Uncle Bob's Build

Uncle Bob's Build

Postby uncle_bob » Wed Jan 01, 2014 3:09 am

This is a log of my "sort of a LISA" build. It's starting as a dupe of a lot of the info over on Rep Rap. Feel free to hop in if you have anything to add or to ask about.
Last edited by uncle_bob on Mon Jan 27, 2014 2:42 am, edited 1 time in total.
uncle_bob
 
Posts: 8
Joined: Tue Dec 31, 2013 7:20 pm

Re: Uncle Bob's Build

Postby uncle_bob » Wed Jan 01, 2014 3:41 pm

Basic idea as a "Mega LISA". Spacing between the drive rods of roughly 3' and rods that are 5' long. The build area should be roughly 16" (400 mm) diameter by 32" (800 mm) tall.

Thread rod cost is a bit of a concern in this tall a design. Roton HiLead is cheap at 3/4" diameter so 3/4" with a 0.5" lead is what I'll be using.

http://www.roton.com/Mating_Components. ... ly=7061226

Since the rod is heavy and not very fast (0.5" vs 1") NEMA-23 steppers at ~ 400 oz-in make sense. Coming up with enough current (and voltage) to drive them gets you into outboard controllers. The Wanti WT57STH115-4204A motors and DQ542MA drivers appear to be a reasonable choice.

http://www.ebay.com/itm/140902534528?ss ... 1439.l2649

Bearings for the shafts will be 6202-2RS-12. Selected purely on price:

http://www.vxb.com/page/bearings/PROD/1 ... OgodnEUAaQ

The main structure will probably be 3/4" plywood for the horizontal plates and tube / pipe for the vertical stuff.

Arm design is a bit up in the air. The arms will be about 24" long. That's bigger than I can print. The choices are down to either some sort of rectangular aluminum tube with printed ends or an aluminum bar with bearings directly mounted in the bar.

Drive nuts will be Roton flanged nuts. The end clamps may be the bronze Roton nuts with set screws.

No drawings at this point. No printed part designs yet.

Cost so far:

$200 for the motors and drivers (delivered)
$100 for various bearings
$500 (pre shipping) for the Roton stuff.

(more to come as things move along)
----------------------------------------------------

At least by my calculations, the 3/4" thread rod should be equal to a standard design with 7/16" rod when it gets to 4.28'. It should be equivalent to a 3/8" design in stiffness at 6'. My *guess* is that the captivated portions of the rod will be a bit longer in my build than on the standard. The "free to swing" region of the rod is what counts.

Speed wise, with 36V on the drivers and running the motors and rod above:

90 mm/ sec is 7.08 rev/sec or 425 rpm
Drive frequency at speed would be 354 Hz
Zero to peak voltage would be 35.7V
To hit 128 micro steps, you need a step pulse rate of 181 KHz

Similar calculations would get you 120mm/sec at 48V. Step pulse rate 241 KHz. The DQ542's may or may not really be happy long term with 48V.

Those are speeds on the rods, not head speeds. It's not clear that you need 128 micro steps or that a Marlin can deliver >150 KHz step rates.

These speeds aren't going to get me super fast prints. It's a big printer, so time is indeed a factor. My *guess* is that printing large objects at 200 mm / sec is not a real good idea...

----------------------

With the screws I will be using the "push it to the top" approach for zero probably will not work. This is *not* going to be a light weight delta. Adding end stop switches isn't going to increase the cost / size / excitement by a noticeable amount. Getting them to work, well that may take some effort.

Adding a bed leveling Z probe is a likely thing. I have not done any modeling at all in that direction yet. Any print I'd do on this will take a while. Doing the level data and then feeding it back into the Python code is fine at this point.

-------------------

The firmware will start out as Marlin running on a Mega and a Ramps board. Ultimately it will morph into something else. Likely candidates are an ARM or a (cheap) dedicated PC. I'm already running enough power into the printer that a 10 or 15W bump up in the power budget for a PC isn't a big deal.

-----------------

I've decided on 4.5" diameter aluminum pipe for the vertical supports on the printer. Even with 0.125" thick walls it's pretty rigid stuff. The price is right ....

Solid aluminum bar looks like the leader in the materials race on the arms. A piece of 3/4" x 2" bar and some 608ZZ flanged bearings is what I'll probably try first.

-----------------

Some dimensions (after measuring doors and the like):

Thread rod mounting diameter 28"
Support pipe mounting diameter 32.5"
Thread rod length 60"
Arm length 18" (est)
Platform width 31.4"
Will fit through door 27"
Height < 6'

-----------------

It would appear that

http://www.mcmaster.com/#35765k99/=q4j491

At roughly 1KW and 12" diameter is the low cost leader for the bed heater.

The power is based on using:

http://www.ebay.com/itm/110859764005?ss ... 1439.l2649

as the build plate / heated bed. Build circle looks like it's right at 18".

Footprint I started with:

Image

After some playing:

Image

Current best guess at a footprint:

Image

Red = supports
Blue = drive shafts with clearance
Green = build plate
Black = circular build area

-------------

One source for plywood up to 1.5"

http://www.buyappleply.com

Looks like good old Home Depot 3/4" birch is the most likely plywood. Apparently the Columbia Forest Products PureBond stuff is pretty good.

