed; it was exactly as much of a PITA as I expected. Extrusion is smooth and positive, and nary a tick in all testing, even with layer fan at 80%. 

A quick check of extrusion travel shows it to be under-extruding by approx 35%; this fits with my measurements of the cog being approx 40% smaller. Noice! A 35% increase in torque may be all that's needed!

However, one of the weak points in my Diggro is no CONTROL menu (and no 128x64 emulation either), so no ESTEPS via front panel; it has to be done by consoling in with my laptop. 

That is a misery for another day.

mnem
*toddles off to ded*

[Zucca]

so no ESTEPS via front panel; it has to be done by consoling in with my laptop. 

Why they do not serial comm interface as standard from the front panel in all 3D printer? I do not understand why.
EDIT: Maybe typing with a push and rotate pot is not the best.
Keep up mnem!

For the mechanical guys here, is this coupler:



bad for D shape shafts? It is rated for 280Ncm I will use it at 46Ncm max. It should be okay even if I do not have an O shaft.

[xrunner]

That's the same coupler I'm using (you can see it in the pic).

Recent print is a tank for the scene I'm modelling now on my model RR. It's for a fire suppression system for the oil storage tanks. It's going to connect to the end of the pipe in the attached scene picture (pipes will be painted red too). Still needs to be sanded a bit and painted.

No recent mods to report here. Might install Marlin 2.0.7.1 if I get bored. 

[mnementh]

so no ESTEPS via front panel; it has to be done by consoling in with my laptop. 

Why they do not serial comm interface as standard from the front panel in all 3D printer? I do not understand why. EDIT: Maybe typing with a push and rotate pot is not the best. Keep up mnem!

For the mechanical guys here, is this coupler:      bad for D shape shafts? It is rated for 280Ncm I will use it at 46Ncm max. It should be okay even if I do not have an O shaft.

That is a drive motor coupler meant for noise reduction on spinny things, NOT meant for steppers.

Unless you design your linear actuator with separate thrust bearings (most 3DP do not because:cost), that is not a good choice. Sometimes you can get away with it if the thrust load ONLY puts the coupler in compression; however 3DP will put load on the coupler on both compression and tensile planes, even on the Z-axis.

To work properly, stepper motors on a 3DP usually need a coupler with a one-piece body; repeat accuracy goes out to lunch on designs where the halves can move in the same axis as the load is applied. Remember, steppers apply load in torsional plane AND longitudinally on the leadscrew axis; that coupler is only meant for torsional loads 90° from the leadscrew axis.

   

There's a reason CReality spent the money on these custom-designed couplers instead of using the common dime-a-dozen spiral-cut ones.

mnem
 

[Zucca]

There's a reason CReality spent the money on these custom-designed couplers instead of using the common dime-a-dozen spiral-cut ones.

The reason is they are cheaper than the others. 
Searching on the www is seems either you have people who love those flex coupler or people that hate them.
I think it is good to have some flex coupler available to test them out in case you want or need to.
I will make my own tests.

[Jay_Diddy_B]

   More Frustion360   .

mnem
I'm gonna go put ice on my junk now.

Mnem,

I think you may get better cooling, less resistance to airflow if you have a smaller number of wider slots. In products I have designed, I would make the slots as wide as the safety regulations would allow. This may also help with 3D printing.

Would you share your STL or F3D file? I would be happy to try the first few layers of this design on my 3D printer.

What filament material are you using?

Regards,
Jay_Diddy_B

[mnementh]

There's a reason CReality spent the money on these custom-designed couplers instead of using the common dime-a-dozen spiral-cut ones.

The reason is they are cheaper than the others. 
Searching on the www is seems either you have people who love those flex coupler or people that hate them.
I think it is good to have some flex coupler available to test them out in case you want or need to.
I will make my own tests.

