New build, pre-design questions, extrusions, shafts, dimensions, electronics

Six months ago I have bought a Malyan M150 as my entrance into the 3D printing technology. The idea was to change a few parts, make a few suggested upgrades and make it much better. A few months later I got aware of it's limitations, so I want to build something myself. Of all the system I have seen I like Ultimaker idea the most, but I am not an mechanical engineer nor have much experience with the 3D printing technology, so I need a few suggestions, before I start drawing, and probably a few afterward.

1. dimensions. Outside dimensions of the cage should be about 500x500x500mm, for about 380-420mm printing capacity in each direction.

2. extrusions. I have read about people making machines with 20x20mm profiles, but also read about people who wanted rigid construction, that they opted for 40x40 and 40x80. I want a heavier machine, not prone to vibration, capable of high speed movement, but do not want to oversize the cage. I am thinking about 30x30mm for lower horizontals and verticals and 30x60mm for the upper horizontals. Would that be enough from the standpoint of rigidity?

3. shafts. I understand that in that configuration the rails hold very small weight and are mostly exposed to high torsion forces. If available I would probably go with anodized aluminum or hollow anodized aluminum, but it is probable that it would not be available, so if anything fails it leaves me with chrome plated hardened steed HRC60. Anyway 8mm should be enough for steel

4. plain bushings. there are three options available, store bought (like INA Elgoglide), store bought generic sintered bronze or privately machined ptfe or uhmw-pe. I have read about Elgoglide and I like more what I have read there then what I have read in similar Igus catalog about Iglidur. Sintered bronze should be the most economic solution, and the machined solution may be in between. Here are the concerns, quietness, longevity, no need for maintenance (or occasional need for maintenance).

5. Z-axis / build plate may present itself as a problem. I would like to take even an oversize trapezoid lead-screw and more then two rails to avoid bending of the build plate, but I would also avoid use of two Z axis motors (one of two things that I most often do on my Malyan is realignment of the Z screws).

6. electronics. I have not been able to find if ARM boards do support Marlin firmware and does Marlin support this kind of machine. I would like it to be a 2560 Arduino based board, as I do want to pair it with a Raspberry I do not need all the functions that faster boards have. After the initial search I found a few boards interesting, but no reviews on any of them. The options that awoke my interest are MakerBase MKS 1.4, Megatronics 3.0 and Rumba. Essentially the board should have a few removable drivers and enough of them for at least two extruders, expansion connectors for card slot and graphic display. I like Rumba at most, but could not really tell why. Is there any real difference between the Rumba from various chinese sellers, highest price I found at a place called ReprapDiscount (with a bull head logo on back side), and for half of that price at a place called BigTreeTech (BIQU logo on the back side) , can there be any real difference between two chinese sellers? The only differences that I saw are blue and green screw terminals and heat sinks on the BIQU board against the black screw terminals and capton tape on the usb connector without the heat sinks on the RD board, both are clean and well soldered, and also one coil has different value between the two beards.

I have read more than a few feuds about original vs clone and similar, generally I do agree and do not want this discussion to be about that. I have almost bought a Flexion extruder for my printer, but at the last moment opted for E3D (have not yet bought it, but I intend it to be an original). I need to have original working on this printer before even considering a clone. I have three Raspberries (all original), but have never bought an original Arduino, Sparkfun or Adafruit module, but I generally agree on the idea of supporting the original authors.

That's probably enough information to start a discussion. I will be adding the drawing when we set at least the extrusion dimensions.

I suggest that rather than starting from the outside (the frame) and working inward, you start at the inside (the extruder carriage) and work your way outward. If you build the frame and then discover it isn't big enough to give you the XY movement you want, you'll be very unhappy. Figure out what you want to be carrying around in the extruder carriage (LEDs, fans, extruders, circuit boards, hot-ends, etc.), get a rough idea of its size, then figure out how long the guide rails need to be to move as far as you want in the X and Y directions, then figure out how big the frame will have to be to hold the rails that allow that movement. If you intend to enclose the machine for printing ABS, think about how you're going to do that and consider how that affects the frame design. Also consider where you're going to put the electronics, how you'll route cables, and how you'll access them for servicing the machine- you're going to be doing a lot of that for a while.

Look into heaters before you decide on the bed size. It is a good idea to stick with off-the-shelf parts, unless you're prepared to shell out money for a custom bed heater.

