Project Logs

Collapse

• Finished the charging circuit!!!

  Dylan Bleier • 05/26/2015 at 23:52 • 0 comments

  I finished the charging circuit / enclosure, sealing everything in with excessive amounts of hot glue and re-used dirty plexiglass. There's even space left over to add the Boarduino when I get around to that. I also cut a tarp to size for the solar panel and have something almost perfect in mind for a rigid box for the solar panel – a large contractor's clipboard. The last crucial item I need now is the encapsulant for the solar panel. Since I don't have a lot of time before I need to move and start a full time summer job until August, in a place where I might not have access to a full shop and all the things I need, I'm going to try to overnight some encapsulant and finish the panel before I have to leave. Then I can use the device over the summer and see if it holds up or if I should change anything. I'll probably be able to do a bit of work on the other parts (the lapdock and the thermoelectric generator) before August, but I'll be really busy with other things. Anyway, the reason I was waiting to do the encapsulant was that the ideal material to use (according to NREL research presentations I saw online) was about a millimeter thick of TPO (thermoplastic olefin) or TPU (thermoplastic urethane). You can't just go to the hardware store to find suitable materials for this, as you probably want something you can melt on in sheet / film form with all the right additives to slow down solar degradation from heat as well as UV and visible light exposure. The only thing I've seen so far is $300 rolls of TPU or TPO on ebay, the quality of which is not guaranteed in the least. There's probably something better that could be found, but since I'm running out of time, I'll just get some EVA (ethylene vinyl acetate), the standard that's been used successfully for decades in the solar industry but probably isn't actually that great because the degradation modes over years can include yellowing and browning due to production of acetic acid which corrodes solar cells. With this project, if I use it for an hour or two a day, it could still potentially last decades.

  Unfortunately, I can't find my camera right now, so I can't take decent pics to share. I'll try to borrow someone else's camera soon.

  solar panel to do list:

  - order EVA (and TEDLAR?) with fast shipping

  - rewire cells using the good quality flexible wire, adding a diode in series before the plug, clip sharp corners (?), test

  - gently swab/wipe cells with methanol to remove corrosive rosin flux and any dirt/dust; allow to dry fully (6+ hrs at room temp with air flow?? or use a heat gun on low)

  - immediately seal in a cool, low humidity room, using a shop vac to pull a vacuum between the EVA and the panel + TPE backing in order to prevent bubbles; seal each cell individually using a heat gun to melt the polymer film. seal the TPE/solar cell/EVA/FEP. trim so EVA is not exposed and the edge is TPE/FEP sealed together. add butyl flashing tape to seal even better.

  - glue or sew (*if there is a suitably large margin of plastic) the solar cells in place on the tarp. sew the flexible wires into place on the tarp to secure them.

  - add foam weather stripping to the sides and if necessary, the middle of each cell to protect them while in the box. Fold it up and put it in the case, stick it in a backpack, and go test it out!
  

• almost finished a lot of stuff

  Dylan Bleier • 05/23/2015 at 21:44 • 0 comments

  I put together the solar panel (minus the coatings and flexible backing. Around mid-day it gave 20V at first and then 19V after heating up in the sun, and it charged the lead acid battery pretty quickly. I redundantly wired all 10 cells in series, using short lengths of stranded wire. some of the wires are too weak, so I'll replace them with more 120VAC cord. The panel folds up lengthwise first, then the two strings of 5 cells fold over each other into one tight bundle. I will need to get a hard case to protect the finished product. As for the charging circuitry, I installed it all in the case with the fans, using a drill and some hot glue.

  .

  enclosure / charging circuit to do list:

  - drill/cut/attach clear plastic front face; mount switches and USB x2 using hot glue, holes for screwdriver to access pots, output cord/jack for lapdock

  - get and attach: replacement LED voltmeter for the one I blew up by replacing the voltage converter with a pot that I overheated, also a female 1/4" audio jack for 12V out.

  - add fine mesh screen to protect fan and everything inside from debris (?), add more holes on top, front and sides to improve air flow (?)
  

