Our regular Thursday meeting this week was pretty much exclusively a Chill Box Build. Cole and I arrived a bit before 5pm. We spent about 20 minutes walking the grounds and looking for places on which to put Our Container. There are not many places with enough space and access for the trucks to drop off. Rai would like it to be not all that visible from “street”. However, there are not a lot of places where we could set up a 40×40′ container building. I think the best site is straight back in the open space. There is a fair bit of Habitat for Humanity material in this area. It looks like the parts to make a steel building, and a whole lotta roofing tiles. The CWD people know the container shows up Tuesday AM and will be planning a spot, at least for temporary siting.
After our walkabout, we got down to business – building the Chill Box. We brought the parts and tools inside and started assembling. The chiller parts (plastic tubing, the Corsair cpu cooler, peltier, valves, etc) are a tight fit. Tubes dont bend as easily. We tried adding a bit of larger tubing on one end, and it fit better. Cole made a quick run to Home Depot for some clamps and silicone sealant.
I worked on building some power cables for the Corsair. It uses small 3pin CPU Fan connectors. Two leads are power and the third is for speed sensing (see pic at left). Cole had previously cut three header pins from our stock and was using them in a breadboard with alligator clips to power the device. Two cables are needed. One connects to the Corsair’s pump, the other connects to the double fans with a vendor provided Y-cable. I used some 4-wire cable, stripping off the green wire. I soldered these to the pins, covering each with small shrink wrap, and then shrink wrapping the entire connector. The other ends were connected to an 8-pin connector strip. We wired up the Peltier and water pump to this 8-pin connector too.
When Cole returned, we clamped the tubes, calked around the edge of Peltier (to avoid condensation shorting it), and filled the tubes with water. Filling the tubing is tricky. The basic fill is easy, but there remains a bubble that takes a fair bit of manipulation to remove. Perhaps there is another arrangement that would make it easier. Perhaps a bleed valve as well as the fill valve? (spoiler – on April 30 build we did major improvement to plumbing and fill).
Brian T, Tommy and Jen R came by about 8pm to see the space, project and and chat.
Then came moment of truth – powering it back up and seeing if it still pumps and cools. We got it hooked up, but went through a couple false starts with power not being connected properly. We’re using a hacked PC 300w power supply – kinda wimpy and it sometimes shuts down. We get it pumping and … well the block seems to be getting cool, but not icy like last week. There is a lot of hot air blowing around in the container. Perhaps the vents are not working well enough? The wave side of vents already has some broken pieces where the laser cuts got too close together, so we break out some larger chunks. This seems to help, although arrangement of components in this area is problematic. The corsair now seems to be pulling cool air in and pushing out only slightly warmer air.
The little IR temp tool Cole got shows it cools down a bit, and the pipes cool, but then it starts warming up. Drat… Maybe its the wimpy former PC power supply? Several times these have kicked off (shorted?) and then reset themselves after sitting for a bit. We pulled out the 12v battery, which had done a better job last week, and hooked it up to the 12v supply. It works good for all the 12v stuff (Corsair and peltier) but the water pump is 5-9v. Since the pc supply was down, we decided to test the pump using 12v. It worked – pumped fast and heated up quick. Wait, it heated up? Oh drat – its heating the water! Shutdown, try the pc supply again – it works and runs the pump ok. The pump still heats up. This little beastie is fully enclosed in plastic, so the heat has really no where to go but into the water. They really sell this thing as good for cooling systems? We let it run for a while but it appears the water pump is countering whatever the peltier can do to cool. Drat! It worked last week right? Ok. dont panic yet. Cole has a little micro-pump he ordered early on. Its got really small hoses but is supposed to be really good (rather expensive too). Unfortunately, he left it home. Since its getting on 10pm and we are fading, we decide to call it a night. Cole takes the container and pump system home to work on it a bit. We plan to get together one more time over weekend to try and get it running.
Here’s the Flicker Set for the April 28 build night:
[slickr-flickr tag="ChillBoxApril28" descriptions="on" flickr_link="on" sort="title" ]