8mm end supported rails are insufficient for any but the smallest printers- they are too flexible. You'll need larger diameter rails to increase stiffness, or use some other type of bearings such as linear guides.

Look into the 32 bit controllers. Replicape is an add-on for BBB to control 3D printers. There are also dedicated 32 bit controllers available (Duet and SmoothieBoard). You're going to be doing a lot of tuning and config adjustments so I'd avoid anything that requires compiling firmware for configuration changes (Arduino/RAMPS).

Edited 1 time(s). Last edit at 02/03/2017 07:17AM by the_digital_dentist.

---

Ultra MegaMax Dominator 3D printer: [drmrehorst.blogspot.com]

Thank you for the pointers, all of them very good.

I would design all in 3D before buying a single part. It's just simpler for me as a novice to start from outside, from the element which can be fixed in space and to constraint the other component to the faces of the frame then to start from the heart and work my way to outside. I completely understand the point, but I will not buy a single part before creating everything in CAD, so I can calculate the exact costs beforehand and order all the individual parts cut to lengths.

I hope to put all the electronics and the power supply at the bottom of the device, bottom 5-6cm, exactly under the build plate. That should allow enough space for servicing when the build plate is all the way up. Cables would go up in one of the back corners, and down from the top to the extruder, where exactly will I know when I have all other parts in the design. The enclosure should be simple enough, screwed plates (2-3mm acrylic or 4mm ply) to the t-nuts in the rails, because of the weight of of the plates the t-nuts can be printed, front and upper plate with hinge on one side and some kind of lock on the other. When I start the drawing I will post the results here, probably with some other questions which may arise during that endeavor.

That is a very good point about the heater, I will look and choose the exact dimensions of the build plate first, and then calculate the outside frame dimensions as 40-50mm offset.

Should I assume that 12mm steel / 16mm aluminum would be enough then? At these dimensions is aluminum even better material because of the weight difference.

I remember reading about Smothieboard, it has nice wiki with a lot of information, Duet seemed much more confusing (but I'll take another look), the Replicape is too pricey, but it may be worth it if silent motor control is more then a sales pitch.

Rather than design first and shop second, I suggest accumulating the parts and designing around them as you acquire them. That lets you take advantage of cheap deals when they're available and keep overall cost lower. Example, I buy used linear guides via ebay. All you need to know to make your selection is the minimum length of the rails and the number of bearing blocks your design will use. Rails can easily be cut to the needed length. I can usually find deals on good Japanese made rails for $40-75 each depending on length and number of bearing blocks. Occasionally NOS rails show up in that price range.

I wouldn't just calculate the outer dimension as 40-50 mm larger than the bed surface. Model key components first as you accumulate the parts and figure out how they're all going to fit together. That will give you your final frame dimensions.

Electronics in the bottom is fine, but you're assuming that you'll be able to raise the bed if the controller isn't working, which may be the very reason why you need to access the controller! If it isn't working, will you be manually turning the screw to lift the bed high enough to get to the controller? You'd probably tire of that process very quickly!

In my last printer I built the electronics into a drawer at the bottom of the machine.



Accessing the controller and all the rest of the electronics is extremely easy. In my new build I am probably going to mount the electronics on a fold-down lid that will be hinged to the printers frame. Since the machine is tall, I'll put the electronics up high where I can reach it easily without having to bend over or crawl on the floor. Most of the cabling has to connect to things (XY motors, extruder, fans, limit switches, etc.) near the top of the machine, so that will keep cable lengths pretty short, too.

Edited 1 time(s). Last edit at 02/03/2017 10:49AM by the_digital_dentist.

---

Ultra MegaMax Dominator 3D printer: [drmrehorst.blogspot.com]

Electronics

What would interface to Raspberry PI? Some kind of special interface that must be Arduino?

Most people use USB interface to printer controller or SD card

Will the Pi be doing slicing and CAM?

Consider:
RE-ARM is a 100Mhz, 32-bit ARM controller with the popular Arduino MEGA footprint.
It's a plug-in replacement for the Arduino Mega in your RAMPS setup
for a quick and easy upgrade to 32-bit goodness.
you would use smoothieware firmware

[www.panucatt.com]

The idea with a drawer is very nice, I did not think about that option.