  .

  solar cell to do list:

  - replace the weak wires on the solar cells with more stranded 120VAC cord

  - add a diode in series with solar panel or step up/down charging circuit to prevent any possible damage

  - TPO or TPU encapsulant (I think these perform much better than EVA if I can find them in the right formulation somewhere?) for the solar cells, apply this and then the FEP using heat gun under mild vacuum (e.g. shop vac) after cleaning cells with suitable solvent, and cleaning off rosin flux from soldering

  - get a tarp or something and cut it to size, melt/seal/sew/tape edges so it won't fall apart; mount the solar panel on this and secure the bendy wires; if necessary add something to make the cells more rigid

  - get/make rigid storage enclosure for solar panel, and foam strip spacers so packing the cells in the box won't damage them while in transit. possible items for an enclosure include a contractor's clipboard box, or a DIY box I could make from scrap bamboo flooring.

  .

  thermoelectric generator to do list:

  - get peltier junctions and clamp a water cooled aluminum block to one side and clamp a series of heavy copper wires to the other side. find a second barrel jack connector for this

  - make a series of experimental high temperature peltier junctions using inexpensive materials (e.g. graphite and aluminum)

  .

  RPI lapdock to do list:

  - research config.txt settings, reconfigure to work with lapdock... this may take some time, and it's possible I soldered the wrong USB data wires since there were 5 or 6 wires in that jumper that came inside the lapdock

  - research startup sequence, add switches where necessary in the USB and/or HDMI wiring.

  - once it all works perfectly, slim down all the parts, finish the desoldering of the ethernet and USB ports and move them, hot glue everything into place inside the RPI including USB drive, GPIO cable, and then close it all up.

• Found enclosure, soldering solar cells, etc.

  Dylan Bleier • 05/23/2015 at 02:47 • 0 comments

  I found a perfect sized enclosure for the circuitry.

  I soldered the Boarduino together for future use.

  I tested all the solar cells and found that they all work, and that they all deliver about 2.0 V in direct sunlight (in the late afternoon). This is with no load; the voltage should drop some with a load. Since I have 10 cells in series it should be around 20V.

  I soldered 5 of the 10 panels together in a string that folds up. Tomorrow I'll test the whole string in direct sunlight and solder together a second string of 5, and then I'll put the two strings together. All the cells will be in series.

  I need to find a source for a TPO (thermoplastic olefin) or TPU (thermoplastic urethane) transparent coating that is suitable for these solar cells. I am thinking about looking in my local hardware stores for something I can just paint on. It just needs to have a the right chemicals, not have any corrosive stuff at all, and withstand UV light for a good 20-50 years. I'll also look around online to see what films are available.

  Also, I found that I don't need the LM317 part of the circuit anymore since I can simply adjust the pot if I need 18V rather than 19V – that step-up converter has constant output voltage even if you vary the input voltage, so my laptop won't get fried even if I overcharge the battery. But I still need a decent 10kΩ pot (or digital pot, eventually) for the other voltage converter because it doesn't give a constant 13.8V output to charge the battery.

  I've also been working on the RPI lapdock. I got my HDMI adapter and soldered two jumper wires to splice the USB signal while still getting power to the microUSB of the RPI. I may need to add a switch to turn the USB power to the RPI on and off. I also need to figure out how to configure the RPI startup menu so I can successfully boot onto the lapdock. Plus whatever sequence of things I need to do to start it up. And I still need to re-route the USB and Ethernet. Desoldering is terrrrrible.

• Main circuit almost complete

  Dylan Bleier • 05/19/2015 at 04:02 • 0 comments

  see gallery for schematic, more pictures to come. circuit diagram does not show potentiometers: there are two, attached to voltage regulators.

  I finished drawing out and assembling the main circuit, except for the linear voltage regulators and toggle switch that go on the 18/19V output to protect a laptop when it's charging. I also need to get regular size pots to replace the tiny ones on the boards. Then I will put it all in an enclosure with lots of hot glue and electrical tape to keep things in place. Once I get everything working well, I'll add a Boarduino or Trinket MPPT controller that will control the pots and switches to maximize efficiency.

  update: I just added some photos. sorry about the potato quality - I could not find my regular camera and had to use a webcam. I'll replace them later. So far I have the main circuit assembled except for replacing the tiny on-board pots and adding the 78XX circuit(s) to ensure the laptop charger doesn't go over 19V (or 18V). The audio jacks are not ideal, as they can short out momentarily while plugging in if you don't do it straight. I have a 5W 12V compact fluorescent bulb hooked up to another audio plug; I need another female 1/4" audio connector to add to the circuit. On the breadboard is a Boarduino and a RPi GPIO T-cobbler, both from Adafruit. The RPi laptop is also coming along nicely but I'm waiting for an HDMI adapter that should arrive later this week.

View all 4 project logs