Thread: 3D Printing - My rapid prototyping

Originally Posted by BobL

Only about a half dozen staff reported the phishing emails to the company IT support staff and by the time they sent out a do-not-reply email to all staff (and who wants to read IT support emails anyway) it was too late.

A couple of years before I was relieved of my duties by my former employer, I had an account for the company network. Mostly used by me to receive emails from various people in the company, mostly read after it was too late. We would be sent IT support emails from the entire company. If there was scheduled down time for maintainance at 2:00am in Margaret River then we were informed. I finally set rules to reroute it to the bin. Talk about stupid. These are supposedly IT specialists.

When IBM was maintaining the computer systems, they complained about an issue with a production computer because I changed the mouse to left hand. I have always used my left hand for the mouse. Once upon a time it was considered the sensible thing to do. Anyway an IBM IT person was confused about left/right hand mouse use. I laughed.

Dean

Originally Posted by RayG

Stepper motors are slow and liable to drop steps, but they are easy to drive and that outweighs the disadvantages of open loop.
Ray

I don't understand why you would need to drive a stepper motor like a brushless DC motor etc??? There are hybrid stepper motors which contain an encoder and driver cards for them that read the encoder signals This gives the advantages of closed loop without the expense of servo drives etc. Mind you the performance is still stepper motor i.e. good for slower stuff (isn't that the description of 3d printing?) but less torque at higher RPM. Perhaps there is something I've mised?? A search of ebay will turn up a hundred chinese made hybrid steppers and driver cards. Some will be rubbish and some a will be good ?

To be honest, I haven't used a hybrid stepper, I have researched them and at this stage intend to use them on my Sherline lathe when I get to finishing it. Depending on the success of that I will use them on a couple of larger CNC jobs I have in mind, though probably not the Deckel FP2LB as I figure there is too much weight in the column for steppers to handle.

With regard to laser sintering of metals in 3d printing etc, not likely to happen anytime soon for home users. The power required is likely beyond the current limits of your electrical service. None the less, pretty nifty.

Cheers

The Berly Bloke

Originally Posted by Theberylbloke

I don't understand why you would need to drive a stepper motor like a brushless DC motor etc??? There are hybrid stepper motors which contain an encoder and driver cards for them that read the encoder signals This gives the advantages of closed loop without the expense of servo drives etc. Mind you the performance is still stepper motor i.e. good for slower stuff (isn't that the description of 3d printing?) but less torque at higher RPM. Perhaps there is something I've mised?? A search of ebay will turn up a hundred chinese made hybrid steppers and driver cards. Some will be rubbish and some a will be good ?
The Berly Bloke

Perhaps I need to be a bit clearer.

The mechaduino project converts cheap stepper motors into a low cost closed loop system, while retaining the step/direction interface common to 3d printer systems. https://hackaday.io/project/11224-mechaduino

Originally Posted by Theberylbloke

To be honest, I haven't used a hybrid stepper, I have researched them and at this stage intend to use them on my Sherline lathe when I get to finishing it. Depending on the success of that I will use them on a couple of larger CNC jobs I have in mind, though probably not the Deckel FP2LB as I figure there is too much weight in the column for steppers to handle.

I wouldn't use steppers on a mill. I went for 240V Yaskawa brushless DC servos for my (unfinished) mill conversion, Panasonic is another good brand in brushless DC, and still keeping to 240V single phase. With amplifiers and cabling you can expect to pay up to $500 - $1000 per axis, although bargains come up on ebay from time to time.

Originally Posted by RayG

Perhaps I need to be a bit clearer.

The mechaduino project converts cheap stepper motors into a low cost closed loop system, while retaining the step/direction interface common to 3d printer systems. https://hackaday.io/project/11224-mechaduino

OK, so are you saying that 3d printers have a different step/direction interface to a normal stepper motors step/direction interface/command??

Hmm, perhaps I better check my references, I thought that was exactly what hybrid stepper motors and their drivers did???