I know the ebay possibility, but it's much more unusable from my standpoint. I am in Europe outside EU (Serbia), so the shipping from US is mostly very expensive. The shipping from EU is sometimes at a reasonable level, but there come the import taxes on everything that is new or looks like new and import denial for every item that has a commercial quantity (most of the time it's three of a kind or five of a kind of same articles). So my options are a local market which is very underdeveloped, shipping from China which is not opened if it comes in envelope. Shipping from EU/US could be ok for the electronics or the extruder, but it would not be available for pillow blocks and linear rails where you have six shafts and 12 pillow blocks, Shaft can be shipped as two 2m shafts, but the pillow blocks would be "a clearly commercial quantity and they would be denied". The options that I have are, a local distributor which is mostly expensive, as an example Item24 1m 20x20 is 12eur, right angle for the same profile without screws is 6eur. So, one of my options is to have all the machine elements calculated and to order the parts from Motedis or similar to a friend who can bring it by car or to make a pickup order form a company where I can drive by when traveling and pick the items up. So, it's all a bit confusing. 3D filament and quatcopters are even more complicated. 3D filament can be denied, because it's not 1kg filament, but a large roll of some kind of plastic wire (clearly commercial quantity ), on one occasion they opened the vacuum bag, so the seller when returned form the customs did not accept the return because of the humidity entering the roll, but enough of that, all of that get's on my nerves and has nothing to do with this topic.

@cozmicray:

The only thing that I have experience with by now is Malyan Melzi, which is Arduino, so I have been reading the Marlin firmware and all the configuration options look contained to one file and simple enough. I could nowhere find an information is Marlin available for the ARM based boards, but I assume that it's not compatible, so that is the only reason for wanting an Arduino based board. I currently use SD card, but do like the option with Octoprint and Raspberry PI, so that's a goal for the next machine (and SD card as fallback option). I would probably still use Cura on my pc and send the gcode files to the Octoprint through the web interface. For the controller I do not have constraints which would need an Arduino, I just thought that I have no need for 32bit electronics, but if there is a reason for going with 32bit electronics who am I to argue. I see the Smothie has it's own configuration platform which seems well documented. so it may be the right option.

I'll try to make a few sketches during the weekend.

While researching I found out that I was mistaken to think that there is no need for 32bit platform inside a Cartesian type printer. I have found an excellent topic discussing it at StrackExchange. So I concluded that it may practical to go 32bit.

The proposition to look at the heater elements was excellent, so I will size down my initial idea. The options that I think about at the moment are 300x200 or 300x300 build plate with silicone based heater, there is also a 400x400 option but it is much more expensive. After thinking thoroughly I came to the conclusion that I never had problem with the height of the build zone, as much as with length and width, so I am thinking to lower the maximum height to no more then 300mm. There should be a heat barrier between the electronics compartment and build chamber and maybe another heat barrier on underside of the silicone heater (have not yet found a reason why are people adding that barrier under the build plate heater).

While researching the GT2 pulleys and shafts I did not find a GT2 pulley with a bore bigger then 12mm. So, the shafts will be 12mm, on plain bearings.

Here's why people put insulation under their print beds:

1) Because their heater is underpowered and unable to get the bed up to print temperature without it. They trade faster heat up for slower cool down.

2) To protect electronics that may be under the bed from radiated heat.

Here are the reasons people put glass on top of their print beds:

1) They want to be able to quickly switch plates so they can start the next print without waiting for the machine to cool down. This is good for about a week, until they're sick of printing yoda heads and tugboats...

2) They like to experiment with different surfaces: Some people will put kapton tape on one glass plate, PEI on another, and some other magical stuff on the third, fourth, and fifth plates, depending on whether they get better first layer adhesion with whatever material they are printing. Changing plates and print surfaces may require releveling and zeroing with each change. Pulling ABS prints out of a warm printer instead of letting it cool slowly may result in checking and delamination.

But the real reason most people use glass on their beds, whether they know it or not, is because the aluminum "heat spreader" that comes on most printers isn't flat enough to print on thanks to its thinness (i.e. cheapness) and the 4 point "leveling" system that bends it and the support structure under it. It's very hard to get prints to stick to an unflat/unlevel surface. By clamping a relatively stiff piece of glass to the bed they can approximate a flat, level surface. Glass is an insulator and will be warm where it touches the heated plate and will be cool where it doesn't, so clamping glass to an unflat surface trades an unflat/unlevel surface for an unevenly heated surface, which can also be difficult to get prints to stick to.

