The Utterly Imbalanced CaseLabs TX10-D build(s)!
- Thread starter vsg28
- Start date
Associate
- Joined
- 9 Jan 2010
- Posts
- 739
- Location
- Sunny Brizzol
Awesome mega crazy case, waiting to see how the beast turns out 
Does this build roll around or just hover around from all the fans lol
I like that idea
LOLDoes this build roll around or just hover around from all the fans lol
Associate
- Joined
- 10 Jan 2004
- Posts
- 155
This is epic!
So the build log just got sponsored (partially) by a US watercooling retailer. This definitely will help when I need to order more pumps, fittings etc. Mick from Mayhems also sent along a Blitz Basic to go along with the Blitz Pro I had previously bought so I got plenty of Part 2 here. Unfortunately the US customs meant Mick had to change his Part 1 solution so I only get 1 liter effective instead of the 2 liters normally. This simply is not enough to clean out all the rads I have. Now I need to see if picking 2-3 more Blitz Pro kits is worth it just for Part 1.
Thanks. As promised, here is a massive update. Now this relates mostly to fan control so if you are not interested then please skip past those sections.
Now then, a couple of international packages came in this past week. First up:
Now what could that be? Oh ya, replacement brackets for the Aquaero 6 so they actually do fit in a standard 5 1/4" bay.
Those didn't look any different from the naked eye, or dSLR lens for that matter. In fact, here is an image with one of the old and new brackets each next to each other. Imaginary brownie points for those who can tell which one is which:
Got them installed on the AQ6, man that red heatsink looks very nice!
But does it actually help out? It was hard to fit the AQ6 in earlier, the display front plate used to bulge out a bit and I was just about to order the Caselabs short non-confirming mounts when I realized I might as well make good use of Shoggy's offer of replacement brackets. This way, I can be rest assured that this would work with any case for anyone in the future- if it does work, that is!
Yes!! Thanks Shoggy, it was a massive relief getting this in easily with no real effort.
Before I go to the next international package, I had something in mind with the Aquaero, Darlene's Diva 'Dapter which was referenced previously in here and the Corsair SP120 "PWM" fans. My other PSU, the Corsair ax1200i, is already in the case so I figured what better way than this to test out the EVGA 1300G2 that I had recently bought:
Customary EVGA logo close up shot above. Speaking of close up shots and small objects, here is the NZXT Grid which is currently at a great price on Amazon (US website anyway) next to a Nexus 4 for comparison. The thing is much smaller in person than what I expected from images:
10 channel 3-pin fan hub, each channel capable of supplying upto 3 watts. It comes with all the cables needed as well. So at <$10 currently, this has to be a steal- not really. Thanks to an OCN member, I realized that the rpm (tach) signal wire is split to all 10 ports in this making it useless to use with most fan controllers because the rpm speed inputs from all of the fans will corrupt the signal and the fan speed will not be correctly reported to the controller. Even cheap 3-pin fan splitter cables will only have all three wires going to one of the fan plugs, and just the 12v and ground to the rest. So I decided to open 'er up to see if newer batches had this corrected by any chance.
Unfortunately not, while faint one can still see the RPM trace going across from channel to channel on the PCB.. Oh well, there was already a solution provided above so I did the same, i.e, cut off the RPM signal prong from 9 of the 10 ports and make sure that single port with the RPM prong is occupied:
Problem solved, the Aquaero displayed RPMs correctly now. So I now started testing the real issue I had: Corsair's PWM implementation on their SP120 fans. Feel free to skip this section if this is not relevant to you. First up, the SP120 QE
Hooked directly to the Aquaero, at 50% PWM control:
Now connected to Channel 1 of the Swiftech 8 way PWM splitter, which is powered straight from the 1300G2:
Ok, that's close enough. Let's add in another fan:
We already start to see the effect of a non-standard PWM implementation. So now with fan 3 added:
Fan 4:
Fan 5:
Getting bad now, what about 6 fans on the splitter?
Ouch! Now adding fan 7:
Full speed already, no control whatsoever possible. Finally, fan 8:
Nothing changes as expected. So can the Double D help out? I hooked up the splitter to the adapter and had the passthrough over to the Aquaero.
