3D Printer yet?

[beanflying]

I did it the lazy way. Sketched 3 concentric lines, a couple pulls, a quick chamfer. A minute or two at most, even for me. 

Problems with limits of a 0.40 nozzle vs fit on tiny, densely populated PCBs now have me on redesign #3. I got the ID of the sleeve good on the 2nd try; I kept having to make the large shoulder smaller and thicker to make clearance for this or that which I couldn't see until I was actually fitting the board. 

Rendering STL now; then off to make another test print.

mnem
 

Clearance on internal holes is a bit of a guess until you workout your own rules of thumb. For me 2.5 and under are +0.3mm diameter and 3-6 works well at about +0.2mm.

With a your model Profile sketch and revolve it and you are done. The first few times might do your head in a little until you figure out what to reference it to (construction line at the axis) then it actually becomes faster and less steps. Give it a bash even with your current simple job to try it.

[GeorgeOfTheJungle]

I got the ID of the sleeve good on the 2nd try; I kept having to make the large shoulder smaller and thicker to make clearance for this or that which I couldn't see until I was actually fitting the board. 

For me the internal diameters turn out to be always smaller, seems like if the slicer centers the spaghetti on the plane of the wall, so the radius ends up being 1/2 the spaghetti less, or IOW, the diameter a whole spaghetti width less => always add that to the holes' IDs in the design. At least that's what works for me!

[beanflying]

If you also think about what an STL is it is straight lines between points. Small Circles have less points so they shrink more is actual terms. Not sure if there is any numbers on it but it will also vary with the Poly Count of the model.

[mnementh]

Yeah; I test-printed a couple designed at spec, then selected the inner wall and pushed .05mm a couple times till I got to 2.2/3.0mm inner/outer. At that point, I was at 0.4mm wall thickness... so really just arguing with physics, not the software. 

It was the large shoulder that wanted to mess with me. I started at 6.5mm OD as that matched the larger grommets; found as I test-fit that it would just hit here or there. I made the mistake of trying to cure with a chamfer, as that was my plan to begin with.

Should have backed out past the original chamfer in the design seen above, but thought I'd get smart and did a new sketch for a new body to pull/cut and make the OD right; that worked but it made something broken in the body and it wanted to slice weird.

Once I had the dimensions "adjusted to meet reality" in my head, it took literally a minute to make a new design; it's waiting for their server to do the render that costs time.  Of course, I'm not daft enough to expect the same time-slice priority as the folks who actually PAY for their services.

mnem
 

[beanflying]

So I had a spare minute/coffee 

I offer up the Holey Intolerant Printer Test. Each step up is 0.2mm from the base number as shown so 1, 1.2, 1.4 etc. so you can have a print and check what size to model your holes/insert/rods etc are to either fit or bind.

Don't want the Vertical (I recommend you print it as printers differ) stop it or chop it in the slicer. Or if you want to crunch the model it is here to download and tweak it to Imperial or add it out to 0.1mm steps https://a360.co/3eJcjfp

For those who just want to print it the STL is in the Zip file below.

[HobGoblyn]

Anyone tried the PID Autotune?

I followed the guide from here

I've only tried the nozzle.

I ran it 3 times with it getting it's result from 25 passes and a further 3 times with it getting its results from 5 passes, all at 215 degrees, all with side fan on.

Results from 25 pass test

1)
PID Autotune finished! Put the last Kp, Ki and Kd constants from below into Configuration.h
Recv: #define DEFAULT_Kp 20.47
Recv: #define DEFAULT_Ki 1.63
Recv: #define DEFAULT_Kd 64.42

2)
PID Autotune finished! Put the last Kp, Ki and Kd constants from below into Configuration.h
Recv: #define DEFAULT_Kp 19.14
Recv: #define DEFAULT_Ki 1.51
Recv: #define DEFAULT_Kd 60.61

3)
PID Autotune finished! Put the last Kp, Ki and Kd constants from below into Configuration.h
Recv: #define DEFAULT_Kp 21.52
Recv: #define DEFAULT_Ki 1.75
Recv: #define DEFAULT_Kd 65.99

Results from 5 pass tests

1)
PID Autotune finished! Put the last Kp, Ki and Kd constants from below into Configuration.h
Recv: #define DEFAULT_Kp 22.52
Recv: #define DEFAULT_Ki 1.87
Recv: #define DEFAULT_Kd 67.70

