UPS for 3D Printing
The three-dimensional printers see a lot of run time here at KautzCraft Studio. The larger prints run for twelve hours or longer. A lot can happen in that time span. Here in Texas and I am sure everywhere else, the weather can change for the worse.
That can lead to power interruptions which can kill a print job in the eleventh hour. It has happened often enough that it became a high priority to find a solution.
Many of the newer three-dimensional printers now offer power interruption protection and can be restarted where they are stopped mid-print and resume the print. That’s not always a perfect solution as the printhead is not parked in a “safe” area above the print and can actually “glue” itself to the print as the heat goes away.
The “pause and resume” works well, but when the power is instantly gone as in a power outage, the pause move (to a safe location) cannot be executed.
The obvious solution is a device called an Uninterruptable Power Supply or UPS. The link will lead to additional information. Often called a battery back-up. Used with any electrical device (like computer equipment) that must continue to operate when normal power is interrupted.
The UPS is a power line (usually 120 volt) battery backup system. A relatively large capacity low voltage rechargeable battery is maintained in a charged condition and is connected to a converter (inverter) that produces 120 volts alternating current.
The converter (inverter) is instantaneously energized and switched to carry the load when the normal power fails as during an electrical storm or other reasons.
A UPS is a perfect solution for saving 3D prints that have a lot of time and material required in their creation.
There are many sizes, brands an features available when selecting the UPS for this task. The choice was a given for me as the brand is owned by my last employer, Schneider Electric -- owners of the APC brand. We used them exclusively as back-up in our energy management control systems.
The proper sizing is the important decision. The UPS is generally intended as instant replacement of power when the normal source in interrupted. Its action is “instant on” invisible and instantaneous supply of the (usual) 120-volt power. The load never “sees” the power loss.
The UPS is NOT generally intended a replacement for back-up mechanical power generators. It is the “gap filler” between a power failure and the startup time for a motor-generator or other long-term power supply.
That reduces the cost and maintenance of needing a high capacity UPS. Batteries are maintenance items and need replacement. But the purchaser is free to choose the capacity and run time they need for the application.
My choice with 3D printing is a UPS run time of not more than one hour. Thirty or as low as fifteen minutes would be adequate. My desire is to ride through the usual one or two second “blink out” that we usually suffer in a weather event. Long term outage is rare and a risk I will accept for a print loss.
It’s the frustration of those few second outages killing a long print that the UPS is intended to relieve. Starting a print in the morning and hearing thunder in the afternoon is not comfortable and cause anguish.
We have two printers connected to the UPS shown. Each will draw 250 watts when the print bed is being heated. Normal run wattage is about 50 for one printer and 100 watts for the other (part of the 250 total load). Not sure why the difference. The nozzle heater may be larger on the 2 color printer. Stand-by is 7 to 10 watts (steppers and fans off). Except for start-up the high wattage print beds cycles with about a 10% duty factor.
Peak load is about 500 watts, but it is not likely the UPS would be used when first warming the print bed. Average run load (per printer) is 50 to 100 watts. 150 watts with both printers running.
I selected a 900 watt APC UPS. The pure sinewave some units feature is not critical. The UPS unit wattage needs to be adequate for the full wattage it is intended to support when it assumes the load. It is also an indicator of how long it will operate under the load. The price class for the UPS shown here is ~$170.00
The Schneider APC has intelligent controls that display the power (wattage) being drawn. It then computes the run time for the present charge of the battery. (Other brands may operate differently) With one printer operating the indication is the UPS can carry the load for one hour on the battery. That more than meets my goals. Two printers would probably run for thirty minutes.
I have produced a video that demonstrates the action of the UPS in a simulated power failure. There is absolutely no hesitation in switching to the UPS when power is interrupted. The same when going back to normal power.
The UPS fully satisfies my needs and peace of mind about running long prints. I must point out my computers are NOT using the UPS so long prints are always run from files loaded by SD memory. There is no particular reason the computer could not also share the UPS but in my case, the printers are not close to the computers.
I thought it was the actual printing. That is only part of it and was a huge part when I started a couple of years ago. The fascination of the process and how the machines operate is what got me interested. Watching a FDM filament printer running under CNC control was mesmerizing. It’s still that way today, but not like when I first started.
I have owned six different printers and still have five of them today. One of them is a DLP, resin and UV light. My first delta style printer I gave away when I upgraded to a slightly larger Delta of the same make and style. The printers are simply a tool. They create but are not themselves creative.
I realize today that printing is not what keeps me going. Once mastering the machine operation process and understanding several dozen variables and how they interact, printing is quite boring. Load a file, check parameters and material, push start and I am finished for one to twenty hours while the printer does its thing. No fun there.
What I really get excited about is the CAD drawing and the total control of creativity the computer drawing provides. The CAD is also just another tool. The 3D printer gives me the ability (the POWER) to produce a tangible item that I know I can make with the CAD drawing I have created on the computer screen.
It’s not the (brand) name of the CAD program that’s important. Like the 3D printers, some are easier to use or have more features than others. What works for one person may be a problem for someone else. The CAD doesn’t create anything. It is just a tool. It’s all about the person using it and their skills with the tool.
I had to learn the limitations of the printing system and design my creations within those limits. That is the same for any creative or artistic process. It’s that knowledge of the tools that separates the pros from the rookies.