Now with all 8 channels filled, and at the same 50% PWM signal:
Darlene is a magician- or someone with a very good electrical engineering background. Personally I prefer the former. Control is back on, the only change being a lower RPM compared to when a single fan was hooked up directly. I can live with that, I don't imagine having to run full speed anyway. The QEs are pretty quiet (who would have thought!) so let's try out a bigger challenge- the Corsair SP120 HPE fans.
First up, a fan hooked directly to the Aquaero at 50% control:
Now the same fan hooked via the Swiftech splitter:
Identical for all intents. Let's add fan 2:
Sigh.. Why did you have to do this, Corsair? Oh that's right- proprietary implementation + making a controller that is the only thing to work with it = Profit :thumbsdow
Fan 3 added:
Fan 4 added:
Now with fan 5:
and the dreaded 6th fan:
As expected, a massive jump with 6 fans on. A few users have reported having no control with 6 fans but I have a wee bit left as was the case with the QE fans. But now with fan 7:
As before, everything now runs at ~100% speed no matter what. Adding in the final 8th fan does nothing really different:
Does the 'Dapter help here too?
Yes, it does. As with the previous case, the RPMs are lowered a bit. That isn't an issue for me as I said before but I am curious as to why this is happening. Moving on, I hooked up both splitters to the adapter PCB and then to the AQ6:
No need to repeat the results here, it was the same as before. So I am confident the PCB will perform as intended with all channels occupied. In fact, this is so good that daisychaining also works. I hooked up 15 fans total (8 HPE, 7 QE) to the adapter and then to 1 channel on the AQ6 and it worked great. There is a caveat though- if I hooked up the 8 HPEs to 1 channel on the Swiftech splitter that housed 7 other QE fans, then there was no control available. If I daisychained it so that the QEs were hooked to the splitter that had the HPE fans, it was fine. Another point of curiosity but since I don't have any plans on having these different fans on the same channel, it does not affect me. Hopefully it will help others who want to try this.
Tl,dr.: Darlene's Diva 'Dapter works great. If you have a PWM controller (Newer Aquaero 6 units have this taken care of apparently) and have Corsair's PWM fans, chances are it won't work as expected. This will help.
Now to the other international package, Mayhems sent me a Blitz Basic kit for coming in close to winning a contest a few weeks back. I had already purchased a Blitz Pro kit from PPC so here they are together:
Blitz Basic has only his Part 2 solution which is enough for a 4 L effective cleaning solution of a loop. Now if you have rads that you suspect haven't been cleaned well from the factory *cough..Alphacool, Phobya..cough* or have purchased 2nd hand rads which have had dyes/colored coolant run through them, I would recommend using Part 1 on the rads alone. Here is the issue I have: Thanks to US customs, Mick was forced to change the formulation of Part 1 so it only gives 1 L effective as opposed to 2 L everywhere else. The Alphacool Monsta 480 alone consumes over 650 mL filled and I got a Mo.Ra 3 9.140 and Black Ice SR 1 560 that were used by a previous owner which I want to clean up as well. While I am at it, I might as well clean the other rads I have too. So I would imagine needing 3-4 L effective Part 1 (I have numbers for all the rads but that Mo.Ra) which means needing another 2-3 Blitz Pro kits just for more Part 1. I am tempted to just do the Monsta and and SR 1 rads now and have the rest cleaned with a dilute phosphoric acid solution but if anyone has suggestions otherwise, I am all ears.
Now then, a couple of international packages came in this past week. First up:
Now what could that be? Oh ya, replacement brackets for the Aquaero 6 so they actually do fit in a standard 5 1/4" bay.
Those didn't look any different from the naked eye, or dSLR lens for that matter. In fact, here is an image with one of the old and new brackets each next to each other. Imaginary brownie points for those who can tell which one is which:
Got them installed on the AQ6, man that red heatsink looks very nice!
But does it actually help out? It was hard to fit the AQ6 in earlier, the display front plate used to bulge out a bit and I was just about to order the Caselabs short non-confirming mounts when I realized I might as well make good use of Shoggy's offer of replacement brackets. This way, I can be rest assured that this would work with any case for anyone in the future- if it does work, that is!