What was the reason you went with servo's rather than steppers? I see lots of stepper kits for Bridgeport style knee mills around the net.

Not having a go at you by the way, just don't understand and want to!

Apologies to Variant22 for going a bit off topic.

Cheers

The Beryl Bloke

Originally Posted by variant22

If I had my time over I would start with one roll of black Torwell PLA in 1.75mm. I print it at 196 degrees with the bed on 55 degrees and a bit of UHU stick. It produces wonderfully smooth layers with really good adhesion. All the black prints in my previous photos are Torwell PLA. You can buy a single roll on Ebay easy enough - or go direct to the factory and get it for half of nothing. That said a minimum purchase order of 8 rolls does apply. It would have saved me a heck of a lot of messing around had I steered clear of some of the name brands and went for this stuff first up.

I will take a photo of my DRO arm extension in a couple of days. It's Torwell PLA and strong as heck. I can't break it by hand and therefore I trust it to hold my DRO screen - which it's done without issue.

Originally Posted by Pete F

Richard I'd suggest PLA to begin with. It's a plant sugar derived plastic so smells nice when it's printing (a bit like maple syrup). It's a good all around plastic and what I mainly use. Other plastics will have other properties that will be better, but I think this is the best place to start.

Thanks. The machine is supposed to come with a free roll of something. But I might get another roll while I wait just in case they forget to put it in.

Originally Posted by Theberylbloke

OK, so are you saying that 3d printers have a different step/direction interface to a normal stepper motors step/direction interface/command??

No, but servo's don't often use step/direction interfaces.

Servo's use, current position, target position, velocity and acceleration limits are used to generate a trapezoidal velocity profile, which then allows a control loop to drive the amplifier to minimize the error comparing actual position with the trajectory, if the loads and accelerations are such that the trajectory error becomes too large a following error will be generated. The servo interface is usually a +-10V signal to the amplifier and optical encoder feedback.

Originally Posted by Theberylbloke

Hmm, perhaps I better check my references, I thought that was exactly what hybrid stepper motors and their drivers did???

Again, No, closed loop vs open loop, Hybrid usually means microstepping these days, says nothing about closed or open.

Originally Posted by Theberylbloke

What was the reason you went with servo's rather than steppers? I see lots of stepper kits for Bridgeport style knee mills around the net.

When you have a system with variable loads, like a mill and you want to do multiaxis co-ordinated moves, you are better off to close the loop for all axes at the one point,

Properly tuned servo systems can track a precise curve in 3 dimensions under variable loading, something that open loop systems can often fail at. closing the loop at the motor helps a bit but still causes inaccuracy.

Originally Posted by Theberylbloke

Not having a go at you by the way, just don't understand and want to!

Understood, there are layers of subtlety that aren't always obvious. The axis loads experienced in 3d printing are nothing like those experienced in milling.

The stepper conversion kits you referred to are all well and good, but aren't a patch on a decent servo system. But for some, the extra cost and complexity of servos means steppers are a viable alternative.

Hope that helps.

Hi Ray,

Ok that helps.

My use of the term "hybrid stepper" is different from yours. There are steppers available with position encoders on the shaft which give feed back to the drivers such that lost steps are taken into account. Those are what I believed were called hybrid steppers rather than just microstepper dirves. I'm glad this came up as I might have ordered hybrid stepper motors and just got microsteppers! In future I shall take care to call closed loop steppers just that and not refer to them as hybrids. My bad.

I am aware that there can be quite a bit involved in setting up CNC equipment. I downloaded one motor manufacturers PDF on selecting the right motor etc. It was 43 pages long and there were a great many calculations involved. I've quietly saved the document for a rainy day or week or something.

So, I'm still at a bit of a loss. Closed loop steppers provide feedback to the drivers from a position encoder on the shaft. Likely the same sort of position sensor used to provide feedback in a servo motor and drive. Both are closed systems. I don't know enough to discuss the electronics that drive the two different systems in detail. I don't get how the servo system will be more accurate given both are closed and running from the same type of sensor. ?? Surely anything one type of driver card can calculate from the feedback input, the other can as well.