No, it's because these couplers are rigid or semi-rigid in the tensile/compression plane, so they are somewhat precise against thrust load, which means the designers can omit thrust bearings in their linear actuator design. THAT is cheaper by far than a $2 coupler vs the 50c spiral-cut ones. A properly designed linear actuator's leadscrew has its own thrust bearings independent of the motor; this is one of the first things that went by the wayside in the China-direct MFG race to the bottom in this arena.

The spiral cut couplers are essentially springs with no rigidity in thrust loading, and THEY are the cheap AF ones. Like 3/$5 delivered to the end-user. They are intended for applications where there is no thrust loading, but they are also the cheapest you can get, so that's what winds up on cheap 3DP.

mnem
 

[mnementh]

   More Frustion360   .

mnem
I'm gonna go put ice on my junk now.

Mnem,

I think you may get better cooling, less resistance to airflow if you have a smaller number of wider slots. In products I have designed, I would make the slots as wide as the safety regulations would allow. This may also help with 3D printing.

Would you share your STL or F3D file? I would be happy to try the first few layers of this design on my 3D printer.

What filament material are you using?

Regards,
Jay_Diddy_B

There's plenty of open cross-section; calculated area is nearly 2X that of the vents directly behind the grill and that of the fan in the back. If anything, I want a little resistance here; this thing has a jet engine inside and needs muffling at both ends.

The came from first, beating Frustion360 into submission on 2 different designs, and then printing problems with my 3DP... I don't think the design is at fault here; I can't even get the little fuckwidget to print a cal cube.

I'm seriously considering just putting the whole hotend back to stock... I'm 98% positive that will work fine. It just drives me nuts that everybody and his grandmomma can get the 'fang to work, but it just keeps kicking me in the junk.

I'm using PLA. Right now, I've started moving the 'fang up 1mm at a time to see if I can find some middle ground where it still works but doesn't freeze out the nozzle.

mnem
 

[mnementh]

Okay... so right now, I have the 'fang slid up higher than I've ever had it... so that approx 1mm (maybe less) of the vent opening is below the heatblock/sock. I have turned the layer fan down to 50% max, 30% regular, hotend at 205° first layer/250° after, and I just set flow control to 90% in Cura because of excess squish/extrusion around the lettering in the front.

Last test was 240°, and I JUST got some layer adhesion. 

mnem
 

[mnementh]

Time to drop back & punt...

After talking with some folks with better knowledge, I've come to the contusion that this 'fang... as popular as it may be... just doesn't put the air where it should be; it puts too much right at the nozzle, which fact appears to be the cause of my misery.

So, I've added some sheet metal deflectors to stop air from going where it shouldn't; I discovered while poking around under the thing that the heat-sink fan dumps a lot of air right on top of the heatblock, so I made a shield to go under the heat-sink. That actually did help a fair bit, but still running crazy hot just to get even a little extrusion, so blobby fugly everything where it used to be pretty clean on the first few layers.

Now I'm trying another shield extension around the nozzle to see if it helps. If still no joy, I think I'm just going to put it all back to stock and stuff the bitch in a box. That should shut it up. 

mnem
All this assache just to get away from some fuggin' noisy-ass fans...

[mnementh]

Okies... something positive here. The above mods, along with dropping the layer fan speeds to 25% and 30%, has resulted in decent layer adhesion at 230°. Only get an occasional pop back from the extruder drive, indicating backpressure is still pretty high, but at least it is printing well enough I'm inclined to let it try and finish.

mnem
*toes, wings, tail & claws crossed for luck*

[mnementh]

Nope... turned to shit minutes later. Constant pop-clink from the extruder, shit adhesion. Couldn't get it to start extruding again even by turning the layer fan off, so something is fucking goofy with this hotend too.

Fuck this noize. I need to get something WORKING here.   

mnem
POS * ANOTHER POS = POS².

[Zucca]

Now do you want to go down the rabbit hole further or wait until the Kickstarter Gods will provide you a "solution"?