No edit summary
No edit summary
Line 223: Line 223:
====August 24<sup>th</sup>, 2012:====
====August 24<sup>th</sup>, 2012:====
:: Printing and slicing software installed on smartbook. Control of a different 3D printer through the computer was successful. Also the slicing (turning a  3D cad file into printable gcode language) was more successful than expected, it only took 2.5min to slice a fairly large printable object. So far the smartbook is meeting and exceeding expectations.<br /><br />
:: Printing and slicing software installed on smartbook. Control of a different 3D printer through the computer was successful. Also the slicing (turning a  3D cad file into printable gcode language) was more successful than expected, it only took 2.5min to slice a fairly large printable object. So far the smartbook is meeting and exceeding expectations.<br /><br />
==Builing and Testing==
====October 29<sup>th</sup>, 2012:====
:: Battery connector built, batteries tested, controller  circuit designed and parts ordered. Still waiting on printer.<br /><br />
==Testing==
===Panel Tests===
===Panel Tests===
First panel test:
First panel test:
Line 270: Line 272:
Image:Batt_in.jpg|Fig. 13 - Battery inside
Image:Batt_in.jpg|Fig. 13 - Battery inside
Image:Batt_bsm.jpg|Fig. 14 - Battery Management System
Image:Batt_bsm.jpg|Fig. 14 - Battery Management System
</gallery>
</gallery><br /><br />
==Building==
== See Also ==
== See Also ==
* [http://reprap.org/wiki/Recyclebot Recyclebot on RepRap wiki]
* [http://reprap.org/wiki/Recyclebot Recyclebot on RepRap wiki]

Revision as of 16:13, 29 October 2012

Template:SEARC

Project Overview

This is a joint project between:

Project created and supervised by:

Project Student Research Assistant:

Research Period:

  • June 2012 - Ongoing

Linked Project(s):
Mobile Solar Powered 3D Printer

Objective

To improve on the design of the original mobile solar powered 3D printer. Build a smaller, more portable charging station, that runs solely on renewable energy for a 3D printer using open source appropriate technology.

Update: 23 July, 2012. Two separate models will be designed.

  • Model A - will be a ultra-low power model, using an SD compatible RepRap. This will be a small, light model; but will only be able to print parts that have been stored on the SD card.

  • Model B - will be capable of new part design in the field.

Fig. 1-RepRap Logo

RepRap Background

The RepRap is a low-cost 3D printer, designed by Dr. Adrian Bowyer. It runs on open-source software. And is self-replicating.

More information on the RepRap can be found on the RepRap at:RepRap.org/wiki

Model Design

Tasks

  • Come up with a design
  • Choose components
    • Printer
    • Computer
    • Panels
    • Batteries
    • Charge Controller
    • Inverter (if needed)
  • Test individual components
  • Build design
  • Test design

Choosing Components

The first step of the planning will involve research all of the possible components that will make up the design(s).

Printer

The printer choice will determine the size, power consumption, and the amount of customization available. A number of different RepRap models will be considered. A Mendel, running Gen6 electronics is already available for testing power circuits with. The RAMPS circuitry will also be explored. RAMPS is made with open-source Arduino electronics and is very customizable.
Other models will be considered, with the criteria being:

  • low power consumption
  • small in size, or foldable or easily disassembled
  • customizable (open-source, available SD card reader, unused i/o pins on board)
  • price
  • availability
  • print quality

One particular printer model that stands out is the FoldaRap. It was designed to fold into a very small size to be completely portable. It is an open-source design, with full build instructions posted by Emmanuel, the creator, and a member of the RepRap community. The FoldaRap could be an ideal candidate for the project.

Update: 23 July, 2012. The FoldaRap has been chosen for Model B.

PV Modules

It has been decided that this model will run on solar power. All current photovoltaic technology will be examined, to determine the appropriate solution for this project. The main factor in deciding which module to go with, will be portability. The sheer size and weight of the common glass covered panels make them unviable for portable applications. Thin film modules will most likely be used. All of the power drawing components must be chosen for the wattage requirement to be determined.

Batteries

In the original mobile solar cart used 4 deep cycle sealed lead acid batteries. They had a combined weight of 400lb, and a volume of over 64,000cm3. This, along with the two bulky panels, was the reason for the cart's extreme size. In this redesign, small battery size and weight will be crucial.
Lithium batteries are a clear choice over SLAs. While Li-ion batteries have only approx. 6% higher energy capacity volume-wise, their energy density is much lower, making Li-ions much lighter. Also Li-ion batteries are more stable and therefor easier to transport. The specific type of Li-ion, LiFePo4, or another type of lithium polymer battery is yet to be determined.