If I couldn’t do the design thinking and the CAD work, I would have lost interest in the whole 3D printing operation. For me there was a period of fun learning to operate the printers (the tool) , but there is no creative outlet in just printing someone else’s CAD designs. Caring for and running the printer(s) is just a job, not an artistic expression. The art is in the design work. The fun for me is creating something original from just a conceptual idea.
The 3D printer is a tool that turns ideas into a tangible reality.
I have had some time to think about several blog articles I recently read about desktop 3D printing. They were about the time and effort (and money) we and the machine manufacturers invest for the results obtained at the hobbyist class machine cost level.
Example, an original project promoted for hobbyist. REPRAP. A desktop machine that can build itself. Really? No… a few plastic pieces of dubious quality, Yes. It helped kick start the hobby interest. It is not the best printer design.
Two products, one made with hobby level desktop 3D printing and the other with professional injection molded parts. I know what I would purchase. A friend and I both looked at recently designed (large) kit DIY printed parts 3D machine. The non-printed components (~$700) were an excellent choice. The self-printed (plastic) parts could make the complete printer a very bad investment.
The elephant in the room is 3D desktop printing as a hobbyist uses it, cannot deliver equivalent consistent output of conventional subtractive manufacturing, injection molding, or profession additive manufacturing. We may pretend it does. I can produce a facsimile of these items. Professional quality production 3D printers are not small desktop sized machines.
I ventured into resin (DLP) 3D printed, Lost Wax Casting models. I used a Wanhoa D7 Replicator. (lead picture) I had some success. But overall, factors beyond the actual printing proved wax is still the most consistent and cost effective method for a hobbyist doing lost wax casting. Making jewelry on a hobbyist 3D printing machine is possible but not practical.
Prototype modeling is a desktop machine's strong suit. Scale model, low volume, specialty, static display components.
What still scares me is some hobbyist using “carbon fiber” PLA and printing propellers for his drone or R/C aircraft engine with his $400 desktop hobby machine. Propeller failure at 20.000 rpm is lethal.
One guy writing about his Moai SLA printer, pretty much summed up what he has. (and all hobbyist have.) A tinkerer’s delight. A good first step in understanding the POTENTIAL of 3D additive manufacturing. The hobbyist grade hardware is basically a toy. A dental lab will not or should not be using a D7 or a Moai.
Technology junkies like myself are attracted to the 3D process and what we and a 3D printer can produce. We invest in our love of things technical and the low-cost desktop printer makes it possible, but so far it is mostly a hobby stuck in “demonstration mode” for me.
I like what I can do… not knocking my interest. Mostly making unnecessary plastic JUNQUE like 50+ plastic owls...
The kid who prints a dog cart, for the pup with no back legs, has not created anything new. The fact it was 3D printed doesn’t make it a better product. What it does show IMHO, is that this 3D print technology is getting these kids (and adults) into tangible, hands on making of things. The story gets press coverage because of the use of the “new” 3D print technology.
If hobbyist desktop 3D printing is anything, it is an enabler for tangible creative effort. It’s the new (toy maker) kid on the block. In an age where home machine shops are extremely rare, it brings affordable manufacturing onto a desktop in a home or a school. That’s what is good about 3D printing as a hobbyist.
What’s actually being printed is not necessarily the best way to make things, but it is a very satisfactory way to see designs come into tangible existance.
Very high-resolution 3D printing is associated mostly with layer height designated as Z in most systems. There is also resolution in the X/Y plane. FDM filament printers control X/Y resolution with nozzle diameter and flow rates.
With SLA resin printers, X/Y resolution is the size of the laser spot or focus point. Laser power levels and slow travel can cause light “bleed” and affect the effective dot size very slightly. Generally, the spot size is fixed by the design of the machine and the laser.
DLP and DLM (Digital Light Projection and Digital Light Masking) resin printer dot size is fixed by the number of pixels and their projected size. One to one is the highest resolution but it is possible and common that the image has less pixels than the hardware. A 1K image on a 4k screen, enlarged to fill the screen, is still a 1k image.
By far the most common and manipulated resolution variable with every type of 3D printing is the Z layer height.
There is a very serious factor to consider when attempting very high-resolution 3D printing. That factor is printing time. Time increase as the cube of the size and in inverse proportion to the layer height with FDM printing. Double the size and halving the layer height (Four time the number of layers) could take 16 time longer!
Large FDM printers (prints) and super fine resolution are just not practical together.
DLP and DLM print time are ONLY affected by height and number of layers, and NOT X/Y size. Doubling the height doubles the time. Reducing layer height by 50% does not again double the time as exposure times decrease as layers become thinner. Rather than 200% longer print time, it may be 190% longer.
Doubling the size and reducing individual layer height by half will be 3.8, say... something less than four times longer print time.
Good FDM printers will produce 100um (micron) layers. I have seen claims for as small as 50um. But doing any print of reasonable size, say... within 64 cubic inches (4x4x4) at 50um will take (just for comparision) perhaps 32 hours. The Resin DLP/DLM printer could probably do it in half the time 16 hours) or better as it does a complete layer at once without X/Y travel.
My general rules:
Large prints, 100um to 400um layers then FDM (filament) printing
Small prints, very fine resolution 10um to 100um then SLA, DLP, DLM (resin) printing.
Otherwise, pay the Zeitmiester (time master) his due.