-------------

After looking at the amount of play in the plastic flanged nuts on the 3/4" shafts, I think there is a need for fairly good anti-backalsh / centering / zero play enhancements. Best guess right now is to mount a second flanged plastic nut below the first and put some springs

http://www.mcmaster.com/#9657k105/=q7mtb6

nuts, and bolts between them. The easy way to do it would be to sacrifice some build height and put the nuts above and below the take off for the arm. If the upper nut travels into the plywood the loss may not be much at all.
uncle_bob
 
Posts: 8
Joined: Tue Dec 31, 2013 7:20 pm

Re: Uncle Bob's Build

Postby uncle_bob » Sat Jan 25, 2014 2:42 pm

After some digging and gyrations, the steppers above probably are not ideal for a 600 rpm sort of application. The higher speed motors seem to be a bit of a specialty item. The:

KL23H276-30-8A
KL23H2100-30-4B
KL23H251-24-8B

All seem to have data graphs indicating they will produce some output at speed. My guess is that 0.5 to 1.0 N-M is a good range to be in to achieve reasonable acceleration. Below that, you will never get to high RPM's. Above that you probably would do better with a DC servo setup.

The middle motor on the list is fairly high inductance. Driving it to high RPM's would be 2 to 3X harder than the other two. It's got the most torque of the three. The first one has a good combination of torque / speed and drive. The third one is cheapest by a wide margin.

All three motors above drive fine with a 4.2A dedicated driver running at about 36 volts. Going up to a higher current driver would bring in the KL23H2100-50-4B. It's got major torque below about 400 rpm. The higher current drivers also are higher voltage.
uncle_bob
 
Posts: 8
Joined: Tue Dec 31, 2013 7:20 pm

Re: Uncle Bob's Build

Postby uncle_bob » Sun Jan 26, 2014 6:28 pm

The vertical axis has some issues at the moment.

The aluminum build plate is > 1" thick. Once I add the McMaster Carr 1.1 KW heater it's up around 1.5". An inch of insulation plus and air gap would get it up into the 3" vicinity. From the plate plots above, there's no room to sling this all below the mid plate. The hole in the plate would be larger than the plate it's self. A larger plate doesn't fit through my magic door. I also have some doubts about a 3" section of plywood plate begin rigid enough to keep everything in X/Y registration.

Next up is the whole speed / screw travel distance tradeoff. If you go for full travel on the screws, the arms go horizontal. As you get to this point, your X/Y resolution gets really good. Your travel speed goes down. There's a lot of details you can fiddle. I think I'm going to stay 4 to 6" (in Z) away from having the arms horizontal. This increases the arm length and I have a build height about 2" shorter than I would with horizontal arms.

Another minor detail is the length and profile of the hot end. I'd like to allow for a fan on the hot end. I also have some that are a bit longer than others. If arm / hub joints are almost on the build plate you get maximum build height. You also have joints that are right on the heated bed. Optimizing all that for a single hot end seems like overkill. That's another couple of inches of build gone. (call it 3")

All these things drive towards less build height, longer drive screws, or raising the end of the drive screw above the mid plate. I'm "over length" on the 3/4" screws as it is (5' vs 4.5') so going to 6' is not real attractive. It also hits that door issue on the Z axis (6' 8" door).

More or less, if the arms are 6" above the mid plate, the only bump I need to worry about is the fan on the hot end below that. The motor body is about 5". Motor plus shaft plus coupling plus bearing plus plate above bearing would get to about 9". It could be 12" and still not restrict travel.

Since the nuts on my lead screws are not zero thickness, they need to be worked into all this. My guess is that they will take up at least 5".

Build wise, I now have the hub arms bumping into the bearing mount plate / coupler / support structure. That's got to be a bigger diameter than the shaft alone. If it adds an inch or so, I can move the shafts out by that amount and still fit through the door. That *assumes* that the mount structure and the arms fits inside the 8" diameter circle I've allowed for clearance around the lead screws.

The structure would have to be fairly rigid if the screw isn't going to push it around. 3/4" thread rod is cheap enough that using three or four pieces in the support would not be out of the question.

If the motors and build plate are above the mid plate, then there really isn't anything below it. You have to mount the electronics somewhere. There's no real need for a bottom plate other than to keep out the dust and dirt. As long as the feet (or casters) are supported with something, the bottom plate could be 1/4" plywood. My guess is that none of the electronics are > 2" thick.

Flipping the printer on it's side to get at the electronics on the bottom may be an issue. This beast is going to weigh well over 100 pounds. Moving some of them to above the mid plate would be nice.

-------

If all that's a bit hazy, well that's the way it is at the moment. A lot depends on just how complicated the mounts get / how massive they need to be. I'm hoping an aluminum plate(s) with some holes in it plus some thread rod does the job. Getting the three lead screws equidistant from each other is pretty important.

If you go back to the top of this thread, the objective is > 32" build height. Some of the stuff above gets up into the > 39" range. I'm willing to trade off anything over 35" (objective + 10%) to optimize in other areas. The build envelope is flat enough that at 35 in the center, I probably can get > 32" over the build area.

--------

Here's what it looks like at the moment
Image

It still fits through the door and there's enough room to at least mount the drivers above the mid plate. The power supplies would go on top of the top plate in this case. Wires would travel inside the support pipes.
uncle_bob
 
Posts: 8
Joined: Tue Dec 31, 2013 7:20 pm


Return to LISA Simpson

Who is online

Users browsing this forum: No registered users and 2 guests

cron