Laptop alternative

A large part of this design will involve reducing the power consumption of both the RepRap and the devices needed to run it. Running the printer from an average laptop more than doubles the watt hours needed.

Possible

Solution

Price Power

Consumption

OS Pros Cons
Raspberry Pi* $35
3w
Linux Very cheap, Large online community support,
RepRap software available on Linux
Long delivery times,
Unknown if processing
speeds are enough to run
slicing software
APC 8750 $49
13w
Android 2.3 Better processor than
Raspberry Pi,
Good price
No available software
would have to write
new program,
Not readily available yet,
High power consumption
Efika MX Smarttop $140
5w
Linux Runs Linux,
Most processing power
of mini pc options,
Already enclosed in housing
Higher cost,
has been recently discontinued
Efika MX Smartbook $199
0w
Linux Runs Linux,
Battery life of up to 7h
so no extra power draw,
Wifi & 3G for downloading new designs,
lowest cost for highest functionality
Higher cost
Control through cell
phone via bluetooth
$29 (with exsisting cell)
0w
Android Cell phones widespread,
cool factor
Current software needs improvement,
can only print designs already in hand
Use only a SD card slot $35
0w
N/A ultra low power can only print designs already in hand,
no community design
Tablet $150-500
0w
Varies No extra power draw
on system,
Readily available,
Higher cost,
Unknown if it can run software
OLPC $100-200
15w
Linux Large user community,
already scaled in developing world
Expense,
difficulty running some software
see notes
Build OS Micro PC Unknown
?
Any Can build for very low power,
Keep cost down by not using
unnecessary components,
Will create new Open source, low power/cost PC
More time to build,
many unknown factors


* See Progress Update, August 2nd, 2012
A low-power monitor would be needed in conjunction with the first three options. There are many different models available at prices starting as low as $20. Different monitors would be chosen for each computer, based on available i/o configurations.

Charge controller & Inverter

The rest of the components needed to complete the power circuit, such as a charge controller, inverter, combiner box, circuit protection, ect, will be chosen or designed after the panels and batteries are chosen. As many of the parts as possible will be built, rather than bought.

CAD Designs

Fig. 2 - 1st draft of design
Fig. 3 - Closed view of 1st draft of design
Fig. 4 - Under view of Uni-solar table
Fig. 5 - Uni-solar table construction

Draft One:

The initial designs did not involve the plan to reduce the size of the printer. Therefor the idea to build a box out of sheet metal was used. The first design (Fig. 2) used a Uni-solar flexible panel. Their ability to completely roll up into a very small volume made the Uni-solar panels the first choice. This design is also showing a Raspberry Pi w/monitor. The green block is a LiFePo4 battery, which was spec'd to keep the setup running for up to 4 days off a single charge.

The wood shown in the model in Fig. 2 is a make-shift table set-up for the PV module (Fig. 4.) The Uni-solar panels are covered in a thin flexible laminate. Although made to be durable, laying them out on a rocky, sandy, or otherwise unsmooth surface repeatedly would not be advisable. Also a table would provide opportunity to tilt the panel, to achieve maximum sun exposure. The panels measure 18ft (5.5m) in length, and provide quite a challenge to fit a table for it into a small space. One solution was to use lengths of 1"x2" cut in the maximum dimensions of the carrying case, fit them together (possibly with printed parts) into two 18' rails. Then using 15.5" (the width of the panel) lengths of wood, connect them together in parallel using a durable ribbon like material, and a staple gun. Make the first piece and the last piece 18' apart, and affix the rest at equal distances in between. Lay the attached wood, fully outstretched, on top of the two 18' rails. All that is needed then is four nails, one on each end of the 18' rails, on the inside of the first and last slots (see Fig. 5.) Be sure after placing the nails behind the first slat, to pull the string of slats tight before placing the other nails before the last slat.

Draft Two:

The second draft was designed to fit in a standard carry-on travel suitcase. This increases its ability to be transported to the remote locations that it will be most useful.

Bill of Materials

Item Model/Description Quantity Unit Price Total Price Website/Store Additional notes
3D Printer FoldaRap
1
$800 CAD to purchase kit,

parts can be purchased & printed separately, and assembled

$800 buy kit

DIY - BOM and build instructions

Genesi Smartbook Efika MX
1
$199+tax $242.34 Genesi Any laptop can be used, as long as it has the battery life to last through a long print
Semi-flexible 20W PV Panels ICO-SMC-20W
5
$86.50 $432.50 Eco-worthy
Batteries 6.6Ah 14.8V Li-ion HSTNN-UB02
4
$31(after shipping) $124 Amazon Any model 6600mAH 14.8V battery will do



Progress Updates

July 24th, 2012:

The FoldaRap has been ordered from the maker's crowd funding campaign. Awaiting confirmation from the maker of the power consumption before ordering the other components.

