Scratch Build – In Progress Project RabidCicada (Mineral Oil Computer) - 20th October 2010

I had some whiteness at the bonding surfaces because I didn't have the polycarbonate clos eenough together.

But everything sure bonded fine.

What in particular are you talking about with the crazing. I know what it is I just don't know where you're expecting a problem.

 

Pump And Full Case Leak Test

Just kicked off a full system leak test. Got the Tank, Radiator, quick disconnects and pump all in place.



I also have those pictures of the Modified Fan Contoller I owe you guys.

You can see that I used larger LED's than were originally on the controller board. It's just a form factor thing...and really is irrelevant to the fan controller unless you have fit issues. I chose 5mm LED's because they had the color I wanted.




And now you get to see the casualties of letting a software engineer do the electrical engineering part. I learned some important lessons:

• Use the right tools. The radio shack sodlering iron didn't seem to get hot enough. I ended up using a soldering iron that reached 750 degrees. Flux is your friend for heating stuff up. Clean up with isopropyl alchohol.
• Make sure to coat your soldering iron in a protective layer of solder between uses to prevent oxidation. This may have been what prevented the radioshack iron from getting hot enough.
• Don't get impatient. I got tired of waiting for the solder to soften up using the radio shack iron and pulled the solder plug out. Thats how I pulled the metal contacts out of the board that are supposed to be there for the two that got damaged. Take your time or you'll cause yourself grief.

To fix the problem with pulling the metal contacts out, I just ran a tiny wire over to his LED partner on the board.

 

I'm also noticing a small bit of rattling from a ball bearing in the slides holding the motherboard. I think when I dampen the pump itself, and have weight mounted on the mobo tray that the noise will disappear.

 

your soldering iron definitely was not hot enough. the solder didnt flow properly into your connections. personally, i use a butane iron. i love it because not only does it not have a cord that jerks it off of whatever precarious perch you set it on, but you can adjust the heat from glowing red to warm-to-the-touch.

wont the bearing on the motherboard tray get lubrication once submerged in mineral oil?

 

Yeah...I figured the bearing would lighten up a little with the mineral oil. But it makes a reasonable noise while submerged in water. So...I'm not counting on too much gain with the switch to mineral oil. If it does fix it..great. If it doesn't...then removing the source of vibration (pump on tank) should fix it.

 

I cant tell from the video, but eventually, whats the running order of the loop going to be? i.e. pump to rad or pump straight into the tank.

And would you also be, like legoman had, using tubing from the top of the pump to direct it towards the CPU heatsink, if indeed thats the loop order youre going to be using?

 

Pump->rad->tank
Pump->bulkhead->rad->bulkhead->tank

Plan is to dump "cool" mineral oil near gfx cards if I really need to worry about circulation.

 

WARP Drive on Build PLEASE

Sweet one heck of a build ya got looking very close.

 

Oh god.

I just helped host a middleschool robotics competition and I happened to run into a guy wearing a northstar hat. Since I had looked at positive displacement pumps before I asked him some questions about positive displacement pumps and related my problems with finding an ideal pump.

He recommended a brand called Grundfos. They look interesting with their submersible line.....

More thinking....

 

wow, im very impressed with what grundfos has to offer. take a look at model number UP-1542F. i dont think its positive displacement, but it can be had for under $100, it has an open flow of 16GPM, a head of 15 feet, and its made entirely from metal (ie, oil proof and STRONG/long lasting). thats a lot of pump for the money. it beats the eheim in every aspect.

although its non submersible so it probably wont work for your application

 

First off- awesome build thread so far. Thanks for the pictures/details.

Also, (sorry if this was covered but I missed it..) do you have plans to paint/ powder-coat the case?.. They can put a nifty finish on the metal which will last forever.. Just a thought..

 

Sorry about the lack of updates recently...familly life.

I didn't plan on painting/coating the metal. My hope is for the end result to look like a metal artifact in a jungle/tribal setting (created by the Tiki Mask and other decorations).

I'm speccing out a lid right now. I'm probably going to do power distribution through the lid but let all other cables rout out through a dust blocking opening (bristles or netting).

I got old computer parts back from my younger brother this last weekend and planned to test them in the mineral oil with the pump running but stuff got in the way. I did put some mineral oil in a small bread pan with aquarium rocks, a small painted wooden tiki toy, and a plastic plant to see how they faired in the mineral oil. I think there's a little color leakage which I will have to consider for my final thing. My self made tiki mask will be raw wood...but I was testing for color fastness in case I wanted to buy some typical aquarium decorations.