Use a flat, rigid, thermally conductive bed (1/4" thick or more cast aluminum tooling plate) sitting on a 3 point leveling system (that actually levels it without bending it), give it adequate heat (about 0.5W per cm^2), and a surface that prints like to stick to (kapton, PEI, etc.) and you'll have very few first layer adhesion problems.

---

Ultra MegaMax Dominator 3D printer: [drmrehorst.blogspot.com]

DD and I like to differ about some aspects of print beds. We both agree that a rigid, flat aluminium plate is essential. We differ on the use of removable print surfaces. I prefer to use glass (on top of my 10mm thick aluminium) for the reasons DD stated (i.e different surfaces for different filaments and being able to quickly resume another print after one has finished). - Oh and I've never yet printed a Yoda head but I do often print a number of items consecutively. With the right sensor for Z homing, there is no need to make any adjustments between changing print surfaces. I have my glass clamped with aluminium channel along the length in Y which is held with wing bolts. So I often simply slacken the bolts and slide the glass forward at the end of a print, and nip the bolts up again, rather than completely remove it. The allows the glass to cool much quicker than if it was left on the bed. The other thing about having a removable (and cheap) print surface is that it will protect the expensive aluminium plate in the event of something accidentally being dropped on it. DD also used to say that glass is an insulator so will sow down the warm up time. Having done some practical tests, I can say that in use, wity the bed heated to 60 degrees C the surface temperature 6mm thick float glass is 58.5 degrees C (1.5 degrees C cooler) so hardly an issue.

DD and I also differ on the use of insulation under the bed. It will prevent heat loss thus enabling the bed to heat up more quickly and save energy as well. I prefer my 800W silicone heater to heat the print bed, not the room that the printer is in.

I print a lot of ABS, so heating the enclosure with the bed heater helps keep print quality high without adding a separate enclosure heater to the machine. Operating cost depends on where you live and how much power costs. In the US electric power is pretty cheap (avg about $0.15/kWh) so operating a 3D printer with a heated bed isn't a big expense if you can afford the printer, you can afford the electricity to operate it. Bed heaters operate at full power for a few minutes when they are running up to print temperature, but then the power is cut back by 50-75% once they have reached it (even mine, with uninsulated bottom surface). I think it's interesting that so many people worry about the cost of electric power to run their 3D printer but never consider the cost of powering their big screen TV for 6-8 hours to watch a football game. In the US, if the electric power operating cost is a concern, maybe 3D printing is just too expensive for you and it would be better to find another hobby.

I don't know where the myth of "delicate" aluminum originated. Have you ever seen a broken aluminum plate? OTOH, think about the fun of cleaning glass shards out of the bearings in your printer (or off the floor, etc.) if you drop a glass plate, a tool, filament spool, etc. It's possible to scratch an aluminum surface (you can scratch glass, too), but I can't see how that's going to happen unless you actually try to scratch through the PEI or kapton print surface. If you did manage to scratch it, it won't matter because you're going to be replacing the ruined cover material before the next print anyway. Whatever you'd have to do to scratch aluminum would either scratch or break glass.

The fast plate change idea is nice if you have a business and need to keep the machine running to earn your income, but it really isn't applicable to most hobbyists. If you're running a business that has so much demand that you have to keep swapping plates, it's time to invest in a second printer. If you're printing ABS or PC, slow cool down is beneficial for print quality. Pulling a warm print out of a warm enclosure as soon as it finishes is asking for it to warp and delaminate.

If you only print PLA or other materials that don't benefit from a warm enclosure, swapping plates is a good way to speed things up. If I were doing that I'd look into swapping aluminum plates instead of glass, but I live near Howard Precision Metals where cast tooling plate cut-offs sell for only $2.50 per lb. A 12" x12" x1/4" plate is about $10 worth of aluminum, not much more than glass, but it's never the exact size you need, so you have to buy a bigger plate and cut it. Not everyone has a place like that nearby, and not everyone can cut the material to the size they need, so for most people glass is ultimately easier and cheaper.

---

Ultra MegaMax Dominator 3D printer: [drmrehorst.blogspot.com]

We'll agree to differ. I'm happy to respect your opinion but that doesn't mean it's the correct solution for everyone, or that it is the only solution that should be adopted.