2)
PID Autotune finished! Put the last Kp, Ki and Kd constants from below into Configuration.h
Recv: #define DEFAULT_Kp 25.87
Recv: #define DEFAULT_Ki 2.18
Recv: #define DEFAULT_Kd 76.73

3)
PID Autotune finished! Put the last Kp, Ki and Kd constants from below into Configuration.h
Recv: #define DEFAULT_Kp 21.59
Recv: #define DEFAULT_Ki 1.74
Recv: #define DEFAULT_Kd 67.09

So the values between the 6 tests vary from

P: 19.14 - 25.87
I:   1.51 - 2.18
D: 60.61 - 76.73

I expected a slight variation, but this seems like autotune isn't really worth it at all, or something in my machine isn't working?

[GeorgeOfTheJungle]

Autotune may not be perfect but bang bang is a disaster (tho it's ok for the bed).

[bitwelder]

I'm doing some maintenance on my Ender 3, and as I paid more attention on the power connections, I noticed that the XT60 connectors used to join the PSU and the controller are such that the male ehm... plug is on the PSU side and the socket on the board side.
Checking through a few YT videos that show them, it seems that's the standard way Creality mounts them.

Why is that? Usually isn't so that the connector that provides power should be as protected as possible from accidental shorts, etc. and so use the socket type (as actually it seems to happen with all LiPo batteries) ?

[Monkeh]

I'm doing some maintenance on my Ender 3, and as I paid more attention on the power connections, I noticed that the XT60 connectors used to join the PSU and the controller are such that the male ehm... plug is on the PSU side and the socket on the board side.
Checking through a few YT videos that show them, it seems that's the standard way Creality mounts them.

Why is that? Usually isn't so that the connector that provides power should be as protected as possible from accidental shorts, etc. and so use the socket type (as actually it seems to happen with all LiPo batteries) ?

I would say that's because, as I pointed out previously (and got a lot of hate for..), Creality have a sore lack of actual engineers. Monkey see, monkey do, just slightly better quality than the other monkeys.

[mnementh]

No comment... 

mnem
mmmmmmnlmmmmmrmmmmphlmmmmrppphhhh!!!

[beanflying]

I'm doing some maintenance on my Ender 3, and as I paid more attention on the power connections, I noticed that the XT60 connectors used to join the PSU and the controller are such that the male ehm... plug is on the PSU side and the socket on the board side.
Checking through a few YT videos that show them, it seems that's the standard way Creality mounts them.

Why is that? Usually isn't so that the connector that provides power should be as protected as possible from accidental shorts, etc. and so use the socket type (as actually it seems to happen with all LiPo batteries) ?

I would say that's because, as I pointed out previously (and got a lot of hate for..), Creality have a sore lack of actual engineers. Monkey see, monkey do, just slightly better quality than the other monkeys.

If you think 'hate' was what you got for making GROSS SWEEPING Generalizations then as you are clearly a 16 year old and have been online for two minutes you should look at your choices. 

Again you make a CLAIM of a Sore lack of Engineers. Please inform the Audience how many Engineers Creality has and then proceed to tell us how many they should have?

[Monkeh]

I'm doing some maintenance on my Ender 3, and as I paid more attention on the power connections, I noticed that the XT60 connectors used to join the PSU and the controller are such that the male ehm... plug is on the PSU side and the socket on the board side.
Checking through a few YT videos that show them, it seems that's the standard way Creality mounts them.

Why is that? Usually isn't so that the connector that provides power should be as protected as possible from accidental shorts, etc. and so use the socket type (as actually it seems to happen with all LiPo batteries) ?

I would say that's because, as I pointed out previously (and got a lot of hate for..), Creality have a sore lack of actual engineers. Monkey see, monkey do, just slightly better quality than the other monkeys.

If you think 'hate' was what you got for making GROSS SWEEPING Generalizations then as you are clearly a 16 year old and have been online for two minutes you should look at your choices. 

Again you make a CLAIM of a Sore lack of Engineers. Please inform the Audience how many Engineers Creality has and then proceed to tell us how many they should have?

Oh here we go..

Look, I've used several, own one, and seen many Creality printers, and I take issue with the lack of engineering. The XT60 being the wrong way around shows nobody thought about it. The long-inherited AVR latchup on their controllers being 'fixed' by swapping the linear regulator for a switcher shows nobody investigated it. The cheap XT60s burning up showed nobody verified the parts were up to task.

