Project: Silent Overkill

Not exactly a progress update but I thought it might amuse people. My PSU (Nesteq Xzero) that I currently plan to use in the new case helpfully uses non-standard 6-pin molex connectors for the modular power leads (molex and SATA). Can't find them available as they're a different pattern to the 6-pin PCI-E shells so I need to get them off to make new sleeved cables when the time comes. Wouldn't budge with the pin extractors rammed in as hard as I could so I ended up giving the cable a good yank with a "what's the worst that could happen?!" attitude.

Turns out that the worst that can happen is the glue suddenly gives (explains why it was difficult) and the pin is fired out of the hole with the sharp wings (that stop it going to far into the hole) ploughing two furrows diagonally across my thumb. The yellow wire is now red :eek:

Plasters are hopeless - don't stick to fingerprints and don't flex. Gorilla Glue (form of superglue with knobs on), however, works wonders. Have a useable thumb again :D
 
Aqua Computer Farbwerk has arrived. Got a 1m strip of RGB LEDs to go with it. Might be able to use them to light the coolant channels in varying colours.

Also ordered fifty RGB LEDs (5mm, 4-pin, common anode) from China. RS only had the common cathode type that isn't compatible with this. They do, however, have the metric f-load of resistors I'll need to wire up a string in possible combingations of 1,2,4 and 8. Plan is to put a load in the bottom of the res - probably a string of 8 of them round the edge and an independent one up the middle core if it works in testing.

If you can smell the scent of solder and hear the sound of swearing, you know it's all arrived :D
 
Cheers. Must try to get the block design completed but work is busy and getting in the way :D Need to ask Parvum if my provisional design is plausible but also need to make sure it's mm perfect too, or it won't fit.
 
Planned to do a premature move to the new case today - need to confirm some measurements with the board in before finalising the block design - but found I needed to order some 90° fittings as people like Flix29 have changed my mind on how the tubing should be routed from GPU to CPU:

flix29 said:
SMA8_WiPl082.jpg
Click to expand...

Also looking for a short, black male to male rotary extension. EK do one but it's got straight knurling instead of diamond and will clash with others. Trying to source a Barrow one in black.

Should get some pics up and some idea of temps once the fittings arrive and I can do a functional but far-from-finished build.
 
Ok, some fittings arrived and I've got a few hours free to make some progress. Plan is to transplant everything from the old case and loop to the new.

While I'm at it I'm going to try to drill out the 3mm LED holes in my GPU and GFX blocks to 5mm if there's space. Why? Well, I wanted to use RGB LEDs (the single type rather than strips) so that I can either change my mind or have them change based on temperatures. This will be done with an Aquaero 6 (has a port for one RGB LED) and a Farbwerk attached for an additional four channels. The channels on the Farbwerk are dead easy to attach RGB strips to....but that's mostly not what I want. You can attach singles or strings of singles (like the eight I plan to have in the res) but you have to find your own common anode RGB LEDs (the normal ones available anywhere seem to be all common cathode) and solder in your own resistors depending on the power requirements of each colour. I'll post more details when I get that far.....assuming my sanity is still intact by then!

I'm also planning to de-lid the CPU, fill it with liquid metal, put the IHS back on and then use Kryonaut between the IHS and the block.....so if there are no more posts for a while, you know what happened! :eek:

So, before I tear everything apart, pictures of the fittings that turned up:



From left to right: A Barrow 90° fitting, same with a 12/10 EK-HDC hardline compression fitting and the hinge bridge that will have soft tubing on each end and should allow me to open the rear door without having to drain and disassemble.
 
Progress so far. Popped down the barbers to cadge a razor blade. Cheers guys. Sliced through the IHS glue very nicely but one 'ear' was a pain to get loose. Perseverance paid off though.






Tomorrow will be liquid metal time and I'll find out whether it survived! I'm not going to get a fair temperature comparison because the loop is going to be massively different - slightly larger rad :D

Also drilled out my 3mm holes to 5mm in the monoblock. 4.8mm drill seems to give a nice grip on the LED as it's slightly tapered.

 
I agree with you. I'm surprised Intel hasn't seen the demand for decent paste or solder from previous generations.
It's a fine line between brave and stupid...and it's not yet clear which side I've landed! ;) Without wishing to be prematurely blazé it wasn't as difficult as I thought it was going to be. The really thin (and sharp) razor seems to be the key. Most of it it just sliced through, it was just the one 'ear' that I had to keep working on. Found the best thing to clean off the black mastic was a fingernail. IPA and methanol didn't touch it.
Hoping to get a working system today - no pretty cabling yet though. Have to drop my son off to my mum's for the day first and glue all the things she's broken - it's usually angel statuettes *sigh* ;)
 
OJ46: Deliberately making something badly just doesn't sit right with me...even with the market economics behind it. Like you say, it'll be interesting if AMD can give them some competition again.

BornEvil: Good to know. I'd heard about 10C. My CPU max core temp hits the high 70's when playing a game that uses the CPU so a 10C drop would be nice. 8Pack commented in another thread that it's probably my board supplying too much voltage when the CPU is on turbo (4.4GHz all cores) so I probably need to sort that out but starting off by dropping the temp isn't going to be a bad thing! :D

Progress pictures :D Cleaned up the existing gunk and applied some Cool Laboratories Liquid Ultra (CLU)



Spread it out with the supplied brush. Was a bit worried it seems a little lumpy but it seems to be the case in their instruction videos as far as I can work out. Assume it'll be fine when squashed and/or heated. Also applied a bit of clear nail varnish (who said wives weren't useful! ;) ) over the uninsulated terminals in case the CLU spreads. Just because you're paranoid.... :D




....and assembled. The black wires are temp sensors I stuck on the bottom of the M.2 SSDs to keep an eye on them. As far as I can tell, once you've stuck them in RAID 0, there's no way of accessing that data. Temp from the top of the chip would be more accurate than the bottom of the PCB but I wasn't going to be sandwiching the probe between the chip and the heatsink.