I also got an air pump to try out the aerator thing. Those puppies are louder than expected...even the "whisper quiet" models. So the aerator may be a no-go. I'll put in some more testing to vet out a possible solution. The problem is the air pump is diaghram and they tend to be louder. I would have to mount it to the lid or above liquid level in the case also.

I'll post a pic or two here shortly of the bread-pan test. Hopefully I'll get to testing running parts here also. I'm not bothering with thermal grease issues for my old test HW. I will need a good solution or maintenance plan when I move to my real HW. Old HW also will let me mess with tubing routing in the case and mineral oil flow.

 

I actually have to be careful about power now. I believe a typical wall circuit (not socket), can handle 15amps. When my power supply kicks on my lights dim a little then return to normal. I'm starting to dance awefully close to the circuit limit. 1.5KW supply AND a mineral oil pump.....

If I hope to use this at a friends house I am going to have to be very careful about top load values

 

Alright,

I've got the lid design mostly down. I've decided I'm going to do AC power distribution through a bus bar with terminals.

I wanted to cut and shorten cables anyways (eheim power cable, and power supply cable) so that I didn't have tons of coiled up cable sitting in my tank. I also didn't want many power cables running from my computer to the wall.

Using a bus bar and maybe even fuses right in the lid will give me less clutter and cables from the tank to the wall and will also look interesting through the clear lid surfaces. I will probably have one large cable from the wall to the lid (using a receptacle), then wire the receptacle directly to the bus bar. I will clearly have be careful and put safeguards in place to avoid electrocution. The polycarbonate will work well as a protector but still allow you to see stuff (hopefully. Don't have all placement and orientation details worked out yet).

I was trying to figure out a good connector scheme and surge protector to fit in the lid...but all the details got really annoying and finding a nifty looking surge protector, proper connectors, then all the work in cutting, and re-connectorizing the cables got to be too much. I think I'm really going to like the busbar thing.

I will also use a similar but way lower grade bus bar idea for power distribution to the fans in the radiator. I'll mount a bus bar in the top, inside of the fan grill, and cut the fan cables to get rid of extra wire and connectors I don't need, then connect the raw wires to the bus bar through screw terminals. With the fan grill in place you should see little of the wires which is the goal. Then I'll feed the power from the fan controller directly to the bus bar and power gets distributed throught the terminals.

I think I finally have a completed design...now it's time to execute. (Which, admittedly...I've taken my sweet time ).

I need to read up more on busbars, terminals, and power junk. I've got some power engineers I can consult here so I should be ok.

But please...if you have experience in this arena....feel free to help me out and provide tips and tricks.

 

Lid construction begins today!

Alright. All parts should be arriving today.

For power distribution in general I've ordered:
A Cinch 9-141
A Cinch 10-141
25 Jumpers to connect terminals together (Daisy chain power across)
50 Lugs (Wire terminators to allow good connection with Terminal bars)

I will be using a 14 AWG cable with standard wall plug and IEC 320-C13 (Standard PC power supply connector) to connect to the receptacle in my lid. That receptacle will be tied directly to the terminal bar for AC power distribution.

For fan power distribution I'll use the 10 Terminal connector. I will make 5 terminals positive and 5 terminals negative, and jump the terminals. I will double stack 2 fans to a terminal. I've way overspecced (inordinately) the terminal capacity for the fan power...but I'm all about overkill.

For AC power distribution I will use the same terminal type but I will use the 9 terminal connector. I will split Hot,Neutral,Ground to 3 of the terminals each with the help of the jumpers. Then I will connect wires for my AC components directly to the terminals. The terminal block and the jumpers are rated to handle the AC current/voltage I plan to put through there. Since I only have 2 AC devices right now (eheim, power supply) that leaves one slot free for expansion.

I will be grounding the case directly to the receptacle that's in the lid.

Those parts are arriving today.

I am also putting together the physical lid today with my Mechanical engineer. I'm probably just a little short on polycarbonate to finish the two smallest face of the lid...but that will be easily fixed here shortly.

Here are the parts arriving today for the lid:
Corner Connnectors to give a nice finished look.
Twist in Cable Tie holders, to allow neat bundling of cables.
Hinges for the lid.

I have aluminum stock and polycarbonate from before.
Ahhh...I can't wait!