I've yet to see any actual electronic development, just rehashed open source designs and straight up copies.

Either they have no EEs or the ones they've got aren't much good. That's my claim. Don't like it? Show me where they've done something original or just done something well, instead of just better than the average clone, rather than attacking me for being unimpressed.

I still think they're far, far from the worst buy. I just don't think they're at all special and I'm not quiet or apologetic about that.

Now, please inform us how many engineers you can see evidence of them having, and how many you think they should have..

[mnementh]

CReality does have a proven track record of poor engineering; just like pretty much every 3DP vendor has had at least a few instances, even *gasp* PRUSA. OTOH, they sell more units than probably the next 10 smaller vendors combined; IMO, that along with the fact of being a recognized brand probably accounts more for this impression of poor design than any single failure in design.

For the most part, CReality is better-than-average at best; yes there certainly are brands with a higher general quality level, even in the hobbyist 3DP market, but nothing that even comes close to their price bracket. Bang/buck they sure are hard to beat; I see them as purveyors of the best-polished turds in a sea of shite. 

This fact is why I consider them the obvious choice for a FIRST 3DP; the fact you can buy a machine from them knowing that it will (almost every time) work correctly the first time for under $300 (INCLUDING a good assortment of consumable supplies), get your feet wet in 3DP with the support of a huge & knowledgeable user base, and find out whether 3DP is FOR YOU so cheaply is a combination that nobody else can offer. 

The CR6-SE is the first product I've seen from them that I could consider a reasonably original design; that and the fact that CReality actually DOES HAVE a minimum level of quality (something the vast majority in this market do not have) was mainly why I bought it instead of waiting for the inevitable cheap clone.

mnem
*puttin' on muh flame-proof jammies*

[ANTALIFE]

Time for me to toot my horn, I finally managed to get our Prusa i3 MK3S + MMU2S to give reliable prints!

If another else has an MMU2S that just won't play nice, here is what I had to do:
https://www.antalife.com/2020/07/project-prusa-i3-mk3s-mmu2s.html

[CatalinaWOW]

I have recently proved that I, like many posting here, am finding 3D printing as much a hobby as a tool.  After most of a year just designing, slicing, printing and using I suddenly couldn't get anything to print.  Poor bed adhesion, prints that did the first several mm perfectly and then began generating air spaghetti and several other maladies.  This was occurring on a design that I print regularly generating a few dozen copies.   A variety of minor problems caused this, all sorted now, but it demonstrated that at least at the Creality Ender 3 end of the scale it isn't a tool you just use.  Frequent and sometimes continuous fettling is required.  Reports from colleagues using these printer in professional settings indicate that the problem is not limited to the low end of the price scale.

I still love it.  Still find it addictive, but newbies be warned.

[mnementh]

LOL... well, yes; it IS called "Hobbyist 3DP" for a reason. 

If you want to escape the "hobbyist" aspect, you need to get into "Additive Process Manufacturing". Be prepared to bring a fat checkbook or mortgage the kids; those turnkey solutions have corporations, CEOs and their families to support.

mnem
 

[CatalinaWOW]

My colleagues are reporting on results from machines that are paying for mortgages, yachts and Carribean vacations.  More stable? Yes.  Easier to use?  Maybe.  Things like cartridge loading and multiple print heads.  But they still find that it isn't like printing ten more copies of the quarterly report.

[Mr. Scram]

LOL... well, yes; it IS called "Hobbyist 3DP" for a reason. 

If you want to escape the "hobbyist" aspect, you need to get into "Additive Process Manufacturing". Be prepared to bring a fat checkbook or mortgage the kids; those turnkey solutions have corporations, CEOs and their families to support.

mnem
 

I don't think any real turnkey solutions exist, other than paying someone else to go through the motions.

[mnementh]

My colleagues are reporting on results from machines that are paying for mortgages, yachts and Carribean vacations.  More stable? Yes.  Easier to use?  Maybe.  Things like cartridge loading and multiple print heads.  But they still find that it isn't like printing ten more copies of the quarterly report.

Well no... I'm not talking about printer farms and serving orders from ThingiVerse; I'm talking about Additive Manufacturing that is being used in industry in place of conventional molded parts, like automobile manufacture. The term "turnkey" there is a service includes a small army of people whose job it is to set up, maintain and manage the hardware, as well as to turn somebody's 3D modeling into a part that can be produced economically via Additive Manufacturing.