Yes!! Thanks Shoggy, it was a massive relief getting this in easily with no real effort.
Before I go to the next international package, I had something in mind with the Aquaero, Darlene's Diva 'Dapter which was referenced previously in here and the Corsair SP120 "PWM" fans. My other PSU, the Corsair ax1200i, is already in the case so I figured what better way than this to test out the EVGA 1300G2 that I had recently bought:
Customary EVGA logo close up shot above. Speaking of close up shots and small objects, here is the NZXT Grid which is currently at a great price on Amazon (US website anyway) next to a Nexus 4 for comparison. The thing is much smaller in person than what I expected from images:
10 channel 3-pin fan hub, each channel capable of supplying upto 3 watts. It comes with all the cables needed as well. So at <$10 currently, this has to be a steal- not really. Thanks to an OCN member, I realized that the rpm (tach) signal wire is split to all 10 ports in this making it useless to use with most fan controllers because the rpm speed inputs from all of the fans will corrupt the signal and the fan speed will not be correctly reported to the controller. Even cheap 3-pin fan splitter cables will only have all three wires going to one of the fan plugs, and just the 12v and ground to the rest. So I decided to open 'er up to see if newer batches had this corrected by any chance.
Unfortunately not, while faint one can still see the RPM trace going across from channel to channel on the PCB.. Oh well, there was already a solution provided above so I did the same, i.e, cut off the RPM signal prong from 9 of the 10 ports and make sure that single port with the RPM prong is occupied:
Problem solved, the Aquaero displayed RPMs correctly now. So I now started testing the real issue I had: Corsair's PWM implementation on their SP120 fans. Feel free to skip this section if this is not relevant to you. First up, the SP120 QE
Hooked directly to the Aquaero, at 50% PWM control:
Now connected to Channel 1 of the Swiftech 8 way PWM splitter, which is powered straight from the 1300G2:
Ok, that's close enough. Let's add in another fan:
We already start to see the effect of a non-standard PWM implementation. So now with fan 3 added:
Fan 4:
Fan 5:
Getting bad now, what about 6 fans on the splitter?
Ouch! Now adding fan 7:
Full speed already, no control whatsoever possible. Finally, fan 8:
Nothing changes as expected. So can the Double D help out? I hooked up the splitter to the adapter and had the passthrough over to the Aquaero.
Now with all 8 channels filled, and at the same 50% PWM signal:
Darlene is a magician- or someone with a very good electrical engineering background. Personally I prefer the former. Control is back on, the only change being a lower RPM compared to when a single fan was hooked up directly. I can live with that, I don't imagine having to run full speed anyway. The QEs are pretty quiet (who would have thought!) so let's try out a bigger challenge- the Corsair SP120 HPE fans.
First up, a fan hooked directly to the Aquaero at 50% control:
Now the same fan hooked via the Swiftech splitter:
Identical for all intents. Let's add fan 2:
Sigh.. Why did you have to do this, Corsair? Oh that's right- proprietary implementation + making a controller that is the only thing to work with it = Profit :thumbsdow
Fan 3 added:
Fan 4 added:
Now with fan 5:
and the dreaded 6th fan:
As expected, a massive jump with 6 fans on. A few users have reported having no control with 6 fans but I have a wee bit left as was the case with the QE fans. But now with fan 7:
As before, everything now runs at ~100% speed no matter what. Adding in the final 8th fan does nothing really different:
Does the 'Dapter help here too?
Yes, it does. As with the previous case, the RPMs are lowered a bit. That isn't an issue for me as I said before but I am curious as to why this is happening. Moving on, I hooked up both splitters to the adapter PCB and then to the AQ6:
No need to repeat the results here, it was the same as before. So I am confident the PCB will perform as intended with all channels occupied. In fact, this is so good that daisychaining also works. I hooked up 15 fans total (8 HPE, 7 QE) to the adapter and then to 1 channel on the AQ6 and it worked great. There is a caveat though- if I hooked up the 8 HPEs to 1 channel on the Swiftech splitter that housed 7 other QE fans, then there was no control available. If I daisychained it so that the QEs were hooked to the splitter that had the HPE fans, it was fine. Another point of curiosity but since I don't have any plans on having these different fans on the same channel, it does not affect me. Hopefully it will help others who want to try this.