Where I live, aluminium tooling plate is expensive whilst glass is cheap. I haven't seen a dented piece of glass but I've see plenty of scratched and dented pieces of aluminium that would not serve as print surfaces (even if they started life flat). I don't worry about the cost of running my printer (and I do own a 50inch plasma TV) but electricity is expensive here so if one can save some, then why not? Why waste electricity just because you can? I do make things and sell them to supplement my income, so being able to resume printing quickly is desirable (as is keeping the production costs low). A second printer would be nice but I'm limited on space and a sheet of glass is somewhat less expensive than another printer.

So, those are the reasons why I choose to use a removable glass print surface. It works for me but I appreciate that others may have different views and/or live in different parts of the world where circumstances may be different. I'm prepared to admit that others may have good reasons to do things differently to me.

i also am doing a larger build with extrusion i went with 3030 extrusion.
18" x 18" heated bed(seperate pus for that)
i will be using the duet bard and touch screen
dual extruder

just trying to find nice acrylic to wrap it all in.
that stuff is exspensive in 24 x24 sheets

MinskMade

Quote
the_digital_dentist

2) They like to experiment with different surfaces: Some people will put kapton tape on one glass plate, PEI on another, and some other magical stuff on the third, fourth, and fifth plates, depending on whether they get better first layer adhesion with whatever material they are printing. Changing plates and print surfaces may require releveling and zeroing with each change. Pulling ABS prints out of a warm printer instead of letting it cool slowly may result in checking and delamination.

y'know... i always like reading your stuff, dd, because it's clear, informative and to the point. since using borosilicate glass (for which i had to get *really* strong perfume-free lacquer to get the 260x12x15mm really long parts i was doing to actually stick... yes, really: 260mm long parts on a 200x200 printbed... across the diagonal), i've discovered printbite and my god does it do the job and take all the mess and hassle away.

as i'm still learning that it's really not ok to use chinese cheap-and-cheerful stuff, i have a MK2 (which is bowed out boinggg boinggg) printbed underneath a 3mm alu plate (yes, i know, why the hell didn't anycubic just sell me a MK3 alu buildplate??) but *sigh* cardboard rolled up under it pushes the centre up, the 3mm alu plate heats up uniformly just fine and the printbite makes damn sure that the prints don't come off until the bed cools down. there's this audible "crink" if you leave it long enough and the part just pops off.

deckingman: i really value your experience in printing multiple parts. i'm going to have to be doing that, soon: i have a design which requires 35 separate parts and i'll need to do 150 of them. even if i do the 12 sets which are for the "early adopter" customers that's still one hell of a lot of printing, so hearing about your use of glass plates and keeping the build plate up to temperature is really handy.

Quote
SaleB
6. electronics. I have not been able to find if ARM boards do support Marlin firmware and does Marlin support this kind of machine. I would like it to be a 2560 Arduino based board, as I do want to pair it with a Raspberry I do not need all the functions that faster boards have. After the initial search I found a few boards interesting, but no reviews on any of them. The options that awoke my interest are MakerBase MKS 1.4, Megatronics 3.0 and Rumba. Essentially the board should have a few removable drivers and enough of them for at least two extruders, expansion connectors for card slot and graphic display. I like Rumba at most, but could not really tell why. Is there any real difference between the Rumba from various chinese sellers, highest price I found at a place called ReprapDiscount (with a bull head logo on back side), and for half of that price at a place called BigTreeTech (BIQU logo on the back side) , can there be any real difference between two chinese sellers? The only differences that I saw are blue and green screw terminals and heat sinks on the BIQU board against the black screw terminals and capton tape on the usb connector without the heat sinks on the RD board, both are clean and well soldered, and also one coil has different value between the two beards.

ok please for god's sake don't use removable drivers. i've mentioned this any number of times but the history behind the original sanguilino-based reprap printer board for the very first mendel was that they used *prototyping* boards supplied by the manufacturer, which you see on those little breadboards. i know someone who actually had one: he showed it to me and i now have it. in actual use they STOPPED WORKING half way through a print because they got so hot. the datasheet CLEARLY STATES that the driver ICs use the exposed pad to wick heat away, and that there is a CERAMIC INSULATOR on the top, and that the *PCB* is supposed to be designed to remove the heat. now, the problem with the insulator being on the top of those ICs is that no amount of heatsinks or fans will help you to stop them from overheating and failing during a printrun. those tiny PCBs are *nowhere near* big enough for the heat to properly dissipate AS THE IC IS DESIGNED AND INTENDED TO BE USED BY THE MANUFACTURER.

so please.

for god's sake.

would people *please* stop designing and selling polulu and polulu "compatible" electronics, and please stop recommending them.

the MKS SBase on the other hand looks to be really good. the company - despite being chinese - also looks like it's actually honouring the GPL software license, and has a github account. i'm amazed.