Yes, it is already happening.

mnem
 

[CatalinaWOW]

My colleagues are reporting on results from machines that are paying for mortgages, yachts and Carribean vacations.  More stable? Yes.  Easier to use?  Maybe.  Things like cartridge loading and multiple print heads.  But they still find that it isn't like printing ten more copies of the quarterly report.

Well no... I'm not talking about printer farms and serving orders from ThingiVerse; I'm talking about Additive Manufacturing that is being used in industry in place of conventional molded parts, like automobile manufacture. The term "turnkey" there is a service includes a small army of people whose job it is to set up, maintain and manage the hardware, as well as to turn somebody's 3D modeling into a part that can be produced economically via Additive Manufacturing.

Yes, it is already happening.

mnem
 

I think we agree on what is happening.  Only disagree on the description.  Hydroelectric dams are a long standing and proven technology.  There is a well developed industry with armies of people ready to implement your dam.  But you can't buy one off the shelf.  Each one is a custom design with variables ranging from available water to available topography and the underlying soil types and strengths.

Same thing with office buildings.  Mature and widely established technology.  But with few exceptions each one is a custom design. 

PWBs fifty years ago were a widely used and large scale industry.  But design rule checkers, standardized processes and the like were manual, spotty and or non-existent.  Multi-layer processes were in their infancy. Designs which required multi-layers were usually iterative and painful.  You negotiated with board fabs (either in house or contract places) over what they could do, and then tested to find out if they were stretching their claims (or had lost the bubble).  It was as much art as science.  But widely used. I think that is not a terrible analogy for 3D printing today.

Perhaps the word printing is the problem.  In todays world it conjures up a trivial task that anybody can do.  Basically compose your document and hit the print buttom.  What comes out will be exactly (with possible minor color precision problems) what you aimed for.  While 3D printing is more like ceramic pottery or finish carpentry.  There are tools which can be used to generate the product, but much knowledge and attention is required.

As a closing thought, another analogy is farming.  The tools are all there, used on an industrial scale.  At the industrial level results are more consistent, but still require constant adjustment to local conditions and requirements and frequent emergency responses due to insects, plant disease, weather, machinery failure and other issues.  At the hobby scale almost all of the tools available to the industrial farmer are available in some watered down form.  And all of the same problems occur with less visibility and less effective responses because of the lower quality of the tools. 

[Mr. Scram]

My colleagues are reporting on results from machines that are paying for mortgages, yachts and Carribean vacations.  More stable? Yes.  Easier to use?  Maybe.  Things like cartridge loading and multiple print heads.  But they still find that it isn't like printing ten more copies of the quarterly report.

Well no... I'm not talking about printer farms and serving orders from ThingiVerse; I'm talking about Additive Manufacturing that is being used in industry in place of conventional molded parts, like automobile manufacture. The term "turnkey" there is a service includes a small army of people whose job it is to set up, maintain and manage the hardware, as well as to turn somebody's 3D modeling into a part that can be produced economically via Additive Manufacturing.

Yes, it is already happening.

mnem
 

That's not really the common definition of turnkey. It would mean buying an Ariane 5 rocket is also turnkey. 

"A turnkey, a turnkey project, or a turnkey operation (also spelled turn-key) is a type of project that is constructed so that it can be sold to any buyer as a completed product. This is contrasted with build to order, where the constructor builds an item to the buyer's exact specifications, or when an incomplete product is sold with the assumption that the buyer would complete it."

[HobGoblyn]

I've had  a week off due to other things.

I've put the SKR V2 board in and it's working fine (the board that is)

I've managed to ding the center of my magnetic bed so have the Creality glass one in a box next to me.

I've bought another TFT screen same one as before {originally bought in package with SKR 1.2 which I returned due to fault).

I've put a dual extruder on and followed a guide by Luke Hatfield on fixing the hot end clogging.

I now have zero extrusion problems and have calibrated my esteps, went from 93 to 143.5.  If I put a black line on filament 100mm behind extruder and feed 100mm through hot end, black line lines up with extruder perfectly.

I also calibrated my x axis according to a guide, where both sides measure exactly the same from the top (mind you I would have thought it's more important to line up with the bottom)

I had the x belt  and the screw thread removed, all sides now moving up and down (and left and right) perfectly, not too loose, not too tight.  When I remnoved the x belt,  moving it left to right I noticed the head stopping every full rotation of the wheels as though one had a nick in. I swapped them around and it's gone away (will order some spares).