Also in terms of stepper driver cards I am not thinking of a parallel printer port. Rather something like a "smoothstepper" or other board fed from USB or ethernet interface.

Cheers

The Beryl Bloke

Originally Posted by Michael G

The trouble with you guys is that you wave temptation in someone's face...

So if not using stepper motors, what would you use to drive them? - or given the cost of glass scales is it better just putting scales on the machine and using those for location purposes?
Designing (and making) a sturdy 3 axis platform should not be all that difficult. The main problem is the 'cheap' part - making something out of thin cardboard (ie MDF) is wonderfully 'hacker' and cheap but there are better ways of getting rigidity.
What are other's thoughts - is it better to have a platform that moves away from a print head or a print head that moves away from a platform?

Michael
(being distracted in speculation from all the other things he should be doing... again.)

Michael with all due respect to some of the other opinions here, I would suggest some are making a mountain out of a molehill.

Firstly in regard rigidity, 3D printers don't need to have especially high rigidity. There are no cutting forces involved, so the only rigidity requirement is with acceleration of a relatively lightweight hot end. Most machines, at least the ones I've seen, now seperate the extruder stepper from hotend via a Bowden cable, so there's relatively low weight being accelerated. Having said that, as Ray suggested, many cheap printers don't even manage to obtain the feeble amount of rigidity required as they are trying to cheap out on everything, and just build something that looks about right. From what I've seen there's rarely any signs of true engineering involved. They also tend to be built from materials that lend themselves to be reproduced in house by other cheap CNC systems eg laser cut acrylic, wood, etc. if I were scratch building one I'd probably use aluminium extrusion for the frame, as sections are available that lend themselves to secondary function within the design. The thing is you don't need to get carried away with this, and a rigid lightweight chassis will be fine, even in acrylic or wood.

Similarly on the motion control, most of the poor engineering and subsequent issues I dealt with were with short linear bearings being spaced too close together, not considering torsional forces (very common error, especially in Chinese gantry style CNC routers for example), and so on, but all these became quite obvious when I even first assembled the machine. Upon use I just replaced the components and modified the design. It wasn't hard and just required a bit of common sense.

Steppers are used in literally millions and millions of small motion control systems and they work perfectly. Servos certainly could be used, and I intend to eventually roll my own small CNC mill at some stage where I expect to use them, but would be complete overkill in a 3D printer like the small ones we're referring to. I can honestly say I've NEVER had a stepper miss a step in a print, indeed the only time I've heard them mis-step is with the extruder when trying to force plastic through when it's too cold/clog/too fast/etc. That's not uncommon but in that case the mis-step is actually a final circuit breaker to protect and give an audible warning that things are not right in the world. The whole motion control system is very cheap, reliable, and easy to program. Trust me on the latter, as if I can figure out how to program one, any idiot can. I don't like coding and am especially crap at it! Lucky for me the Marlin firmware is already established, so you just go through and program the required characteristics you need. It's a no-brainer to go with an Arduino mega and RAMPS.

Yes you could certainly modify a CNC mill to work as a 3D printer, but the general consensus when this comes up (as it does constantly on 3D printing forums) is that it's a spectacularly stupid idea. I've watched some really fast VMCs run (some Brothers were especially impressive) and their motion is fast by CNC machining standards, blindingly fast in fact and I've been literally shocked, but the average CNC mill is glacial by comparison to a 3D printer. Again they have very different needs to a 3D printer in terms of required rigidity ... they also have absolutely massive motion control systems to get that acceleration. Want to hook up a massive 3 phase multi-tonne machine and tie it up for days doing a print that would take even the crappiest dedicated desktop 3D printer a few hours? Knock yourself out I guess!