Hope to find out whether it lives later today :D
 
So I've got a functional system now, right? Er, no....but you're not that surprised are you?! :D
Apart from work, I hit a couple of stumbling blocks. Just before I mounted the motherboard it occurred to me that I wouldn't be able to insert the LED on the IO panel side of the monoblock once the IO shield was screwed on and that I wouldn't be able to screw that on after mounting the motherboard. I had to make up an RGB LED on an arbitrary length of cable and sleeve it. I also had to assemble the two pumps and sleeve the two cables - I tried not to but I had to cut the molex off and put a 4-pin on anyway...so it would have been rude not to! :D I can always shorten them once I know final lengths.

Pump and LED cables each have four wires. Technically you can get four wires through a paracord sleeve. Should you? Well, only if you can answer "Hell no!" to the question "Do I value my sanity?"!

So motherboard is in and pump is ready. More progress tomorrow afternoon.
 
Progress? Er, some. It's been one of those days where every time you try to get something done someone interrupts. I've got a bend done from cpu to gpu and most of one from cpu up and over the motherboard tray. Must have got through half my tubing and I'm not sure I'm happy with the bends I've got to be honest. Had to make up a bending guide for the up-and-over because the 2nd bend was too close to the first one. Will try to get some pictures up tomorrow.
 
whiterabbit said:
Subbed! Might give this delidding lark a go.
Click to expand...

Maybe wait until I confirm (or not!) that it still works before you take this as a good reference! :D I think it should as there's no damage been done.

orbitalwalsh said:
great board, but had issues with dual bios, and when you update the bios it only did one. In the end I just disabled dual and left it running with one. stopped the booting to default speeds etc.
Click to expand...

I didn't have many issues with mine but I did have with the identical board I used to build for a friend. I think that once you get a boot failure it flips to the other bios which then defaults and then you get stuck in a loop. A slightly older bios worked best for me - think it f6 but I can dig it out if you need it. Had issues with RAID 0'd M.2's otherwise. Like you I stuck his board in single bios mode but mine's still in dual for the minute.
 
Cheers. It's the holes I've drilled in it that really make it ;) I like the clear ones as you can see if/when it's getting clogged up with any gunk....although hopefully the filter going in should take care of that.
 
Just a mini update with hopefully more to come today.
I got the GPU to CPU hardline bent yesterday after a few false starts. Also got the CPU up-and-over bent after some misses, some fails, some snapping and some swearing :D



Looking to do most of the rest in soft tubing at the moment (unless I run out!) simply because it's only temporary so no point in making it harder than it needs be!

Res mount is mounted and I managed to find some existing holes - again, it's temporary so I don't want to be drilling more. Have got the res reassembled too - just need to leak test it before it goes in as I took apart the level sensor end I don't want 880ml of coolant in a leaky res! :eek:
Pump is assembled, sleeved and in place. Cables are too long at the moment but can be shortened Once I've got final positions for everything.



To get the pump at that height, I needed some legs. Some 4mm threaded rod and some 6mm aluminium tubing from B&Q plus a hacksaw equals legs :D Remembering the old adage: measure twice, cut once, swear once and drive to ScrewFix for additional washers so the inlet fitting doesn't hit the PSU shroud :rolleyes:




...and mounted. Why six legs? Well, clearly six is better than four and I couldn't get the additional two in without drilling holes in the bottom that I don't need there long term.



Well, back to the grindstone. More later, wife permitting! :D
 
And it.....leaks! Bit of a leak round the bottom of the glass that was sorted by tightening it up but also a weeping round the level sensor. Hopefully a smear of LS-X should sort it out. Will find out when it sets.

First time I've ever leak tested something outside the case...looks like it was a good job I did!
 
Well over 12 hours later the LS-X hasn't set. I dunno, you have a tube lying in your water cooling box for just ten short years and what's in the bottom of the tube, after you've drilled through the solid stuff at the top, doesn't set! I demand my money back! ;) Well, off to buy another tube and then its wash, rinse, repeat.

Also found that I'm going to need more Tygon and some more fittings for the rather awkward angles so I'm not going to be able to complete the loop until that arrives :mad: Will try to post some progress of mounting the bottom half of the Aquaero and the flow meter in the meantime.
 
Actually that's not a bad idea. Normally I go with the kitchen towel method but it might be reassuring to test my dodgy-looking up-and-over acrylic bend!
Might be able to work out a way of getting them to send a replacement rubber grommet thing for the res to sort it properly.
 
New tube seem a to have set better. My old tube I had to drill through the solid stuff about 3" to coax out some unset stuff. I'm guessing what I got hadn't set because it was never going to. Did seem to stop the leak ironically...but wasn't happy leaving it as mush.
Have ordered a Dr Drop (cheers for the suggestion OJ46), some more Tygon and enough fittings to (hopefully!) finish the loop and find out if the delidding worked!
 
You must log in or register to reply here.
Back
Top Bottom