And I put it together, did a test print and it's perfect.

Then I did another and it wouldn't stick

I'm having levelling problems again. I can do 4 corner levelling 5 times in a row, 4 times it will be perfect, one time it will be to high.

I've noticed that the right hand side x axis bracket, when it homes, it moves slightly and the bottom wheel is not touching (all hex nuts tight), this is after I've already gone through adjusting this when I did the x axis calibration.

I also notice that if I push the right hand up with a tiny amount of pressure with one finger, it rises by a couple of mm, I presume something isn't right.

Tomorrow I'm taking the entire thing to pieces, reassembling and making sure every bit is square.  I have also bought a BLtouch and will fit at the same time.

As a last resort I could fit a 2nd z axis but will see how I go.

Will let you know

[CatalinaWOW]

I've had  a week off due to other things.

I've put the SKR V2 board in and it's working fine (the board that is)

I've managed to ding the center of my magnetic bed so have the Creality glass one in a box next to me.

I've bought another TFT screen same one as before {originally bought in package with SKR 1.2 which I returned due to fault).

I've put a dual extruder on and followed a guide by Luke Hatfield on fixing the hot end clogging.

I now have zero extrusion problems and have calibrated my esteps, went from 93 to 143.5.  If I put a black line on filament 100mm behind extruder and fees 100mm through hot end, black line lines up with extruder perfectly.

I also calibrated my x axis according to a guide, where both sides measure exactly the same from the top (mind you I would have thought it's more important to line up with the bottom)

I had the x belt  and the screw thread removed, all sides now moving up and down (and left and right) perfectly, not too loose, not too tight.  When I remnoved the x belt,  moving it left to right I noticed the head stopping every full rotation of the wheels as though one had a nick in. I swapped them around and it's gone away (will order some spares).

And I put it together, did a test print and it's perfect.

Then I did another and it wouldn't stick

I'm having levelling problems again. I can do 4 corner levelling 5 times in a row, 4 times it will be perfect, one time it will be to high.

I've noticed that the right hand side x axis bracket, when it homes, it moves slightly and the bottom wheel is not touching (all hex nuts tight), this is after I've already gone through adjusting this when I did the x axis calibration.

I also notice that if I push the right hand up with a tiny amount of pressure with one finger, it rises by a couple of mm, I presume something isn't right.

Tomorrow I'm taking the entire thing to pieces, reassembling and making sure every bit is square.  I have also bought a BLtouch and will fit at the same time.

As a last resort I could fit a 2nd z axis but will see how I go.

Will let you know

One issue I had with non-repeatable leveling was cable routing.  If not carefully arranged they can contact the leveling wheels as the bed is moved about.  They take almost no torque to turn so a little rub can result in an uncommanded automatic level change.  A little thought and a few zip ties permanently eliminated the problem.

[mnementh]

@CatalinaWOW: Exactly where I was going. Things have evolved; the March of Progress etc.     In today's corporate world, "Turnkey" often includes a service contract. 

Here's some actual "hobbyist-grade" 3DP being sold in production quantities as part of a hobbyist kit:

                  

https://www.eevblog.com/forum/testgear/test-equipment-anonymous-(tea)-group-therapy-thread/msg3126386/#msg3126386

mnem
 

[HobGoblyn]

6 hours later and the frame is square.

After this experience, I wonder about a lot of the info out there on the web, of course there's always the possibility I squared it the hard way.

Loosen the 4 bolts either side to get your printer feet level, but 1/4 of an extra turn on one bolt when tightening them,  can throw the Y axis out, tighten the top two a bit more and it can throw the vertical bar out by about 3mm from top to bottom.

If I loosen the two bolts (that hold the vertical in place) on just one side, depending on how I tighten them, measuring between the two verticals can be as much as 1/4" difference from top to bottom.  I presume if I did this both sides, it could be even more out.

I've used Aluminium foil as shims to get everything exactly square   (and clamped top bracket against both verticals to maker sure they were properly aligned with each other) , but even then depending on how much I tighten a particular bolt, it can throw things off by a couple of mm.

I finally got the verticals square and realised I'd done it the wrong way around (cut out for cable on wrong side) so undid the cross beam and turned it around   still took me ages to get it back all square though.



My point is that so many peoples advice involve loosening bolts here and there to adjust whatever, but I suspect by doing so, most people will end up with a less square printer than they started with.