 

Bring the pics

 

Mineral Oil and Decorations compatibiility Test

Right. So I needed to start scoping out how well mineral oil will play with the intended decorations and I needed to try to find a good air pump for the air bubbles etc. I had stated this before but I owe you some pictures....so here they are.

My supplies minus the pump. This includes a backstop air flow thingamagigger to stop liquid from working it's way into the pump if pressur issues occur...or capillary action


The bread baking dish that my wife apparently planned to use that night, *snigger*, and the air pump itself


The Aquarium rocks that I'm testing for color fastness in the mineral oil. I gave them a good wash with the hose to make sure they don't have junk in them.


The unsuspecting hair dryer I mangled. It has the funnel on it because I also used it to dry out the radiator from the water leak test run. I wanted to make sure the radiator was completely dry inside. I used a fanning tip to spread heat over the aquarium rocks (Thank god for all those damn hair dryer assecories *phew*)


AND...The final setup.


I'll probably check for color fastness and other issues on Monday. I don't know if I'll be using an air pump at all now if there aren't any that are quieter than the "whisper" model one that I got.

 

It has a lid....Kind of

Today my mechanical engineer and I got working on the lid. It turns out that my leftover polycarbonate is a little to short...and by a little I mean like under 2 centimeters...so that was no fun.

What I intended to do was to complete the lid and spend the weekend tinkering with getting my power distribution stuff into the lid itself. Instead, I have a new sheet of polycarbonate on order and the AC power stuff will have to wait till some other time.

What we could do though...was make the lid frame itself with my leftover aluminum stock. I decided I wanted the lid to accomodate 2 5-1/2 dirve bays. This was so that I could fit the fan controller and probably an optical drive (or whatever else I want) in the narrow face of the case.

The corner pieces I bought require you to tap the middle of the aluminum extrusion so we did that. Then we found out that the bolts that came with the corner pieces were too long for our tap length...so we ended having to shorten the bolts. Luckily we had a tool made specifically to do that. You screw the bolt through, then twist, and it cleaves off the extra amount. Those mechanical guys think of everything.

As an aside...I have to say I am very lucky to have access to the people I do. It is because of them that I've been able to put together the stuff I have. I really appreciate the electrical engineers I consulted to deciding my power stuff. They make sure I don't blow myself, my house, or anyone else up.

Pictures below:

Cutting more Metal


Tapping the Aluminum Extrusion


Bolt threaded through bolt cleaving device. It has a center piece that gets gradually larger as you rotate it. As it gets larger, it pushes into the bolt and literally cleaves the part off.


Mid-cleave on Bolt. The device is in a vice and being driven by a large hex wrench.


Cleave Finished.


All Bolts Shortened


First lid extrusion assembled. One corner piece attached.


One plane of the lid assembled (blurry)


Single lid plane attached to tank with hinges


Because of the lid hinges and wanting the lid to lay flat on the top, we had to retro-fit the existing connection from internal frame to external frame. We got rid of the two flat plates that connected the frames and instead we used two angled gussets. This is a close up of one.


From the Side


Testing hinge


Mostly closed lid part. I have the other lid plane completed sitting vertically and preventing the first plane from closing completely. Best view of hinges in this shot.


Whats this.....It's growing!


Definitely growing


And the lid frame is complete



I need to order some bumpers to put on the lid so that the lid touches the side of the tank when it's fully open and doesn't over-extend the hinges. I also think I'm going to slide the hinge location out towards the edges of the case...there was a little more twisting pressure on the aluminum stock than I wanted from the weight of the lid (and I've got lots more to put in).

I have caps for the holes in the corner pieces so those will look nice and sharp when I get the polycarbonate cut and in place.

 

Electrical bits

My trusty electrical engineers have helped me through this part. Knowing the exact part numbers and names is a huge hassle though it is clearly necessary to have all the details considered especially when designing and using electrical parts.

I got my terminal bars, jumpers, and lugs today. These will handle all my electrical stuff. I can actually get started on my fan controller power distribution this weekend if I feel like it. The AC power stuff will have to wait till I have a polycarbonate in the lid to mount to.

Terminals


Lugs (AKA ring Terminators). These get crimped onto the wires then the terminal screws are threaded through, guaranteeing the wire has a greate contact and wont come off by accident.


Jumpers. I will use these to turn portions of the the terminal block into hot, neutral, or ground (AC) pieces, or positive and negative (DC) pieces.

 

OH !!! You got a Lister Twin Engine How SWEET they Sound !

Got any info on the Motor ?

Nice Lid on Case ya making .