[www.aliexpress.com]

so that's an MKS Sbase 1.3, with an LPC 32-bit ARM processor, built-in ethernet, DRV8825 (2.5A max) steppers... all for $USD 45, wow! DRV8825s are a TSOP-28 so the legs take the heat out through the PCB: they've got a whole bank of heatsinks on top of them though, which will help. the board also goes up to 24V so if you ever want to do 24V electronics all's good there.

[www.aliexpress.com]

this one's the MKS-BASE-L V1.0, $19!! wow! and it uses the TSOP-28 variant of the A4982 (2A max) so that's really really good as well. the QFN variant can only really handle up to 1.3A (which is why you always always get the overheating problems on those "prototyping" style polulu boards).

i'd never heard of MKS before, so thank you! i might have to get one of the SBase 100mhz ARM boards to try it out (i'm in taiwan so that would be quite quick and easy to do).


Rumba: i got one, i tried it out, it actually had an electronics design fault (in the ADC circuits for the thermistors), which took me ages to resolve. the guy did give me a full refund, bless him, but it cost me a lot of time and investigation. not recommended, basically.

Megatronics 3: polulu-style "prototyping" driver boards. don't do it. just.... don't.


generally these days you really should be getting an ARM-based board, because Arduino's, at only 16mhz, can barely cope with driving PWM steppers above around 100khz. that sets a hard limit on the maximum move speed (i forget exactly what that is, apologies).

i say that despite just having got a Melzi 2.0 only last month (because i'd never found a chinese printer driver board manufacturer that honoured the GPL... until today! really, i'm deeply grateful for you making me aware of MKS!) because the Melzi 2.0 is cheap, cheerful, and i know it works.

also i've tried the Duet 0.8.5 and i love it: built-in Ethernet, built-in Web interface, just absolutely awesome, and you can even edit the configuration *from* the web interface. deeply impressed with it, and the price now is not unreasonable for the chinese clones. the DuetWifi... awesomely-awesome, monster drivers, extensible... aaand just too costly. if the electronics cost *as much as a 3D printer clone from china*, there's something wrong.


basically you should be evaluating which stepper motors you want (dc42's advice is extremely good here) - which is basically "use 900 steps/rev motors and run them at 24v", and then make sure that you get a board which has *matched* drivers for the rated current of the stepper motors. i remember reading dc42's advice about this over a year ago and he stressed that the current matching is really quite important. getting 2A high torque stepper motors and a board with only 1.3A drivers will get you into trouble, for example: the lower resistance will cause the underpowered drivers to overheat.

in short: there is a *lot* that you'll need to learn. you've picked an extremely rewarding challenge for yourself, and you're in the right place. this forum has a lot of really experienced people who are happy to help, you can tell very quickly who they are. stick at it, be patient- keep in touch: you'll do ok

Edited 2 time(s). Last edit at 05/13/2017 04:03PM by lkcl.

A couple things to know about the MKS Sbase boards:

1) DRV8825 drivers require some simple modification for them to work properly otherwise they will skip steps when moving slowly. See: [www.engineerination.com]#! 3D printers are not the right application for those driver chips, which should tell you something about the engineering that went into any 3D printer controller/driver module that uses them.

2) When you buy the MKS Sbase boards you don't get any support. The board runs smoothieware which is open source, but the MKS folks don't contribute to the project at all so the people who designed the smoothieboard and wrote smoothieware are not enthusiastic about helping people set up and debug their MKS Sbase boards. They'll help, but only after you try (and fail) to get help from MKS. See [smoothieware.org][]=mks

---

Ultra MegaMax Dominator 3D printer: [drmrehorst.blogspot.com]

Quote
the_digital_dentist
A couple things to know about the MKS Sbase boards:

1) DRV8825 drivers require some simple modification for them to work properly otherwise they will skip steps when moving slowly. See: [www.engineerination.com]#! 3D printers are not the right application for those driver chips, which should tell you something about the engineering that went into any 3D printer controller/driver module that uses them.