Currently researching DIY charge controllers (with MPPT) and inverters.

August 2nd, 2012:

Raspberry Pi tested for viability. Tested board was very glitchy, could not source enough power to run peripherals (ie: mouse, keyboard.)The oscillating quartz crystal was not working properly. Most likely just got a lemon, but from reading the forums, this was not the only one. Because of this, the slow lead times for ordering, and the likelihood that they would not run the needed software anyway, the raspberry Pi will not be used for this project.

A few different high efficiency MPPT controller circuits have been found. Still awaiting power consumption numbers for the FoldaRap to begin adjusting circuit to meet specific power requirements.

Exploring the option of designing a bare bones pc to run the RepRap host software, then comparing the prices to buying a SBC (single board computer).

August 7th, 2012:

After researching many SBCs and finding the average cost to be around $150 for the necessary specs, it was decided to settle on the Efika Smartbook by Genesi. This laptop is $50 more, has the necessary specs, but would not draw any power from the circuit, since it has its own battery life of approx 7 hours. And would not require the purchase of a display. Genesi is a company whose sole focus is making low-power, low-cost computing solutions, geared for developing countries.

August 9th, 2012:

Batteries and panels ordered. Still unsure of FoldaRap power consumption, so a power circuit was chosen that can run a 60W printer for 6.5h, or up to a 90W printer for 4.3h. Still much longer than the average 2h print.

August 16th, 2012:

Panels and laptop arrived today. Preliminary testing shows both are working as expected.

August 21st, 2012:

The batteries arrived last week. Dismantling them revealed a good onboard BMS. Now it will have to be decided whether our system can simply communicate with the current BMS, or whether it will be scrapped and the controller will be used to control the batteries. See Battery Tests for more details.

It has been decided to base our charge controller design off current open source designs available. The two controllers under consideration are the Free Charge Controller and the Arduino Peak Power Tracker Solar Charger. The Arduino build is a much simpler design, and would therefore be easier to modify to for this projects specific requirements.

August 24th, 2012:

Printing and slicing software installed on smartbook. Control of a different 3D printer through the computer was successful. Also the slicing (turning a 3D cad file into printable gcode language) was more successful than expected, it only took 2.5min to slice a fairly large printable object. So far the smartbook is meeting and exceeding expectations.

October 29th, 2012:

Battery connector built, batteries tested, controller circuit designed and parts ordered. Still waiting on printer.

Testing

Panel Tests

First panel test:

Spec'd Measured
Voc 20.6V 21.02V
Isc 1.38A 1.04A

Also tested under a 20.5ohm load and got 18.09V. Will do further testing with a larger selection of power resistors.

The short circuit current (Isc) was not as high as spec'd, but the panel was not optimally angled toward the sun. Another test will be performed to gauge Isc.

The panels are very light weight, they will be easily transportable. They appear to be quite durable; and they bend to the 30° as advertised (See Fig. 8.)

Smartbook Tests

The Genesi Efika MX Smartbook comes loaded with it's own unique Linux distribution. It loads fairly quickly, and connected to the nearest unsecured wireless network automatically. The speed is not the best, but as good as expected for it's hardware. It is extremely lightweight, and has very small overall dimensions, even when compared to a 13.5" netbook (See Fig. 9.)

Its power consumption is lower than the 12W advertised.

Start up: 7-8W
Idle (with wireless on): 6-7W
Idle (with wireless off): 5-6W
Playing a movie(wireless off): 6w
Suspended: 1W



Battery Tests

The 14.8V 6600mAh Li-ion batteries, are tested, and are outputting 14.6V.

The batteries come with their own battery management system (BMS), which controls the way they are charged, monitors cell levels to avoid over charging or over discharging, and monitors temperature with a thermistor. It is designed to communicate with its specific laptop model via a 7-prong output (See Fig. 10 & 11) Batt out pic.jpg
Fig. 10 - Battery connector pin out

As per Fig. 10, it was tested that A and B are connected and make up the negative terminal, as F and G are connected and make up the positive terminal.

According to the product website the battery uses SMBus for communication. SMBus is a two-wire communication, loosely based off I2C communication. It would be preferred to build a controller that talks to the current BMS that comes with the batteries, as it is already tailored to the specific cells, and will save time in programming, and avoid potential problems and damage to the cells with improper coding. Li-ions charge different from sealed lead acid batteries (SLAs) found in larger solar projects. While SLAs can be charged by simply applying a small constant trickle current until full, Li-ions have up to 3 different charging stages, depending on how far they have been discharged. This required a more complex BMS, which is another reason it would be ideal to keep the current BMS in place.



Building

See Also

Cookies help us deliver our services. By using our services, you agree to our use of cookies.