Tl,dr.: Darlene's Diva 'Dapter works great. If you have a PWM controller (Newer Aquaero 6 units have this taken care of apparently) and have Corsair's PWM fans, chances are it won't work as expected. This will help.
Now to the other international package, Mayhems sent me a Blitz Basic kit for coming in close to winning a contest a few weeks back. I had already purchased a Blitz Pro kit from PPC so here they are together:
Blitz Basic has only his Part 2 solution which is enough for a 4 L effective cleaning solution of a loop. Now if you have rads that you suspect haven't been cleaned well from the factory *cough..Alphacool, Phobya..cough* or have purchased 2nd hand rads which have had dyes/colored coolant run through them, I would recommend using Part 1 on the rads alone. Here is the issue I have: Thanks to US customs, Mick was forced to change the formulation of Part 1 so it only gives 1 L effective as opposed to 2 L everywhere else. The Alphacool Monsta 480 alone consumes over 650 mL filled and I got a Mo.Ra 3 9.140 and Black Ice SR 1 560 that were used by a previous owner which I want to clean up as well. While I am at it, I might as well clean the other rads I have too. So I would imagine needing 3-4 L effective Part 1 (I have numbers for all the rads but that Mo.Ra) which means needing another 2-3 Blitz Pro kits just for more Part 1. I am tempted to just do the Monsta and and SR 1 rads now and have the rest cleaned with a dilute phosphoric acid solution but if anyone has suggestions otherwise, I am all ears.
Last edited:
I took a break to watch the Google I/O event and then got a few more things done.
There was always doubt of the best orientation of a CPU block, especially in the case of the Koolance 380i where Stren and a few other reviewers noted a difference of 2-3 ºC from orientation. Now one can attribute this to a lot of things but I wanted to be sure. I had the block oriented as shown below assuming that was the way to go:
Since I wanted to replace the Arctic Cooling MX-4 paste on there with Thermene anyway, I called up Koolance and got the ok to open up the block without voiding any warranty.
Not too bad a spread, but let's get rid of it. One of the benefits of having access to chemicals is this:
That's much better. Now let's take a look inside that 380i:
This purple stuff along with the left over coolant/dirt was quickly taken care of. Some of that purple stuff (anyone have any idea what it is?) was still left behind despite my best efforts but not enough to worry me unless someone can confirm that as possible corrosion. But seeing how the jetplate was oriented, and knowing that the 4770k has the die oriented vertically, I knew I had the orientation wrong previously. Now let's have some Thermene on:
Assembling the block is way easier with the backplate already on:
With the CPU taken care of, I now had one KPE left to put the water block on. Removing the stock heatsink is very easy, and then immediately we get to see the factory TIM application:
Not as bad as was the case with my Asus 290x but definitely can be better. Got it cleaned up, and then it was time to introduce it to the EK FC 780 GTX Classy (Shouldn't it have been the EK FC GTX 780 Classy? Oh well!):
All done, and with the stock backplate on again it was time to put it back into place for now:
It will come back out soon since I will be testing those cards individually first before putting them together. But for now, it is ok. Meanwhile, a package arrived from Stren of XtremeRigs.net:
Two pieces of his Project Thief! A 480mm drop-in rad mount that will be used on the XSPC AX480 and a 180.3 drop-in rad mount just in case I decided to expand side 2 with GPUs. I also bought this off him:
By itself, it makes no sense. But combine it with the massive TX10-D HDD super mount, and things get clearer:
So that's one cage taken care of, and there is still enough space for 4 more such cages if need be. That's it for now, I will install in some of the drives in this cage soon and put the assembly in place. I really need to get the rads cleaned up too. Hopefully more TX10 parts are on the way to me from another source!
There was always doubt of the best orientation of a CPU block, especially in the case of the Koolance 380i where Stren and a few other reviewers noted a difference of 2-3 ºC from orientation. Now one can attribute this to a lot of things but I wanted to be sure. I had the block oriented as shown below assuming that was the way to go:
Since I wanted to replace the Arctic Cooling MX-4 paste on there with Thermene anyway, I called up Koolance and got the ok to open up the block without voiding any warranty.