While I would never recommend the printer I bought, if you were going to seriously consider starting to design and build one from scratch, a kit like I bought could be the easiest and cheapest way to go about it. As much as I'm loathe to admit it. While many of the parts will finish up in the bin, they will function well enough in the beginning to get you started, and those that don't bespoke Australian landfill couldn't be purchased as cheaply individually, so there's value for money there. I think mine owes me around $6-700 in total, and something like the E3D hotend didn't technically NEED to be replaced, but I'm glad I did. The prints it produces now are as good as any I've seen, but it was a bit of a long and bumpy road to get there.

Designing and building a 3D printer wouldn't be especially difficult these days, and kits are available with all the motion control available. But it would certainly be a considerable project, no I hear of just as many of these toes of projects getting stuck in a corner somewhere as actually completed, so it wouldn't be something to take up on a whim.

Originally Posted by Theberylbloke

Hi Ray,
Ok that helps.

I'm feeling that we are taking this thread too far away from the 3D printing topic. Best start another thread.

Back on 3D printing, a friend who is mad on 3D printing dropped around and showed me a thing he printed and then smoothed with solvent. Looked pretty good, he says there is a system that uses a plastic that can be smoothed with iso-propanol, anyone else heard of this?

Ray

"Vapour polishing" is the term used by some for this. I have seen this work well with ABS prints using acetone, and can give a glossy wet look finish if you get it just right.
The part to be polished is suspended in a sealable container with a small amount of acetone in the base, and left for a suitable time.
Bill

Yes vapour polishing is the term as mentioned by Bill. A solvent needs to be used that reacts with the particular plastic, so it will vary. Some plastics need some nasty solvents so they're not the sort of thing you want to splash on as an aftershave!

I've used acetone with abs but haven't ever bothered with other plastics. It probably depends on what you're using the printed parts for, mine tend to be functional so while I like a good printed surface I'm not interested in fussing with them beyond that. Other people want to print out endless plastic trinkets they download off Thingiverse. Goodness knows what they then do with them, having a house cluttered full of cheap plastic 3D crap is about my definition of hell, but I guess everyone is different. Vapour polishing is probably worthwhile trying if you want to print things like useless plastic vases that will leak (something the media fail to mention in the hype).

Originally Posted by Pete F

Other people want to print out endless plastic trinkets they download off Thingiverse. Goodness knows what they then do with them, having a house cluttered full of cheap plastic 3D crap is about my definition of hell, but I guess everyone is different. Vapour polishing is probably worthwhile trying if you want to print things like useless plastic vases that will leak (something the media fail to mention in the hype).

That's a bit disappointing. I was thinking that a 2m plastic crocodile I could leave floating in my driveway pond would provide a welcoming ambience for visitors....

PDW

Originally Posted by Pete F

Other people want to print out endless plastic trinkets they download off Thingiverse. Goodness knows what they then do with them, having a house cluttered full of cheap plastic 3D crap is about my definition of hell, but I guess everyone is different.

Sounds like there will be so use for an appliance that automatically turns undated objects back into printer filament?

Originally Posted by Pete F

having a house cluttered full of cheap plastic 3D crap is about my definition of hell

The clear impression I get from looking at places like thingiverse, is that 3D printing is a solution in desperate search for a problem.

Printing patterns for casting seems to me the best use I've seen so far, well except maybe the silver jewelry my son got printed by shapeways. The ability to print complex structures with internal moving parts is something other technologies can't handle as well as 3D, I'm thinking of things like the Spacex Rocket Engine parts, I saw parts for a 3D printed A320 bulkhead in Austria, lighter and stronger than conventional.

There should be a whole class of engineering solutions previously unexplored, that should be opened up by 3D printing, instead what do we get.... plastic Darth Vaders ad-infinitum.

Originally Posted by BobL

Sounds like there will be so use for an appliance that automatically turns undated objects back into printer filament?

Yes they've been experimented with. The ability to recycle plastic "fails" and turn them back inti filament is the holy grail of this area, but so far it hasn't been much of a success. Apparently getting the filement a very precise size isn't as easy as it sounds. The commercial ones are the size of a bus, so there must be something going on with them.