*splutter* that's.. that's a truly dreadful well-researched hard-hack: two diodes to lose 1.4v across each so that you get the proper sine-wave characteristics, eurghh um thank you dd, i will not be using DRV8825-based boards. once again i'm really appreciative of your depth of knowledge and willingness to share it, saving me time, money and aggravation in the process.

Quote

2) When you buy the MKS Sbase boards you don't get any support. The board runs smoothieware which is open source, but the MKS folks don't contribute to the project at all so the people who designed the smoothieboard and wrote smoothieware are not enthusiastic about helping people set up and debug their MKS Sbase boards. They'll help, but only after you try (and fail) to get help from MKS. See [smoothieware.org][]=mks

y'know... i realise it's software libre but come on, this is china. the number of chinese hardware companies that have actually released full source code under the GPL software license as you are supposed to do - without requiring a court case or threats from customers in literally the thousands or requiring the FSF or the Copyright holders to go after them - you can count on the fingers of one hand. like... in *twenty years* of working with free software as a software libre advocate and developer, MKS is *literally* the first chinese hardware company that i've ever heard of that's fulfilled its copyright obligations voluntarily and without threats.

so i am therefore a bit puzzled by the smoothieware team's reactions. chinese companies that don't violate software licenses should be cherised, encouraged and supported for the rarity that they are. *confused*

Should the smoothieware/smoothieboard developers be answering questions about print problems caused by DRV8825 drivers? I think they have enough work to do to support the stuff they produce and to produce more and new stuff. The smoothieware folks were nice enough to open source their work. It doesn't mean they have to provide free tech support for every company that chooses to use their stuff.

---

Ultra MegaMax Dominator 3D printer: [drmrehorst.blogspot.com]

Quote
the_digital_dentist
Should the smoothieware/smoothieboard developers be answering questions about print problems caused by DRV8825 drivers? I think they have enough work to do to support the stuff they produce and to produce more and new stuff. The smoothieware folks were nice enough to open source their work. It doesn't mean they have to provide free tech support for every company that chooses to use their stuff.

y'know, that's a *really* good question, that entirely depends on your perspective and life-experience. if you've never worked with software libre, you're just a hardware person, doing CNC, 3D printing and so on, you just want to buy stuff that works well and allows you to do what you want, the answer's "no of course not!". if on the other hand like myself and many others who have dedicated getting on for two decades of our lives to working with software libre, saving businesses world-wide hundreds of millions if not billions of dollars in proprietary software license fees and being f*****d over by arsehole criminals stealing the source code *and* not receiving a damn cent from the very businesses whose money you've saved, with us in some cases ending up $45k or greater in debt then the answer's obviously much more complex and certainly less straightforward to answer either yes _or_ no.

personally i got into hardware design, from a software libre background, *precisely* because people can rip off software libre source code much more easily than they can rip off a hardware design, but even in getting into the hardware design and manufacturing business i am not going to abandon the ethical roots from which i started.

so yeah, it's genuinely a really good question, dd. anyway, back to 3D printing...

Real life took precedence for a few months, but I am still actively reading all the comments.

Quote
lkcl

ok please for god's sake don't use removable drivers. i've mentioned this any number of times but the history behind the original sanguilino-based reprap printer board for the very first mendel was that they used *prototyping* boards supplied by the manufacturer, which you see on those little breadboards. i know someone who actually had one: he showed it to me and i now have it. in actual use they STOPPED WORKING half way through a print because they got so hot. the datasheet CLEARLY STATES that the driver ICs use the exposed pad to wick heat away, and that there is a CERAMIC INSULATOR on the top, and that the *PCB* is supposed to be designed to remove the heat. now, the problem with the insulator being on the top of those ICs is that no amount of heatsinks or fans will help you to stop them from overheating and failing during a printrun. those tiny PCBs are *nowhere near* big enough for the heat to properly dissipate AS THE IC IS DESIGNED AND INTENDED TO BE USED BY THE MANUFACTURER.

so please.

for god's sake.

would people *please* stop designing and selling polulu and polulu "compatible" electronics, and please stop recommending them.

I did not know that.

My safe bet was Azteeg, but this new information disqualifies most of their products.

Quote
the_digital_dentist
1) DRV8825 drivers require some simple modification for them to work properly otherwise they will skip steps when moving slowly. See: [www.engineerination.com]#! 3D printers are not the right application for those driver chips, which should tell you something about the engineering that went into any 3D printer controller/driver module that uses them.