Not too bad a spread, but let's get rid of it. One of the benefits of having access to chemicals is this:
That's much better. Now let's take a look inside that 380i:
This purple stuff along with the left over coolant/dirt was quickly taken care of. Some of that purple stuff (anyone have any idea what it is?) was still left behind despite my best efforts but not enough to worry me unless someone can confirm that as possible corrosion. But seeing how the jetplate was oriented, and knowing that the 4770k has the die oriented vertically, I knew I had the orientation wrong previously. Now let's have some Thermene on:
Assembling the block is way easier with the backplate already on:
With the CPU taken care of, I now had one KPE left to put the water block on. Removing the stock heatsink is very easy, and then immediately we get to see the factory TIM application:
Not as bad as was the case with my Asus 290x but definitely can be better. Got it cleaned up, and then it was time to introduce it to the EK FC 780 GTX Classy (Shouldn't it have been the EK FC GTX 780 Classy? Oh well!):
All done, and with the stock backplate on again it was time to put it back into place for now:
It will come back out soon since I will be testing those cards individually first before putting them together. But for now, it is ok. Meanwhile, a package arrived from Stren of XtremeRigs.net:
Two pieces of his Project Thief! A 480mm drop-in rad mount that will be used on the XSPC AX480 and a 180.3 drop-in rad mount just in case I decided to expand side 2 with GPUs. I also bought this off him:
By itself, it makes no sense. But combine it with the massive TX10-D HDD super mount, and things get clearer:
So that's one cage taken care of, and there is still enough space for 4 more such cages if need be. That's it for now, I will install in some of the drives in this cage soon and put the assembly in place. I really need to get the rads cleaned up too. Hopefully more TX10 parts are on the way to me from another source!
So I previously unboxed the EVGA 1300G2 but as fate would have it, I got this yesterday:
Ok that wasn't exactly the halo effect I wanted so let's get right to opening up the EVGA 1600G2:
Look at all these cables:
That's enough PCI-E power for 3 MSI 290x Lightnings/EVGA 780Ti Kingpins without needing a daisy chained cable
Oh, I forgot to mention I also got a 750G2 for Side 2. So here's a EVGA fanboy shot:
But wait, there's MORE:
So how do they look all mounted on the PSU mount?
~5000 W of PSU in there, enough to burn down any circuit I have in this place. The ax1200i is pretty much spoken for already but the 1300G2 needs a new owner who will care for it and keep it longer than the ~6-7 days it has been opened
Anyway, so here's the final PSU configuration:
The EVGA 1600G2 for Side 1 and EVGA 750G2 for Side 2. I got these both at a tremendous price so no complaints. But not everything worked out as planned, I had a set of Lamptron handles that I was hoping to install on the PSU mount to make it "drop-in" style too. Unfortunately the only place I can have them without interfering on the 4 mounting locations are the extreme ends and the handle holes match exactly the screw in locations of the mount to the case:
So there is no way I can think of making this a drop-in mount unless I am missing something obvious. Oh well! Next up, this goes in the case and I see if the stock cables are long enough. Side 2 has the same motherboard that I currently have but for Side 1 I have to assume a similar layout and see.
Ok that wasn't exactly the halo effect I wanted so let's get right to opening up the EVGA 1600G2:
Look at all these cables:
That's enough PCI-E power for 3 MSI 290x Lightnings/EVGA 780Ti Kingpins without needing a daisy chained cable
Oh, I forgot to mention I also got a 750G2 for Side 2. So here's a EVGA fanboy shot:
But wait, there's MORE:
So how do they look all mounted on the PSU mount?