Taking the this into account there are only a few controller boards left, both Duets, Replicabe, Smothieboard, but non of them cheap. And all of that makes much better case for a Duet.

In the mean time a have found a few more controllers, one of them being Aligator board which may be abandoned project or still a work in progress and something new that I have found in last few days.

Another Chinese product, a Pibot uses a new approach, independent motor drivers which may or may not be an advantage. It does not use the DRV8825's, but rather Toshiba 6600. It can handle 24V, and has more powerful processor.

Edited 1 time(s). Last edit at 05/16/2017 04:52AM by SaleB.

Quote
SaleB
I did not know that.

My safe bet was Azteeg, but this new information disqualifies most of their products.

i found out only yesterday that reprappro - the business that adrian started when he began the 3d printing revolution - has sadly had to shut down. why? because there is *so much* out there - most of it not really very good - that they decided to quit whilst they were still in the black. there's a huge amount that people don't know - me included, we're all learning, but i find it quite sad that we can't get our (plural) act together and financially support the people who provide the best information and hardware.

[www.panucatt.com]

all of these are polulu-based (prototyping boards), only one of them is 32-bit. don't do it

Quote

Quote
the_digital_dentist
1) DRV8825 drivers require some simple modification for them to work properly otherwise they will skip steps when moving slowly. See: [www.engineerination.com]#! 3D printers are not the right application for those driver chips, which should tell you something about the engineering that went into any 3D printer controller/driver module that uses them.

Taking the this into account there are only a few controller boards left, both Duets, Replicabe, Smothieboard, but non of them cheap. And all of that makes much better case for a Duet.

yehyeh. you should still be able to find the duet 0.6 if you're on a budget, but do consider the 0.8.5, it's damn good. uses 2A A4982 drivers. the smoothieboard... also looks pretty good, again 2A drivers, again a 32-bit controller.

both these boards (a) respect the open hardware community spirit (b) are 32-bit and over 80mhz so will be able to do decent PWM rates (run really fast basically) (c) use 2A drivers so will work great with higher-torque motors if you want to use them

Quote

In the mean time a have found a few more controllers, one of them being Aligator board which may be abandoned project or still a work in progress and something new that I have found in last few days.

[www.reprap.org]

not an open hardware product. they've released some PDFs but it's not in the spirit of the open reprap movement. basically they're spongeing off of everybody's expertise.

Quote

Another Chinese product, a Pibot uses a new approach, independent motor drivers which may or may not be an advantage. It does not use the DRV8825's, but rather Toshiba 6600. It can handle 24V, and has more powerful processor.

oo nice, you notice it's a modular design, but the driver boards with the toshiba 6600s are *much* bigger? that means they will have decent heat dissipation, using the PCB to wick heat away from those high-current chips. that's how it's *supposed* to be done.

holeee cr** 4.5A output, zoweeee! that's really impressive you'd reasily be able to drive high-torque 900step/rev NEMA17s and even consider NEMA23 motors with those. _wow_.

taking a brief look at their page, they do have software downloads - aiyaaa they recommend slic3r, don't use slic3r, i can send you a part to try out if you want to know why (slic3r makes an absolute dog's dinner of it)... do make sure you investigate the software *properly*, make *absolutely* sure it's got full source. if you get *any* problems you want there to be an active community of developers around the board, not just the one person who really isn't well-known.


so, anyway, the summary is, yes you can buy a cheap china RAMPS clone just to get things up and running and tested, but i guarantee you'll regret it and should *only* get one knowing that you are going to replace it. it'll stall in the middle of prints if you go over around 100mm/sec due to overheating, and the extruder will REGULARLY skip (making this sort-of "tkk, tkk" sound) if you try a 0.3mm layer height at only 50mm/sec for example because there will not be enough current to drive it. also it will be extremely unlikely that you'll be able to upgrade the firmware in order to fix any problems... because it'll almost certainly be Copyright violating (illegal) firmware.

orr... you can buy something that's risk-free and supports the people who brought you - and sustain - the reprap community spirit. that will *appear* to cost you with a larger initial financial outlay but in reality long-term it will save you a lot of time, hassle and money. really that means a duet or perhaps a smoothieboard. as dc42 is on here a lot and is incredibly helpful i lean in favour of the duet (and bought one last year).