~5000 W of PSU in there, enough to burn down any circuit I have in this place. The ax1200i is pretty much spoken for already but the 1300G2 needs a new owner who will care for it and keep it longer than the ~6-7 days it has been opened
Anyway, so here's the final PSU configuration:
The EVGA 1600G2 for Side 1 and EVGA 750G2 for Side 2. I got these both at a tremendous price so no complaints. But not everything worked out as planned, I had a set of Lamptron handles that I was hoping to install on the PSU mount to make it "drop-in" style too. Unfortunately the only place I can have them without interfering on the 4 mounting locations are the extreme ends and the handle holes match exactly the screw in locations of the mount to the case:
So there is no way I can think of making this a drop-in mount unless I am missing something obvious. Oh well! Next up, this goes in the case and I see if the stock cables are long enough. Side 2 has the same motherboard that I currently have but for Side 1 I have to assume a similar layout and see.
Another surprise, this time of the good kind: UPS man showed up way earlier today compared to 8 pm last night with the 1600/750G2 units. Here is what I got:
Definitely need to gift the UPS man come Christmas time. Above is 1 XXL window door for Side 1 (because XL just isn't enough to show off 2 FrozenQ reservoirs and the respective loops), 2 480mm rad mounts and a few flexbay/HDD mounts thrown in. Now I have enough for all my rads as well as enough for an additional 480mm and 540mm (180x3) rad if need be. Come to think of it, I got the space for those two as well. Hmmm....
Now I need to clean up the rads with the Mayhems cleaning kits and my own dilute phosphoric acid solution and start putting them all in. I have also decided to go hardline in the motherboard area, at least for Side 1. Copper or acrylic is the question now. What do you guys think?
Definitely need to gift the UPS man come Christmas time. Above is 1 XXL window door for Side 1 (because XL just isn't enough to show off 2 FrozenQ reservoirs and the respective loops), 2 480mm rad mounts and a few flexbay/HDD mounts thrown in. Now I have enough for all my rads as well as enough for an additional 480mm and 540mm (180x3) rad if need be. Come to think of it, I got the space for those two as well. Hmmm....
Now I need to clean up the rads with the Mayhems cleaning kits and my own dilute phosphoric acid solution and start putting them all in. I have also decided to go hardline in the motherboard area, at least for Side 1. Copper or acrylic is the question now. What do you guys think?
I have a buyer for the ax1200i more or less confirmed but still looking to sell the 1300G2. Trying it out locally first since shipping this will be pretty expensive.
The 750G2 is for Side 2, I love the eco mode on it and combined with QE fans and a silent pump that entire rig should be whisper quiet. The 1600G2 is for when I bench the 5930k/5960x and 2 EVGA Kingpins which will consume 1400+ watts easily. I got this PSU at $199 and possibly another $25 off in rebates soon thanks to an EVGA contest so I am not complaining.
The 750G2 is for Side 2, I love the eco mode on it and combined with QE fans and a silent pump that entire rig should be whisper quiet. The 1600G2 is for when I bench the 5930k/5960x and 2 EVGA Kingpins which will consume 1400+ watts easily. I got this PSU at $199 and possibly another $25 off in rebates soon thanks to an EVGA contest so I am not complaining.
Nice to see the size comparison between the PSU's, I have the 1300G2 myself and found EVGA customer service are great when I messed up a cable trying to sleeve it, they sent me a new one for free.
Would try using EVGA as much as possible in the future for that reason alone.
On I side note they also now offer cable sets in various colours now for P & G PSU's which is handy if you are a clumsy sleever like me lol.
Would try using EVGA as much as possible in the future for that reason alone.
On I side note they also now offer cable sets in various colours now for P & G PSU's which is handy if you are a clumsy sleever like me lol.
It's not you, the EVGA cables are pretty hard to sleeve. I did consider the cable sets to be honest but I had the Corsair kit before and wasn't impressed at all. Mass produced kits tend to have imperfections.
There is a local custom sleever who I am talking to, he's done great work so far and I may ask him to make custom length cables using 16 AWG wiring for the PCI-E part and then just extensions for the rest. Not sure yet!
There is a local custom sleever who I am talking to, he's done great work so far and I may ask him to make custom length cables using 16 AWG wiring for the PCI-E part and then just extensions for the rest. Not sure yet!
Cheers! At the moment, I am cleaning all the rads with the Mayhems Blitz kits. Part 1 done on some of them, got a vinegar solution in the rest. Next up is part 2 and then it's done. Shame I had to dismantle the Mo.Ra and all fans for this, that will take a while putting back together.